SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
LatticeECP3 Family Data Sheet
Preliminary DS1021 Version 01.6, March 2010
LatticeECP3 Family Data Sheet
Introduction
November 2009
Preliminary Data Sheet DS1021
Features
• Dedicated read/write levelling functionality
• Dedicated gearing logic
• Source synchronous standards support
– ADC/DAC, 7:1 LVDS, XGMII
– High Speed ADC/DAC devices
• Dedicated DDR/DDR2/DDR3 memory with DQS
support
• Optional Inter-Symbol Interference (ISI)
correction on outputs
Higher Logic Density for Increased System
Integration
• 17K to 149K LUTs
• 133 to 586 I/Os
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Embedded SERDES
• 150 Mbps to 3.2 Gbps for Generic 8b10b, 10-bit
SERDES, and 8-bit SERDES modes
• Data Rates 230 Mbps to 3.2 Gbps per channel
for all other protocols
• Up to 16 channels per device: PCI Express,
SONET/SDH, Ethernet (1GbE, SGMII, XAUI),
CPRI, SMPTE 3G and Serial RapidIO
Programmable sysI/O™ Buffer Supports
Wide Range of Interfaces
•
•
•
•
•
•
•
sysDSP™
• Fully cascadable slice architecture
• 12 to 160 slices for high performance multiply
and accumulate
• Powerful 54-bit ALU operations
• Time Division Multiplexing MAC Sharing
• Rounding and truncation
• Each slice supports
– Half 36x36, two 18x18 or four 9x9 multipliers
– Advanced 18x36 MAC and 18x18 Multiply-
Multiply-Accumulate (MMAC) operations
On-chip termination
Optional equalization filter on inputs
LVTTL and LVCMOS 33/25/18/15/12
SSTL 33/25/18/15 I, II
HSTL15 I and HSTL18 I, II
PCI and Differential HSTL, SSTL
LVDS, Bus-LVDS, LVPECL, RSDS, MLVDS
Flexible Device Configuration
•
•
•
•
•
•
Flexible Memory Resources
Dedicated bank for configuration I/Os
SPI boot flash interface
Dual-boot images supported
Slave SPI
TransFR™ I/O for simple field updates
Soft Error Detect embedded macro
System Level Support
• Up to 6.85Mbits sysMEM™ Embedded Block
RAM (EBR)
• 36K to 303K bits distributed RAM
•
•
•
•
•
sysCLOCK Analog PLLs and DLLs
• Two DLLs and up to ten PLLs per device
Pre-Engineered Source Synchronous I/O
IEEE 1149.1 and IEEE 1532 compliant
Reveal Logic Analyzer
ORCAstra FPGA configuration utility
On-chip oscillator for initialization & general use
1.2V core power supply
• DDR registers in I/O cells
Table 1-1. LatticeECP3™ Family Selection Guide
Device
ECP3-17
ECP3-35
ECP3-70
ECP3-95
ECP3-150
LUTs (K)
17
33
67
92
149
sysMEM Blocks (18Kbits)
38
72
240
240
372
Embedded Memory (Kbits)
700
1327
4420
4420
6850
Distributed RAM Bits (Kbits)
36
68
145
188
303
18X18 Multipliers
24
64
128
128
320
SERDES (Quad)
1
1
3
3
4
2/2
4/2
10 / 2
10 / 2
10 / 2
4 / 295
4 / 295
PLLs/DLLs
Packages and SERDES Channels/ I/O Combinations
256 ftBGA (17x17 mm)
4 / 133
4 / 133
484 fpBGA (23x23 mm)
4 / 222
4 / 295
672 fpBGA (27x27 mm)
4 / 310
1156 fpBGA (35x35 mm)
8 / 380
8 / 380
8 / 380
12 / 490
12 / 490
16 / 586
© 2009 Lattice Semiconductor Corp. All Lattice trademarks, registered trademarks, patents, and disclaimers are as listed at www.latticesemi.com/legal. All other brand
or product names are trademarks or registered trademarks of their respective holders. The specifications and information herein are subject to change without notice.
www.latticesemi.com
1-1
DS1021 Introduction_01.3
Introduction
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Introduction
The LatticeECP3™ (EConomy Plus Third generation) family of FPGA devices is optimized to deliver high performance features such as an enhanced DSP architecture, high speed SERDES and high speed source synchronous
interfaces in an economical FPGA fabric. This combination is achieved through advances in device architecture
and the use of 65nm technology making the devices suitable for high-volume, high-speed, low-cost applications.
The LatticeECP3 device family expands look-up-table (LUT) capacity to 149K logic elements and supports up to
486 user I/Os. The LatticeECP3 device family also offers up to 320 18x18 multipliers and a wide range of parallel
I/O standards.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
The LatticeECP3 FPGA fabric is optimized with high performance and low cost in mind. The LatticeECP3 devices
utilize reconfigurable SRAM logic technology and provide popular building blocks such as LUT-based logic, distributed and embedded memory, Phase Locked Loops (PLLs), Delay Locked Loops (DLLs), pre-engineered source
synchronous I/O support, enhanced sysDSP slices and advanced configuration support, including encryption and
dual-boot capabilities.
The pre-engineered source synchronous logic implemented in the LatticeECP3 device family supports a broad
range of interface standards, including DDR3, XGMII and 7:1 LVDS.
The LatticeECP3 device family also features high speed SERDES with dedicated PCS functions. High jitter tolerance and low transmit jitter allow the SERDES plus PCS blocks to be configured to support an array of popular
data protocols including PCI Express, SMPTE, Ethernet (XAUI, GbE, and SGMII) and CPRI. Transmit Pre-emphasis and Receive Equalization settings make the SERDES suitable for transmission and reception over various
forms of media.
The LatticeECP3 devices also provide flexible, reliable and secure configuration options, such as dual-boot capability, bit-stream encryption, and TransFR field upgrade features.
The ispLEVER® design tool suite from Lattice allows large complex designs to be efficiently implemented using the
LatticeECP3 FPGA family. Synthesis library support for LatticeECP3 is available for popular logic synthesis tools.
The ispLEVER tool uses the synthesis tool output along with the constraints from its floor planning tools to place
and route the design in the LatticeECP3 device. The ispLEVER tool extracts the timing from the routing and backannotates it into the design for timing verification.
Lattice provides many pre-engineered IP (Intellectual Property) ispLeverCORE™ modules for the LatticeECP3
family. By using these configurable soft core IPs as standardized blocks, designers are free to concentrate on the
unique aspects of their design, increasing their productivity.
1-2
LatticeECP3 Family Data Sheet
Architecture
March 2010
Preliminary Data Sheet DS1021
Architecture Overview
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Each LatticeECP3 device contains an array of logic blocks surrounded by Programmable I/O Cells (PIC). Interspersed between the rows of logic blocks are rows of sysMEM™ Embedded Block RAM (EBR) and rows of sysDSP™ Digital Signal Processing slices, as shown in Figure 2-1. In addition, the LatticeECP3 family contains
SERDES Quads on the bottom of the device.
There are two kinds of logic blocks, the Programmable Functional Unit (PFU) and Programmable Functional Unit
without RAM (PFF). The PFU contains the building blocks for logic, arithmetic, RAM and ROM functions. The PFF
block contains building blocks for logic, arithmetic and ROM functions. Both PFU and PFF blocks are optimized for
flexibility, allowing complex designs to be implemented quickly and efficiently. Logic Blocks are arranged in a twodimensional array. Only one type of block is used per row.
The LatticeECP3 devices contain one or more rows of sysMEM EBR blocks. sysMEM EBRs are large, dedicated
18Kbit fast memory blocks. Each sysMEM block can be configured in a variety of depths and widths as RAM or
ROM. In addition, LatticeECP3 devices contain up to two rows of DSP slices. Each DSP slice has multipliers and
adder/accumulators, which are the building blocks for complex signal processing capabilities.
The LatticeECP3 devices feature up to 16 embedded 3.2Gbps SERDES (Serializer / Deserializer) channels. Each
SERDES channel contains independent 8b/10b encoding / decoding, polarity adjust and elastic buffer logic. Each
group of four SERDES channels, along with its Physical Coding Sub-layer (PCS) block, creates a quad. The functionality of the SERDES/PCS quads can be controlled by memory cells set during device configuration or by registers that are addressable during device operation. The registers in every quad can be programmed via the
SERDES Client Interface (SCI). These quads (up to four) are located at the bottom of the devices.
Each PIC block encompasses two PIOs (PIO pairs) with their respective sysI/O buffers. The sysI/O buffers of the
LatticeECP3 devices are arranged in seven banks, allowing the implementation of a wide variety of I/O standards.
In addition, a separate I/O bank is provided for the programming interfaces. 50% of the PIO pairs on the left and
right edges of the device can be configured as LVDS transmit/receive pairs. The PIC logic also includes pre-engineered support to aid in the implementation of high speed source synchronous standards such as XGMII, 7:1
LVDS, along with memory interfaces including DDR3.
Other blocks provided include PLLs, DLLs and configuration functions. The LatticeECP3 architecture provides two
Delay Locked Loops (DLLs) and up to ten Phase Locked Loops (PLLs). In addition, each LatticeECP3 family member provides two DLLs per device. The PLL and DLL blocks are located at the end of the EBR/DSP rows.
The configuration block that supports features such as configuration bit-stream decryption, transparent updates
and dual-boot support is located toward the center of this EBR row. Every device in the LatticeECP3 family supports a sysCONFIG™ port located in the corner between banks one and two, which allows for serial or parallel
device configuration.
In addition, every device in the family has a JTAG port. This family also provides an on-chip oscillator and soft error
detect capability. The LatticeECP3 devices use 1.2V as their core voltage.
© 2010 Lattice Semiconductor Corp. All Lattice trademarks, registered trademarks, patents, and disclaimers are as listed at www.latticesemi.com/legal. All other brand
or product names are trademarks or registered trademarks of their respective holders. The specifications and information herein are subject to change without notice.
www.latticesemi.com
2-1
DS1021 Architecture_01.5
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-1. Simplified Block Diagram, LatticeECP3-35 Device (Top Level)
sysIO
Bank 0
sysIO
Bank 1
Configuration Logic:
Dual-boot, Encryption
and Transparent Updates
JTAG
On-chip Oscillator
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Pre-engineered Source
Synchronous Support:
DDR3 - 800Mbps
Generic - Up to 1Gbps
sysIO
Bank
7
sysIO
Bank
2
Enhanced DSP
Slices: Multiply,
Accumulate and ALU
sysCLOCK
PLLs & DLLs:
Frequency Synthesis
and Clock Alignment
Flexible sysIO:
LVCMOS, HSTL,
SSTL, LVDS
Up to 486 I/Os
sysMEM Block
RAM: 18Kbit
Flexible Routing:
Optimized for speed
and routability
Programmable
Function Units:
Up to 149K LUTs
SERDES/PCS SERDES/PCS
CH 3
CH 2
SERDES/PCS SERDES/PCS
CH 1
CH 0
sysIO Bank 6
sysIO Bank 3
3.2Gbps SERDES
Note: There is no Bank 4 or Bank 5 in LatticeECP3 devices.
PFU Blocks
The core of the LatticeECP3 device consists of PFU blocks, which are provided in two forms, the PFU and PFF.
The PFUs can be programmed to perform Logic, Arithmetic, Distributed RAM and Distributed ROM functions. PFF
blocks can be programmed to perform Logic, Arithmetic and ROM functions. Except where necessary, the remainder of this data sheet will use the term PFU to refer to both PFU and PFF blocks.
Each PFU block consists of four interconnected slices numbered 0-3 as shown in Figure 2-2. Each slice contains
two LUTs. All the interconnections to and from PFU blocks are from routing. There are 50 inputs and 23 outputs
associated with each PFU block.
2-2
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-2. PFU Diagram
From
Routing
LUT4 &
CARRY
LUT4 &
CARRY
LUT4 &
CARRY
LUT4 &
CARRY
Slice 1
LUT4 &
CARRY
LUT4
LUT4
Slice 3
Slice 2
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Slice 0
LUT4 &
CARRY
D
FF
D
FF
D
FF
D
D
FF
FF
D
FF
To
Routing
Slice
Slice 0 through Slice 2 contain two LUT4s feeding two registers, whereas Slice 3 contains two LUT4s only. For
PFUs, Slice 0 through Slice 2 can be configured as distributed memory, a capability not available in the PFF.
Table 2-1 shows the capability of the slices in both PFF and PFU blocks along with the operation modes they
enable. In addition, each PFU contains logic that allows the LUTs to be combined to perform functions such as
LUT5, LUT6, LUT7 and LUT8. There is control logic to perform set/reset functions (programmable as synchronous/
asynchronous), clock select, chip-select and wider RAM/ROM functions.
Table 2-1. Resources and Modes Available per Slice
PFU BLock
Slice
Resources
PFF Block
Modes
Resources
Modes
Slice 0
2 LUT4s and 2 Registers Logic, Ripple, RAM, ROM 2 LUT4s and 2 Registers
Logic, Ripple, ROM
Slice 1
2 LUT4s and 2 Registers Logic, Ripple, RAM, ROM 2 LUT4s and 2 Registers
Logic, Ripple, ROM
Slice 2
2 LUT4s and 2 Registers Logic, Ripple, RAM, ROM 2 LUT4s and 2 Registers
Logic, Ripple, ROM
Slice 3
2 LUT4s
Logic, ROM
2 LUT4s
Logic, ROM
Figure 2-3 shows an overview of the internal logic of the slice. The registers in the slice can be configured for positive/negative and edge triggered or level sensitive clocks.
Slices 0, 1 and 2 have 14 input signals: 13 signals from routing and one from the carry-chain (from the adjacent
slice or PFU). There are seven outputs: six to routing and one to carry-chain (to the adjacent PFU). Slice 3 has 10
input signals from routing and four signals to routing. Table 2-2 lists the signals associated with Slice 0 to Slice 2.
2-3
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-3. Slice Diagram
FCO To Different Slice/PFU
SLICE
FXB
FXA
OFX1
A1
B1
C1
D1
CO
F1
F/SUM
Q1
D
LUT4 &
CARRY*
FF*
To
Routing
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
CI
M1
M0
LUT5
Mux
From
Routing
OFX0
A0
B0
C0
D0
CO
LUT4 &
CARRY*
F0
F/SUM
Q0
D
FF*
CI
CE
CLK
LSR
* Not in Slice 3
FCI From Different Slice/PFU
For Slices 0 and 1, memory control signals are generated from Slice 2 as follows:
WCK is CLK
WRE is from LSR
DI[3:2] for Slice 1 and DI[1:0] for Slice 0 data from Slice 2
WAD [A:D] is a 4-bit address from slice 2 LUT input
Table 2-2. Slice Signal Descriptions
Function
Type
Signal Names
Input
Data signal
A0, B0, C0, D0
Inputs to LUT4
Description
Input
Data signal
A1, B1, C1, D1
Inputs to LUT4
Input
Multi-purpose
M0
Multipurpose Input
Input
Multi-purpose
M1
Multipurpose Input
Input
Control signal
CE
Clock Enable
Input
Control signal
LSR
Local Set/Reset
Input
Control signal
CLK
System Clock
Input
Inter-PFU signal
FC
Fast Carry-in1
Input
Inter-slice signal
FXA
Intermediate signal to generate LUT6 and LUT7
Intermediate signal to generate LUT6 and LUT7
Input
Inter-slice signal
FXB
Output
Data signals
F0, F1
LUT4 output register bypass signals
Output
Data signals
Q0, Q1
Register outputs
Output
Data signals
OFX0
Output of a LUT5 MUX
Output
Data signals
OFX1
Output of a LUT6, LUT7, LUT82 MUX depending on the slice
Output
Inter-PFU signal
FCO
Slice 2 of each PFU is the fast carry chain output1
1. See Figure 2-3 for connection details.
2. Requires two PFUs.
2-4
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Modes of Operation
Each slice has up to four potential modes of operation: Logic, Ripple, RAM and ROM.
Logic Mode
In this mode, the LUTs in each slice are configured as 4-input combinatorial lookup tables. A LUT4 can have 16
possible input combinations. Any four input logic functions can be generated by programming this lookup table.
Since there are two LUT4s per slice, a LUT5 can be constructed within one slice. Larger look-up tables such as
LUT6, LUT7 and LUT8 can be constructed by concatenating other slices. Note LUT8 requires more than four
slices.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Ripple Mode
Ripple mode supports the efficient implementation of small arithmetic functions. In ripple mode, the following functions can be implemented by each slice:
• Addition 2-bit
• Subtraction 2-bit
• Add/Subtract 2-bit using dynamic control
• Up counter 2-bit
• Down counter 2-bit
• Up/Down counter with asynchronous clear
• Up/Down counter with preload (sync)
• Ripple mode multiplier building block
• Multiplier support
• Comparator functions of A and B inputs
– A greater-than-or-equal-to B
– A not-equal-to B
– A less-than-or-equal-to B
Ripple Mode includes an optional configuration that performs arithmetic using fast carry chain methods. In this configuration (also referred to as CCU2 mode) two additional signals, Carry Generate and Carry Propagate, are generated on a per slice basis to allow fast arithmetic functions to be constructed by concatenating Slices.
RAM Mode
In this mode, a 16x4-bit distributed single port RAM (SPR) can be constructed using each LUT block in Slice 0 and
Slice 1 as a 16x1-bit memory. Slice 2 is used to provide memory address and control signals. A 16x2-bit pseudo
dual port RAM (PDPR) memory is created by using one Slice as the read-write port and the other companion slice
as the read-only port.
LatticeECP3 devices support distributed memory initialization.
The Lattice design tools support the creation of a variety of different size memories. Where appropriate, the software will construct these using distributed memory primitives that represent the capabilities of the PFU. Table 2-3
shows the number of slices required to implement different distributed RAM primitives. For more information about
using RAM in LatticeECP3 devices, please see TN1179, LatticeECP3 Memory Usage Guide.
Table 2-3. Number of Slices Required to Implement Distributed RAM
Number of slices
SPR 16X4
PDPR 16X4
3
3
Note: SPR = Single Port RAM, PDPR = Pseudo Dual Port RAM
2-5
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
ROM Mode
ROM mode uses the LUT logic; hence, Slices 0 through 3 can be used in ROM mode. Preloading is accomplished
through the programming interface during PFU configuration.
For more information, please refer to TN1179, LatticeECP3 Memory Usage Guide.
Routing
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
There are many resources provided in the LatticeECP3 devices to route signals individually or as busses with
related control signals. The routing resources consist of switching circuitry, buffers and metal interconnect (routing)
segments.
The LatticeECP3 family has an enhanced routing architecture that produces a compact design. The ispLEVER
design tool suite takes the output of the synthesis tool and places and routes the design.
sysCLOCK PLLs and DLLs
The sysCLOCK PLLs provide the ability to synthesize clock frequencies. All the devices in the LatticeECP3 family
support four to ten full-featured General Purpose PLLs.
General Purpose PLL
The architecture of the PLL is shown in Figure 2-4. A description of the PLL functionality follows.
CLKI is the reference frequency (generated either from the pin or from routing) for the PLL. CLKI feeds into the
Input Clock Divider block. The CLKFB is the feedback signal (generated from CLKOP, CLKOS or from a user clock
pin/logic). This signal feeds into the Feedback Divider. The Feedback Divider is used to multiply the reference frequency.
Both the input path and feedback signals enter the Voltage Controlled Oscillator (VCO) block. In this block the difference between the input path and feedback signals is used to control the frequency and phase of the oscillator. A
LOCK signal is generated by the VCO to indicate that the VCO has locked onto the input clock signal. In dynamic
mode, the PLL may lose lock after a dynamic delay adjustment and not relock until the tLOCK parameter has been
satisfied.
The output of the VCO then enters the CLKOP divider. The CLKOP divider allows the VCO to operate at higher frequencies than the clock output (CLKOP), thereby increasing the frequency range. The Phase/Duty Select block
adjusts the phase and duty cycle of the CLKOS signal. The phase/duty cycle setting can be pre-programmed or
dynamically adjusted. A secondary divider takes the CLKOP or CLKOS signal and uses it to derive lower frequency
outputs (CLKOK).
The primary output from the CLKOP divider (CLKOP) along with the outputs from the secondary dividers (CLKOK
and CLKOK2) and Phase/Duty select (CLKOS) are fed to the clock distribution network.
The PLL allows two methods for adjusting the phase of signal. The first is referred to as Fine Delay Adjustment.
This inserts up to 16 nominal 125ps delays to be applied to the secondary PLL output. The number of steps may
be set statically or from the FPGA logic. The second method is referred to as Coarse Phase Adjustment. This
allows the phase of the rising and falling edge of the secondary PLL output to be adjusted in 22.5 degree steps.
The number of steps may be set statically or from the FPGA logic.
2-6
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-4. General Purpose PLL Diagram
FDA[3:0]
WRDEL
3
Phase/
Duty Cycle/
Duty Trim
CLKI
Divider
CLKI
VCO/
Loop Filter
CLKOP
Divider
CLKOS
CLKOP
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
PFD
CLKOK2
CLKFB
CLKFB
Divider
Duty Trim
CLKOK
CLKOK
Divider
Lock
Detect
RSTK
RST
DRPAI[3:0]
DFPAI[3:0]
LOCK
Table 2-4 provides a description of the signals in the PLL blocks.
Table 2-4. PLL Blocks Signal Descriptions
Signal
I/O
Description
CLKI
I
Clock input from external pin or routing
CLKFB
I
PLL feedback input from CLKOP, CLKOS, or from a user clock (pin or logic)
RST
I
“1” to reset PLL counters, VCO, charge pumps and M-dividers
RSTK
I
“1” to reset K-divider
WRDEL
I
DPA Fine Delay Adjust input
CLKOS
O
PLL output to clock tree (phase shifted/duty cycle changed)
CLKOP
O
PLL output to clock tree (no phase shift)
CLKOK
O
PLL output to clock tree through secondary clock divider
CLKOK2
O
PLL output to clock tree (CLKOP divided by 3)
LOCK
O
“1” indicates PLL LOCK to CLKI
FDA [3:0]
I
Dynamic fine delay adjustment on CLKOS output
DRPAI[3:0]
I
Dynamic coarse phase shift, rising edge setting
DFPAI[3:0]
I
Dynamic coarse phase shift, falling edge setting
Delay Locked Loops (DLL)
In addition to PLLs, the LatticeECP3 family of devices has two DLLs per device.
CLKI is the input frequency (generated either from the pin or routing) for the DLL. CLKI feeds into the output muxes
block to bypass the DLL, directly to the DELAY CHAIN block and (directly or through divider circuit) to the reference
input of the Phase Detector (PD) input mux. The reference signal for the PD can also be generated from the Delay
Chain signals. The feedback input to the PD is generated from the CLKFB pin or from a tapped signal from the
Delay chain.
The PD produces a binary number proportional to the phase and frequency difference between the reference and
feedback signals. Based on these inputs, the ALU determines the correct digital control codes to send to the delay
2-7
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
chain in order to better match the reference and feedback signals. This digital code from the ALU is also transmitted via the Digital Control bus (DCNTL) bus to its associated Slave Delay lines (two per DLL). The ALUHOLD input
allows the user to suspend the ALU output at its current value. The UDDCNTL signal allows the user to latch the
current value on the DCNTL bus.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
The DLL has two clock outputs, CLKOP and CLKOS. These outputs can individually select one of the outputs from
the tapped delay line. The CLKOS has optional fine delay shift and divider blocks to allow this output to be further
modified, if required. The fine delay shift block allows the CLKOS output to phase shifted a further 45, 22.5 or 11.25
degrees relative to its normal position. Both the CLKOS and CLKOP outputs are available with optional duty cycle
correction. Divide by two and divide by four frequencies are available at CLKOS. The LOCK output signal is
asserted when the DLL is locked. Figure 2-5 shows the DLL block diagram and Table 2-5 provides a description of
the DLL inputs and outputs.
The user can configure the DLL for many common functions such as time reference delay mode and clock injection
removal mode. Lattice provides primitives in its design tools for these functions.
Figure 2-5. Delay Locked Loop Diagram (DLL)
Delay Chain
ALUHOLD
Delay0
Duty
Cycle
50%
CLKOP
Delay1
÷4
÷2
(from routing
or external pin)
CLKI
from CLKOP (DLL
internal), from clock net
(CLKOP) or from a user
clock (pin or logic)
Output
Muxes
Delay2
Reference
Phase
Detector
Duty
Cycle
50%
Delay3
Arithmetic
Logic Unit
CLKOS
÷4
÷2
Delay4
Feedback
CLKFB
LOCK
Lock
Detect
6
Digital
Control
Output
UDDCNTL
RSTN
INCI
GRAYI[5:0]
DCNTL[5:0]*
DIFF
INCO
GRAYO[5:0]
* This signal is not user accessible. This can only be used to feed the slave delay line.
2-8
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Table 2-5. DLL Signals
Signal
I/O
Description
CLKI
I
Clock input from external pin or routing
CLKFB
I
DLL feed input from DLL output, clock net, routing or external pin
RSTN
I
Active low synchronous reset
ALUHOLD
I
Active high freezes the ALU
I
Synchronous enable signal (hold high for two cycles) from routing
CLKOP
O
The primary clock output
CLKOS
O
The secondary clock output with fine delay shift and/or division by 2 or by 4
LOCK
O
Active high phase lock indicator
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
UDDCNTL
INCI
I
Incremental indicator from another DLL via CIB.
GRAYI[5:0]
I
Gray-coded digital control bus from another DLL in time reference mode.
DIFF
O
Difference indicator when DCNTL is difference than the internal setting and update is needed.
INCO
O
Incremental indicator to other DLLs via CIB.
GRAYO[5:0]
O
Gray-coded digital control bus to other DLLs via CIB
LatticeECP3 devices have two general DLLs and four Slave Delay lines, two per DLL. The DLLs are in the lowest
EBR row and located adjacent to the EBR. Each DLL replaces one EBR block. One Slave Delay line is placed adjacent to the DLL and the duplicate Slave Delay line (in Figure 2-6) for the DLL is placed in the I/O ring between
Banks 6 and 7 and Banks 2 and 3.
The outputs from the DLL and Slave Delay lines are fed to the clock distribution network.
For more information, please see TN1178, LatticeECP3 sysCLOCK PLL/DLL Design and Usage Guide.
Figure 2-6. Top-Level Block Diagram, High-Speed DLL and Slave Delay Line
HOLD
GRAY_IN[5:0]
INC_IN
RSTN
GSRN
UDDCNTL
DCPS[5:0]
CLKOP
CLKOS
TPIO0 (L) OR TPIO1 (R)
GPLL_PIO
CIB (DATA)
CIB (CLK)
GDLL_PIO
4
Top ECLK1 (L) OR Top ECLK2 (R)
FB CIB (CLK)
Internal from CLKOP
GDLLFB_PIO
ECLK1
4
3
2
CLKI
LatticeECP3
High-Speed DLL
LOCK
1
0
3
2
GRAY_OUT[5:0]
INC_OUT
CLKFB
DIFF
1
0
DCNTL[5:0]*
DCNTL[5:0]
4
3
2
1
CLKI
Slave Delay Line
CLKO (to edge clock
muxes as CLKINDEL)
0
* This signal is not user accessible. It can only be used to feed the slave delay line.
2-9
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
PLL/DLL Cascading
LatticeECP3 devices have been designed to allow certain combinations of PLL and DLL cascading. The allowable
combinations are:
• PLL to PLL supported
• PLL to DLL supported
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
The DLLs in the LatticeECP3 are used to shift the clock in relation to the data for source synchronous inputs. PLLs
are used for frequency synthesis and clock generation for source synchronous interfaces. Cascading PLL and DLL
blocks allows applications to utilize the unique benefits of both DLLs and PLLs.
For further information about the DLL, please see the list of technical documentation at the end of this data sheet.
PLL/DLL PIO Input Pin Connections
All LatticeECP3 devices contains two DLLs and up to ten PLLs, arranged in quadrants. If a PLL and a DLL are next
to each other, they share input pins as shown in the Figure 2-7.
Figure 2-7. Sharing of PIO Pins by PLLs and DLLs in LatticeECP3 Devices
PLL_PIO
PLL
DLL
DLL_PIO
Note: Not every PLL has an associated DLL.
Clock Dividers
LatticeECP3 devices have two clock dividers, one on the left side and one on the right side of the device. These are
intended to generate a slower-speed system clock from a high-speed edge clock. The block operates in a ÷2, ÷4 or
÷8 mode and maintains a known phase relationship between the divided down clock and the high-speed clock
based on the release of its reset signal. The clock dividers can be fed from selected PLL/DLL outputs, the Slave
Delay lines, routing or from an external clock input. The clock divider outputs serve as primary clock sources and
feed into the clock distribution network. The Reset (RST) control signal resets input and asynchronously forces all
outputs to low. The RELEASE signal releases outputs synchronously to the input clock. For further information on
clock dividers, please see TN1178, LatticeECP3 sysCLOCK PLL/DLL Design and Usage Guide. Figure 2-8 shows
the clock divider connections.
2-10
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-8. Clock Divider Connections
ECLK1
ECLK2
÷1
CLKOP (PLL)
CLKOP (DLL)
÷2
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
CLKDIV
÷4
RST
÷8
RELEASE
Clock Distribution Network
LatticeECP3 devices have eight quadrant-based primary clocks and eight secondary clock/control sources. Two
high performance edge clocks are available on the top, left, and right edges of the device to support high speed
interfaces. These clock sources are selected from external I/Os, the sysCLOCK PLLs, DLLs or routing. These clock
sources are fed throughout the chip via a clock distribution system.
Primary Clock Sources
LatticeECP3 devices derive clocks from six primary source types: PLL outputs, DLL outputs, CLKDIV outputs, dedicated clock inputs, routing and SERDES Quads. LatticeECP3 devices have two to ten sysCLOCK PLLs and two
DLLs, located on the left and right sides of the device. There are six dedicated clock inputs: two on the top side, two
on the left side and two on the right side of the device. Figures 2-9, 2-10 and 2-11 show the primary clock sources
for LatticeECP3 devices.
Figure 2-9. Primary Clock Sources for LatticeECP3-17
Clock Input
Clock Input
From Routing
Clock
Input
Clock
Input
Clock
Input
Clock
Input
CLK
DIV
Primary Clock Sources
to Eight Quadrant Clock Selection
CLK
DIV
DLL Input
DLL
DLL
DLL Input
PLL Input
PLL
PLL
PLL Input
SERDES
Quad
From Routing
Note: Clock inputs can be configured in differential or single-ended mode.
2-11
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-10. Primary Clock Sources for LatticeECP3-35
Clock Input
Clock Input
From Routing
PLL Input
PLL
PLL
PLL Input
Clock
Input
Clock
Input
Clock
Input
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Clock
Input
CLK
DIV
DLL Input
DLL
PLL Input
PLL
Primary Clock Sources
to Eight Quadrant Clock Selection
CLK
DIV
DLL
DLL Input
PLL
PLL Input
SERDES
Quad
From Routing
Note: Clock inputs can be configured in differential or single-ended mode.
Figure 2-11. Primary Clock Sources for LatticeECP3-70, -95, -150
Clock Input
Clock Input
From Routing
PLL Input
PLL
PLL
PLL Input
PLL Input
PLL
PLL
PLL Input
Clock
Input
Clock
Input
Clock
Input
Clock
Input
CLK
DIV
Primary Clock Sources
to Eight Quadrant Clock Selection
CLK
DIV
DLL Input
DLL
DLL
DLL Input
PLL Input
PLL
PLL
PLL Input
PLL Input
PLL
PLL
PLL Input
PLL Input
PLL
PLL
PLL Input
SERDES
Quad
SERDES
Quad
SERDES
Quad
SERDES
Quad
(ECP3-150 only)
From Routing
Note: Clock inputs can be configured in differential or single-ended mode.
2-12
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Primary Clock Routing
The purpose of the primary clock routing is to distribute primary clock sources to the destination quadrants of the
device. A global primary clock is a primary clock that is distributed to all quadrants. The clock routing structure in
LatticeECP3 devices consists of a network of eight primary clock lines (CLK0 through CLK7) per quadrant. The primary clocks of each quadrant are generated from muxes located in the center of the device. All the clock sources
are connected to these muxes. Figure 2-12 shows the clock routing for one quadrant. Each quadrant mux is identical. If desired, any clock can be routed globally.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Figure 2-12. Per Quadrant Primary Clock Selection
PLLs + DLLs + CLKDIVs + PCLK PIOs + SERDES Quads
63:1
63:1
63:1
63:1
63:1
63:1
DCC
DCC
DCC
DCC
DCC
DCC
CLK0
CLK1
CLK2
CLK3
CLK4
CLK5
58:1
58:1
DCS
CLK6
58:1
58:1
DCS
CLK7
8 Primary Clocks (CLK0 to CLK7) per Quadrant
Dynamic Clock Control (DCC)
The DCC (Quadrant Clock Enable/Disable) feature allows internal logic control of the quadrant primary clock network. When a clock network is disabled, all the logic fed by that clock does not toggle, reducing the overall power
consumption of the device.
Dynamic Clock Select (DCS)
The DCS is a smart multiplexer function available in the primary clock routing. It switches between two independent
input clock sources without any glitches or runt pulses. This is achieved regardless of when the select signal is toggled. There are two DCS blocks per quadrant; in total, there are eight DCS blocks per device. The inputs to the
DCS block come from the center muxes. The output of the DCS is connected to primary clocks CLK6 and CLK7
(see Figure 2-12).
Figure 2-13 shows the timing waveforms of the default DCS operating mode. The DCS block can be programmed
to other modes. For more information about the DCS, please see the list of technical documentation at the end of
this data sheet.
Figure 2-13. DCS Waveforms
CLK0
CLK1
SEL
DCSOUT
2-13
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Secondary Clock/Control Sources
LatticeECP3 devices derive eight secondary clock sources (SC0 through SC7) from six dedicated clock input pads
and the rest from routing. Figure 2-14 shows the secondary clock sources. All eight secondary clock sources are
defined as inputs to a per-region mux SC0-SC7. SC0-SC3 are primary for control signals (CE and/or LSR), and
SC4-SC7 are for clock and high fanout data.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
In an actual implementation, there is some overlap to maximize routability. In addition to SC0-SC3, SC7 is also an
input to the control signals (LSR or CE). SC0-SC2 are also inputs to clocks along with SC4-SC7. High fanout logic
signals (LUT inputs) will utilize the X2 and X0 switches where SC0-SC7 are inputs to X2 switches, and SC4-SC7
are inputs to X0 switches. Note that through X0 switches, SC4-SC7 can also access control signals CE/LSR.
Figure 2-14. Secondary Clock Sources
Clock
Input
From
Routing
Clock
Input
From
Routing
From
Routing
From
Routing
From Routing
From Routing
From Routing
From Routing
Clock Input
Clock Input
Secondary Clock Sources
Clock Input
Clock Input
From Routing
From Routing
From Routing
From Routing
From
Routing
From
Routing
From
Routing
From
Routing
Note: Clock inputs can be configured in differential or single-ended mode.
Secondary Clock/Control Routing
Global secondary clock is a secondary clock that is distributed to all regions. The purpose of the secondary clock
routing is to distribute the secondary clock sources to the secondary clock regions. Secondary clocks in the
LatticeECP3 devices are region-based resources. Certain EBR rows and special vertical routing channels bind the
secondary clock regions. This special vertical routing channel aligns with either the left edge of the center DSP
slice in the DSP row or the center of the DSP row. Figure 2-15 shows this special vertical routing channel and the
20 secondary clock regions for the LatticeECP3 family of devices. All devices in the LatticeECP3 family have eight
2-14
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
secondary clock resources per region (SC0 to SC7). The same secondary clock routing can be used for control
signals.
Table 2-6. Secondary Clock Regions
Number of Secondary Clock
Regions
ECP3-17
16
ECP3-35
16
ECP3-70
20
ECP3-95
20
ECP3-150
36
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Device
Figure 2-15. LatticeECP3-70 and LatticeECP3-95 Secondary Clock Regions
Vertical Routing Channel
Regional Boundary
Secondary Clock
Region R1C2
Secondary Clock
Region R1C3
Secondary Clock
Region R1C4
Secondary Clock
Region R2C1
Secondary Clock
Region R2C2
Secondary Clock
Region R2C3
Secondary Clock
Region R2C4
Secondary Clock
Region R3C1
Secondary Clock
Region R3C2
Secondary Clock
Region R3C3
Secondary Clock
Region R3C4
Secondary Clock
Region R4C1
Secondary Clock
Region R4C2
Secondary Clock
Region R4C3
Secondary Clock
Region R4C4
Secondary Clock
Region R5C1
Secondary Clock
Region R5C2
Secondary Clock
Region R5C3
Secondary Clock
Region R5C4
SERDES
Spine Repeaters
2-15
EBR Row
Regional Boundary
sysIO Bank 2
Secondary Clock
Region R1C1
Configuration Bank
sysIO Bank 1
sysIO Bank 3
sysIO Bank 6
sysIO Bank 7
sysIO Bank 0
EBR Row
Regional Boundary
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-16. Per Region Secondary Clock Selection
Secondary Clock Feedlines: 8 PIOs + 16 Routing
8:1
8:1
8:1
8:1
8:1
8:1
8:1
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
8:1
SC0
SC1
SC2
SC3
SC4
SC5
SC6
SC7
8 Secondary Clocks (SC0 to SC7) per Region
Clock/Control
Slice Clock Selection
Figure 2-17 shows the clock selections and Figure 2-18 shows the control selections for Slice0 through Slice2. All
the primary clocks and seven secondary clocks are routed to this clock selection mux. Other signals can be used
as a clock input to the slices via routing. Slice controls are generated from the secondary clocks/controls or other
signals connected via routing.
If none of the signals are selected for both clock and control then the default value of the mux output is 1. Slice 3
does not have any registers; therefore it does not have the clock or control muxes.
Figure 2-17. Slice0 through Slice2 Clock Selection
Primary Clock
Secondary Clock
8
Clock to Slice
7
28:1
Routing
12
Vcc
1
Figure 2-18. Slice0 through Slice2 Control Selection
Secondary Control
5
Slice Control
Routing
20:1
14
Vcc
1
2-16
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Edge Clock Sources
Edge clock resources can be driven from a variety of sources at the same edge. Edge clock resources can be
driven from adjacent edge clock PIOs, primary clock PIOs, PLLs, DLLs, Slave Delay and clock dividers as shown in
Figure 2-19.
Figure 2-19. Edge Clock Sources
Clock Input
Clock Input
From
Routing
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
From
Routing
Sources for top
edge clocks
From Routing
From Routing
Clock
Input
Clock
Input
Clock
Input
Clock
Input
From Routing
Slave Delay
Six Edge Clocks (ECLK)
Two Clocks per Edge
From Routing
Slave Delay
DLL
Input
DLL
DLL
DLL
Input
PLL
Input
PLL
PLL
PLL
Input
Sources for right edge clocks
Sources for left edge clocks
Notes:
1. Clock inputs can be configured in differential or single ended mode.
2. The two DLLs can also drive the two top edge clocks.
3. The top left and top right PLL can also drive the two top edge clocks.
Edge Clock Routing
LatticeECP3 devices have a number of high-speed edge clocks that are intended for use with the PIOs in the
implementation of high-speed interfaces. There are six edge clocks per device: two edge clocks on each of the top,
left, and right edges. Different PLL and DLL outputs are routed to the two muxes on the left and right sides of the
device. In addition, the CLKINDEL signal (generated from the DLL Slave Delay Line block) is routed to all the edge
clock muxes on the left and right sides of the device. Figure 2-20 shows the selection muxes for these clocks.
2-17
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-20. Sources of Edge Clock (Left and Right Edges)
Input Pad
Left and Right
PLL Input Pad
Edge Clocks
DLL Output CLKOP
ECLK1
6:1
PLL Output CLKOP
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Routing
CLKINDEL
from DLL Slave Delay
Input Pad
Left and Right
PLL Input Pad
Edge Clocks
DLL Output CLKOS
ECLK2
6:1
PLL Output CLKOS
Routing
CLKINDEL
from DLL Slave Delay
Figure 2-21. Sources of Edge Clock (Top Edge)
Input Pad
Top left PLL_CLKOP
Top Right PLL_CLKOS
ECLK1
7:1
Left DLL_CLKOP
Right DLL_CLKOS
Routing
CLKINDEL
(Left DLL_DEL)
Input Pad
Top Right PLL_CLKOP
ECLK2
Top Left PLL_CLKOS
7:1
Right DLL_CLKOP
Left DLL_CLKOS
Routing
CLKINDEL
(Right DLL_DEL)
The edge clocks have low injection delay and low skew. They are used to clock the I/O registers and thus are ideal
for creating I/O interfaces with a single clock signal and a wide data bus. They are also used for DDR Memory or
Generic DDR interfaces.
2-18
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
The edge clocks on the top, left, and right sides of the device can drive the secondary clocks or general routing
resources of the device. The left and right side edge clocks also can drive the primary clock network through the
clock dividers (CLKDIV).
sysMEM Memory
LatticeECP3 devices contain a number of sysMEM Embedded Block RAM (EBR). The EBR consists of an 18-Kbit
RAM with memory core, dedicated input registers and output registers with separate clock and clock enable. Each
EBR includes functionality to support true dual-port, pseudo dual-port, single-port RAM, ROM and FIFO buffers
(via external PFUs).
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
sysMEM Memory Block
The sysMEM block can implement single port, dual port or pseudo dual port memories. Each block can be used in
a variety of depths and widths as shown in Table 2-7. FIFOs can be implemented in sysMEM EBR blocks by implementing support logic with PFUs. The EBR block facilitates parity checking by supporting an optional parity bit for
each data byte. EBR blocks provide byte-enable support for configurations with18-bit and 36-bit data widths. For
more information, please see TN1179, LatticeECP3 Memory Usage Guide.
Table 2-7. sysMEM Block Configurations
Memory Mode
Configurations
Single Port
16,384 x 1
8,192 x 2
4,096 x 4
2,048 x 9
1,024 x 18
512 x 36
True Dual Port
16,384 x 1
8,192 x 2
4,096 x 4
2,048 x 9
1,024 x 18
Pseudo Dual Port
16,384 x 1
8,192 x 2
4,096 x 4
2,048 x 9
1,024 x 18
512 x 36
Bus Size Matching
All of the multi-port memory modes support different widths on each of the ports. The RAM bits are mapped LSB
word 0 to MSB word 0, LSB word 1 to MSB word 1, and so on. Although the word size and number of words for
each port varies, this mapping scheme applies to each port.
RAM Initialization and ROM Operation
If desired, the contents of the RAM can be pre-loaded during device configuration. By preloading the RAM block
during the chip configuration cycle and disabling the write controls, the sysMEM block can also be utilized as a
ROM.
Memory Cascading
Larger and deeper blocks of RAM can be created using EBR sysMEM Blocks. Typically, the Lattice design tools
cascade memory transparently, based on specific design inputs.
2-19
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Single, Dual and Pseudo-Dual Port Modes
In all the sysMEM RAM modes the input data and address for the ports are registered at the input of the memory
array. The output data of the memory is optionally registered at the output.
EBR memory supports the following forms of write behavior for single port or dual port operation:
1. Normal – Data on the output appears only during a read cycle. During a write cycle, the data (at the current
address) does not appear on the output. This mode is supported for all data widths.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
2. Write Through – A copy of the input data appears at the output of the same port during a write cycle. This
mode is supported for all data widths.
3. Read-Before-Write (EA devices only) – When new data is written, the old content of the address appears at
the output. This mode is supported for x9, x18, and x36 data widths.
Memory Core Reset
The memory array in the EBR utilizes latches at the A and B output ports. These latches can be reset asynchronously or synchronously. RSTA and RSTB are local signals, which reset the output latches associated with Port A
and Port B, respectively. The Global Reset (GSRN) signal can reset both ports. The output data latches and associated resets for both ports are as shown in Figure 2-22.
Figure 2-22. Memory Core Reset
Memory Core
D
SET
Q
Port A[17:0]
LCLR
Output Data
Latches
D
SET
Q
Port B[17:0]
LCLR
RSTA
RSTB
GSRN
Programmable Disable
For further information on the sysMEM EBR block, please see the list of technical documentation at the end of this
data sheet.
sysDSP™ Slice
The LatticeECP3 family provides an enhanced sysDSP architecture, making it ideally suited for low-cost, high-performance Digital Signal Processing (DSP) applications. Typical functions used in these applications are Finite
Impulse Response (FIR) filters, Fast Fourier Transforms (FFT) functions, Correlators, Reed-Solomon/Turbo/Convolution encoders and decoders. These complex signal processing functions use similar building blocks such as multiply-adders and multiply-accumulators.
sysDSP Slice Approach Compared to General DSP
Conventional general-purpose DSP chips typically contain one to four (Multiply and Accumulate) MAC units with
fixed data-width multipliers; this leads to limited parallelism and limited throughput. Their throughput is increased by
higher clock speeds. The LatticeECP3, on the other hand, has many DSP slices that support different data widths.
2-20
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
This allows designers to use highly parallel implementations of DSP functions. Designers can optimize DSP performance vs. area by choosing appropriate levels of parallelism. Figure 2-23 compares the fully serial implementation
to the mixed parallel and serial implementation.
Figure 2-23. Comparison of General DSP and LatticeECP3 Approaches
Operand
A
Operand
A
Operand
B
Operand
B
Operand
B
Operand
B
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Operand
A
Operand
A
Single
Multiplier
x
M loops
x
Multiplier
0
x
x
Multiplier
1
m/k
loops
Multiplier
k
Accumulator
(k adds)
Function Implemented in
General Purpose DSP
+
m/k
accumulate
Output
Function Implemented in
LatticeECP3
LatticeECP3 sysDSP Slice Architecture Features
The LatticeECP3 sysDSP Slice has been significantly enhanced to provide functions needed for advanced processing applications. These enhancements provide improved flexibility and resource utilization.
The LatticeECP3 sysDSP Slice supports many functions that include the following:
• Multiply (one 18x36, two 18x18 or four 9x9 Multiplies per Slice)
• Multiply (36x36 by cascading across two sysDSP slices)
• Multiply Accumulate (up to 18x36 Multipliers feeding an Accumulator that can have up to 54-bit resolution)
• Two Multiplies feeding one Accumulate per cycle for increased processing with lower latency (two 18x18 Multiplies feed into an accumulator that can accumulate up to 52 bits)
• Flexible saturation and rounding options to satisfy a diverse set of applications situations
• Flexible cascading across DSP slices
– Minimizes fabric use for common DSP and ALU functions
– Enables implementation of FIR Filter or similar structures using dedicated sysDSP slice resources only
– Provides matching pipeline registers
– Can be configured to continue cascading from one row of sysDSP slices to another for longer cascade
chains
• Flexible and Powerful Arithmetic Logic Unit (ALU) Supports:
– Dynamically selectable ALU OPCODE
– Ternary arithmetic (addition/subtraction of three inputs)
– Bit-wise two-input logic operations (AND, OR, NAND, NOR, XOR and XNOR)
– Eight flexible and programmable ALU flags that can be used for multiple pattern detection scenarios, such
2-21
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
as, overflow, underflow and convergent rounding, etc.
– Flexible cascading across slices to get larger functions
• RTL Synthesis friendly synchronous reset on all registers, while still supporting asynchronous reset for legacy
users
• Dynamic MUX selection to allow Time Division Multiplexing (TDM) of resources for applications that require
processor-like flexibility that enables different functions for each clock cycle
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
For most cases, as shown in Figure 2-24, the LatticeECP3 DSP slice is backwards-compatible with the
LatticeECP2™ sysDSP block, such that, legacy applications can be targeted to the LatticeECP3 sysDSP slice. The
functionality of one LatticeECP2 sysDSP Block can be mapped into two adjacent LatticeECP3 sysDSP slices, as
shown in Figure 2-25.
Figure 2-24. Simplified sysDSP Slice Block Diagram
From FPGA Core
Slice 0
Input Registers
from SRO of
Left-side DSP
Slice 1
Casc
A0
IR
IR
IR
IR
MULTA
MULTB
9x9
9x9
9x9
IR
IR
IR
IR
MULTA
9x9
9x9
9x9
Casc
A1
MULTB
9x9
9x9
Mult18-0
Mult18-1
Mult18-0
Mult18-1
PR
PR
PR
PR
Intermediate Pipeline
Registers
Cascade from
Left DSP
Carry
Out
Reg.
Accumulator/ALU (54)
Carry
Out
Reg.
Accumulator/ALU (54)
ALU Op-Codes
Output Registers
OR
OR
OR
OR
OR
To FPGA Core
2-22
OR
OR
One of
these
OR
Cascade to
Right DSP
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-25. Detailed sysDSP Slice Diagram
From FPGA Core
C
AA
AB
OPCODE
BA
BB
SRIB
SROB
IR
IR
IR
IR
IR SROA
SRIA
MULTA
MULTB
IR
IR
PR
PR
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
PR
A_ALU
0
Previous
DSP Slice
B_ALU
0
Next
DSP Slice
BMUX
AMUX
C_ALU
CMUX
CIN
A_ALU
Rounding
COUT
R= A ± B ± C
R = Logic (B, C)
ALU
0
==
IR = Input Register
PR = Pipeline Register
OR = Output Register
FR = Flag Register
OR
OR
FR
OR
To FPGA Core
Note: A_ALU, B_ALU and C_ALU are internal signals generated by combining bits from AA, AB, BA BB and C
inputs. See TN1182, LatticeECP3 sysDSP Usage Guide, for further information.
The LatticeECP2 sysDSP block supports the following basic elements.
• MULT (Multiply)
• MAC (Multiply, Accumulate)
• MULTADDSUB (Multiply, Addition/Subtraction)
• MULTADDSUBSUM (Multiply, Addition/Subtraction, Summation)
Table 2-8 shows the capabilities of each of the LatticeECP3 slices versus the above functions.
Table 2-8. Maximum Number of Elements in a Slice
Width of Multiply
x9
x18
x36
MULT
4
2
1/2
MAC
1
1
—
MULTADDSUB
2
1
—
MULTADDSUBSUM
11
1/2
—
1. One slice can implement 1/2 9x9 m9x9addsubsum and two m9x9addsubsum with two slices.
Some options are available in the four elements. The input register in all the elements can be directly loaded or can
be loaded as a shift register from previous operand registers. By selecting “dynamic operation” the following operations are possible:
• In the Add/Sub option the Accumulator can be switched between addition and subtraction on every cycle.
• The loading of operands can switch between parallel and serial operations.
2-23
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
For further information, please refer to TN1182, LatticeECP3 sysDSP Usage Guide.
MULT DSP Element
This multiplier element implements a multiply with no addition or accumulator nodes. The two operands, AA and
AB, are multiplied and the result is available at the output. The user can enable the input/output and pipeline registers. Figure 2-26 shows the MULT sysDSP element.
Figure 2-26. MULT sysDSP Element
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
From FPGA Core
C
AA
AB
OPCODE
BA
BB
SRIB
SROB
IR
IR
IR
IR
IR SROA
SRIA
MULTA
MULTB
IR
IR
PR
PR
A_ALU
0
Previous
DSP Slice
CMUX
CIN
PR
0
Next
DSP Slice
BMUX
AMUX
C_ALU
A_ALU
Rounding
B_ALU
COUT
R= A ± B ± C
R = Logic (B, C)
ALU
0
==
IR = Input Register
PR = Pipeline Register
OR = Output Register
FR = Flag Register
OR
OR
To FPGA Core
2-24
FR
OR
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
MAC DSP Element
In this case, the two operands, AA and AB, are multiplied and the result is added with the previous accumulated
value. This accumulated value is available at the output. The user can enable the input and pipeline registers, but
the output register is always enabled. The output register is used to store the accumulated value. The ALU is configured as the accumulator in the sysDSP slice in the LatticeECP3 family can be initialized dynamically. A registered overflow signal is also available. The overflow conditions are provided later in this document. Figure 2-27
shows the MAC sysDSP element.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Figure 2-27. MAC DSP Element
From FPGA Core
C
AA
AB
OPCODE
BA
BB
SRIB
SROB
IR
IR
IR
IR
IR SROA
SRIA
MULTA
MULTB
IR
I
PR
PR
A_ALU
0
Previous
DSP Slice
CMUX
CIN
PR
0
Next
DSP Slice
BMUX
AMUX
C_ALU
A_ALU
Rounding
B_ALU
COUT
R= A ± B ± C
R = Logic (B, C)
ALU
0
==
IR = Input Register
PR = Pipeline Register
OR = Output Register
FR = Flag Register
OR
OR
To FPGA Core
2-25
FR
OR
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
MMAC DSP Element
The LatticeECP3 supports a MAC with two multipliers. This is called Multiply Multiply Accumulate or MMAC. In this
case, the two operands, AA and AB, are multiplied and the result is added with the previous accumulated value and
with the result of the multiplier operation of operands BA and BB. This accumulated value is available at the output.
The user can enable the input and pipeline registers, but the output register is always enabled. The output register
is used to store the accumulated value. The ALU is configured as the accumulator in the sysDSP slice. A registered
overflow signal is also available. The overflow conditions are provided later in this document. Figure 2-28 shows the
MMAC sysDSP element.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Figure 2-28. MMAC sysDSP Element
From FPGA Core
C
AA
AB
OPCODE
BA
BB
SRIB
SROB
IR
IR
IR
IR
IR SROA
SRIA
MULTA
MULTB
IR
IR
PR
PR
A_ALU
0
Previous
DSP Slice
B_ALU
0
Next
DSP Slice
BMUX
AMUX
C_ALU
PR
CMUX
CIN
A_ALU
Rounding
COUT
R= A ± B ± C
R = Logic (B, C)
ALU
0
==
IR = Input Register
PR = Pipeline Register
OR = Output Register
FR = Flag Register
OR
OR
To FPGA Core
2-26
FR
OR
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
MULTADDSUB DSP Element
In this case, the operands AA and AB are multiplied and the result is added/subtracted with the result of the multiplier operation of operands BA and BB. The user can enable the input, output and pipeline registers. Figure 2-29
shows the MULTADDSUB sysDSP element.
Figure 2-29. MULTADDSUB
From FPGA Core
C
AA
AB
OPCODE
BA
BB
SROB
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
SRIB
IR
IR
IR
IR
IR SROA
SRIA
MULTA
MULTB
IR
IR
PR
PR
A_ALU
0
Previous
DSP Slice
CMUX
PR
0
Next
DSP Slice
BMUX
AMUX
C_ALU
CIN
A_ALU
Rounding
B_ALU
COUT
R= A ± B ± C
R = Logic (B, C)
ALU
0
==
IR = Input Register
PR = Pipeline Register
OR = Output Register
FR = Flag Register
OR
OR
To FPGA Core
2-27
FR
OR
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
MULTADDSUBSUM DSP Element
In this case, the operands AA and AB are multiplied and the result is added/subtracted with the result of the multiplier operation of operands BA and BB of Slice 0. Additionally, the operands AA and AB are multiplied and the
result is added/subtracted with the result of the multiplier operation of operands BA and BB of Slice 1. The results
of both addition/subtractions are added by the second ALU following the slice cascade path. The user can enable
the input, output and pipeline registers. Figure 2-30 and Figure 2-31 show the MULTADDSUBSUM sysDSP element.
Figure 2-30. MULTADDSUBSUM Slice 0
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
From FPGA Core
C
AA
AB
OPCODE
BA
BB
SRIB
SROB
IR
IR
IR
IR
IR SROA
SRIA
MULTA
MULTB
I
IR
PR
PR
A_ALU
0
Previous
DSP Slice
CMUX
PR
Next
DSP Slice
BMUX
AMUX
C_ALU
CIN
A_ALU
Rounding
B_ALU
0
COUT
R= A ± B ± C
R = Logic (B, C)
ALU
0
==
IR = Input Register
PR = Pipeline Register
OR = Output Register
FR = Flag Register
OR
OR
To FPGA Core
2-28
FR
OR
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-31. MULTADDSUBSUM Slice 1
From FPGA Core
C
AA
AB
OPCODE
BA
BB
SRIB
SROB
IR
IR
IR
IR
IR SROA
SRIA
MULTA
MULTB
IR
IR
PR
PR
A_ALU
0
PR
0
Next
DSP Slice
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Previous
DSP Slice
B_ALU
CMUX
CIN
A_ALU
Rounding
BMUX
AMUX
C_ALU
COUT
R= A ± B ± C
R = Logic (B, C)
ALU
0
==
IR = Input Register
PR = Pipeline Register
OR = Output Register
FR = Flag Register
OR
OR
FR
OR
To FPGA Core
Advanced sysDSP Slice Features
Cascading
The LatticeECP3 sysDSP slice has been enhanced to allow cascading. Adder trees are implemented fully in sysDSP slices, improving the performance. Cascading of slices uses the signals CIN, COUT and C Mux of the slice.
Addition
The LatticeECP3 sysDSP slice allows for the bypassing of multipliers and cascading of adder logic. High performance adder functions are implemented without the use of LUTs. The maximum width adders that can be implemented are 54-bit.
Rounding
The rounding operation is implemented in the ALU and is done by adding a constant followed by a truncation operation. The rounding methods supported are:
• Rounding to zero (RTZ)
• Rounding to infinity (RTI)
• Dynamic rounding
• Random rounding
• Convergent rounding
2-29
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
ALU Flags
The sysDSP slice provides a number of flags from the ALU including:
• Equal to zero (EQZ)
• Equal to zero with mask (EQZM)
• Equal to one with mask (EQOM)
• Equal to pattern with mask (EQPAT)
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
• Equal to bit inverted pattern with mask (EQPATB)
• Accumulator Overflow (OVER)
• Accumulator Underflow (UNDER)
• Either over or under flow supporting LatticeECP2 legacy designs (OVERUNDER)
Clock, Clock Enable and Reset Resources
Global Clock, Clock Enable and Reset signals from routing are available to every sysDSP slice. From four clock
sources (CLK0, CLK1, CLK2, and CLK3) one clock is selected for each input register, pipeline register and output
register. Similarly Clock Enable (CE) and Reset (RST) are selected at each input register, pipeline register and output register.
Resources Available in the LatticeECP3 Family
Table 2-9 shows the maximum number of multipliers for each member of the LatticeECP3 family. Table 2-10 shows
the maximum available EBR RAM Blocks in each LatticeECP3 device. EBR blocks, together with Distributed RAM
can be used to store variables locally for fast DSP operations.
Table 2-9. Maximum Number of DSP Slices in the LatticeECP3 Family
Device
DSP Slices
9x9 Multiplier
18x18 Multiplier
36x36 Multiplier
ECP3-17
12
48
24
6
ECP3-35
32
128
64
16
ECP3-70
64
256
128
32
ECP3-95
64
256
128
32
ECP3-150
160
640
320
80
Table 2-10. Embedded SRAM in the LatticeECP3 Family
Device
EBR SRAM Block
Total EBR SRAM
(Kbits)
ECP3-17
38
700
ECP3-35
72
1327
ECP3-70
240
4420
ECP3-95
240
4420
ECP3-150
372
6850
2-30
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Programmable I/O Cells (PIC)
Each PIC contains two PIOs connected to their respective sysI/O buffers as shown in Figure 2-32. The PIO Block
supplies the output data (DO) and the tri-state control signal (TO) to the sysI/O buffer and receives input from the
buffer. Table 2-11 provides the PIO signal list.
Figure 2-32. PIC Diagram
I/Os in a DQS-12 Group, Except DQSN (Complement of DQS) I/Os
PIOA
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
TS
ONEGB
IOLT0
Tristate
Register
Block
OPOSA
OPOSB
ONEGA**
ONEGB**
PADA
“T”
IOLD0
Output
Register
Block
(ISI)
INDD
INCK
INB
IPB
INA
IPA
DEL[3:0]
ECLK1, ECLK2
SCLK
CE
LSR
GSRN
Input
Register
Block
Control
Muxes
CLK
CEOT
LSR
GSR
CEI
sysIO
Buffer
DI
PADB
“C”
PIOB
DQS Control Block
(One per DQS Group of 12 I/Os)***
Read Control
ECLK1
ECLK2
SCLK
READ
DCNTL[5:0]
DYNDEL[7:0]
DQSI
PRMDET
DDRLAT*
DDRCLKPOL*
ECLKDQSR*
Write Control
DQCLK0*
DQCLK1*
DQSW*
* Signals are available on left/right/top edges only.
** Signals are available on the left and right sides only
*** Selected PIO.
2-31
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Two adjacent PIOs can be joined to provide a differential I/O pair (labeled as “T” and “C”) as shown in Figure 2-32.
The PAD Labels “T” and “C” distinguish the two PIOs. Approximately 50% of the PIO pairs on the left and right
edges of the device can be configured as true LVDS outputs. All I/O pairs can operate as LVDS inputs.
Table 2-11. PIO Signal List
Name
Type
Description
INDD
Input Data
Register bypassed input. This is not the same port as INCK.
IPA, INA, IPB, INB
Input Data
Ports to core for input data
1
OPOSA, ONEGA ,
OPOSB, ONEGB1
Output Data
CE
PIO Control
Clock enables for input and output block flip-flops.
SCLK
PIO Control
System Clock (PCLK) for input and output/TS blocks. Connected from clock ISB.
LSR
PIO Control
Local Set/Reset
ECLK1, ECLK2
PIO Control
Edge clock sources. Entire PIO selects one of two sources using mux.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Output signals from core. An exception is the ONEGB port, used for tristate logic
at the DQS pad.
ECLKDQSR1
Read Control
DDRCLKPOL1
Read Control
Ensures transfer from DQS domain to SCLK domain.
DDRLAT1
Read Control
Used to guarantee INDDRX2 gearing by selectively enabling a D-Flip-Flop in datapath.
DEL[3:0]
Read Control
Dynamic input delay control bits.
INCK
From DQS_STROBE, shifted strobe for memory interfaces only.
To Clock Distribution PIO treated as clock PIO, path to distribute to primary clocks and PLL.
and PLL
TS
Tristate Data
Tristate signal from core (SDR)
DQCLK01, DQCLK11
Write Control
Two clocks edges, 90 degrees out of phase, used in output gearing.
DQSW2
Write Control
Used for output and tristate logic at DQS only.
DYNDEL[7:0]
Write Control
Shifting of write clocks for specific DQS group, using 6:0 each step is approximately 25ps, 128 steps. Bit 7 is an invert (timing depends on input frequency).
There is also a static control for this 8-bit setting, enabled with a memory cell.
DCNTL[6:0]
1
DATAVALID
READ
PIO Control
Original delay code from DDR DLL
Output Data
Status flag from DATAVALID logic, used to indicate when input data is captured in
IOLOGIC and valid to core.
For DQS_Strobe
Read signal for DDR memory interface
DQSI
For DQS_Strobe
Unshifted DQS strobe from input pad
PRMBDET
For DQS_Strobe
DQSI biased to go high when DQSI is tristate, goes to input logic block as well as
core logic.
GSRN
Control from routing Global Set/Reset
1. Signals available on left/right/top edges only.
2. Selected PIO.
PIO
The PIO contains four blocks: an input register block, output register block, tristate register block and a control logic
block. These blocks contain registers for operating in a variety of modes along with the necessary clock and selection logic.
Input Register Block
The input register blocks for the PIOs, in the left, right and top edges, contain delay elements and registers that can
be used to condition high-speed interface signals, such as DDR memory interfaces and source synchronous interfaces, before they are passed to the device core. Figure 2-33 shows the input register block for the left, right and
top edges. The input register block for the bottom edge contains one element to register the input signal and no
DDR registers. The following description applies to the input register block for PIOs in the left, right and top edges
only.
2-32
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Input signals are fed from the sysI/O buffer to the input register block (as signal DI). If desired, the input signal can
bypass the register and delay elements and be used directly as a combinatorial signal (INDD), a clock (INCK) and,
in selected blocks, the input to the DQS delay block. If an input delay is desired, designers can select either a fixed
delay or a dynamic delay DEL[3:0]. The delay, if selected, reduces input register hold time requirements when
using a global clock.
The input block allows three modes of operation. In single data rate (SDR) the data is registered with the system
clock by one of the registers in the single data rate sync register block.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
In DDR mode, two registers are used to sample the data on the positive and negative edges of the modified DQS
(ECLKDQSR) in the DDR Memory mode or ECLK signal when using DDR Generic mode, creating two data
streams. Before entering the core, these two data streams are synchronized to the system clock to generate two
data streams.
A gearbox function can be implemented in each of the input registers on the left and right sides. The gearbox function takes a double data rate signal applied to PIOA and converts it as four data streams, INA, IPA, INB and IPB.
The two data streams from the first set of DDR registers are synchronized to the edge clock and then to the system
clock before entering the core. Figure 2-30 provides further information on the use of the gearbox function.
The signal DDRCLKPOL controls the polarity of the clock used in the synchronization registers. It ensures adequate timing when data is transferred to the system clock domain from the ECLKDQSR (DDR Memory Interface
mode) or ECLK (DDR Generic mode). The DDRLAT signal is used to ensure the data transfer from the synchronization registers to the clock transfer and gearbox registers.
The ECLKDQSR, DDRCLKPOL and DDRLAT signals are generated in the DQS Read Control Logic Block. See
Figure 2-37 for an overview of the DQS read control logic.
Further discussion about using the DQS strobe in this module is discussed in the DDR Memory section of this data
sheet.
Please see TN1180, LatticeECP3 High-Speed I/O Interface for more information on this topic.
2-33
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-33. ECP3-70/95 (E or EA) Input Register Block for Left, Right and Top Edges
Clock Transfer &
Gearing Registers*
INCLK**
INDD
To DQSI**
Fixed Delay
DI
Dynamic Delay
(From sysIO
Buffer)
Synch Registers
DDR Registers
0
1
A
D D Q
D Q
F D Q
DDRLAT
H D Q
Config bit
X0
01
11
L
D Q
CE
INB
R
L
B
DEL[3:0]
D Q
C D Q
E D Q
G
D Q
X0
01
11
K
D Q
IPB
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
D Q
L
L
ECLKDQSR
ECLK2
D Q J
D Q
INA
D Q
IPA
CLKP
L
DDRCLKPOL
ECLK1
ECLK2
D Q
1
0
10
D Q I
L
L
SCLK
* Only on the left and right sides.
** Selected PIO.
Note: Simplified diagram does not show CE/SET/REST details.
Output Register Block
The output register block registers signals from the core of the device before they are passed to the sysI/O buffers.
The blocks on the left and right PIOs contain registers for SDR and full DDR operation. The topside PIO block is the
same as the left and right sides except it does not support ODDRX2 gearing of output logic. ODDRX2 gearing is
used in DDR3 memory interfaces.The PIO blocks on the bottom contain the SDR registers and generic DDR interface without gearing.
Figure 2-34 shows the Output Register Block for PIOs on the left and right edges.
In SDR mode, OPOSA feeds one of the flip-flops that then feeds the output. The flip-flop can be configured as a
Dtype or latch. In DDR mode, two of the inputs are fed into registers on the positive edge of the clock. At the next
clock cycle, one of the registered outputs is also latched.
A multiplexer running off the same clock is used to switch the mux between the 11 and 01 inputs that will then feed
the output.
A gearbox function can be implemented in the output register block that takes four data streams: OPOSA, ONEGA,
OPOSB and ONEGB. All four data inputs are registered on the positive edge of the system clock and two of them
are also latched. The data is then output at a high rate using a multiplexer that runs off the DQCLK0 and DQCLK1
clocks. DQCLK0 and DQCLK1 are used in this case to transfer data from the system clock to the edge clock
domain. These signals are generated in the DQS Write Control Logic block. See Figure 2-37 for an overview of the
DQS write control logic.
Please see TN1180, LatticeECP3 High-Speed I/O Interface for more information on this topic.
Further discussion on using the DQS strobe in this module is discussed in the DDR Memory section of this data
sheet.
2-34
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-34. ECP3-70/95 (E or EA) Output and Tristate Block for Left and Right Edges
Tristate Logic
TS
TO
D Q
D Q
CE
R
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Output Logic
OPOSA
D Q
CE
R
ONEGA
D Q
A
B
OPOSB
C1
D Q
D Q
11
DO
10
C
00
D
01
ISI
L
ONEGB
D Q
D1
D Q
L
DDR Gearing &
ISI Correction
Clock
Transfer
Registers
SCLK
DQCLK1
Config Bit
DQCLK0
Tristate Register Block
The tristate register block registers tri-state control signals from the core of the device before they are passed to the
sysI/O buffers. The block contains a register for SDR operation and an additional register for DDR operation.
In SDR and non-gearing DDR modes, TS input feeds one of the flip-flops that then feeds the output. In DDRX2
mode, the register TS input is fed into another register that is clocked using the DQCLK0 and DQCLK1 signals. The
output of this register is used as a tristate control.
ISI Calibration
The setting for Inter-Symbol Interference (ISI) cancellation occurs in the output register block. ISI correction is only
available in the DDRX2 modes. ISI calibration settings exist once per output register block, so each I/O in a DQS12 group may have a different ISI calibration setting.
The ISI block extends output signals at certain times, as a function of recent signal history. So, if the output pattern
consists of a long strings of 0's to long strings of 1's, there are no delays on output signals. However, if there are
quick, successive transitions from 010, the block will stretch out the binary 1. This is because the long trail of 0's will
cause these symbols to interfere with the logic 1. Likewise, if there are quick, successive transitions from 101, the
block will stretch out the binary 0. This block is controlled by a 3-bit delay control that can be set in the DQS control
logic block.
For more information about this topic, please see the list of technical documentation at the end of this data sheet.
2-35
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Control Logic Block
The control logic block allows the selection and modification of control signals for use in the PIO block.
DDR Memory Support
Certain PICs have additional circuitry to allow the implementation of high-speed source synchronous and DDR1,
DDR2 and DDR3 memory interfaces. The support varies by the edge of the device as detailed below.
Left and Right Edges
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
The left and right sides of the PIC have fully functional elements supporting DDR1, DDR2, and DDR3 memory
interfaces. One of every 12 PIOs supports the dedicated DQS pins with the DQS control logic block. Figure 2-35
shows the DQS bus spanning 11 I/O pins. Two of every 12 PIOs support the dedicated DQS and DQS# pins with
the DQS control logic block.
Bottom Edge
PICs on the bottom edge of the device do not support DDR memory and Generic DDR interfaces.
Top Edge
PICs on the top side are similar to the PIO elements on the left and right sides but do not support gearing on the
output registers. Hence, the modes to support output/tristate DDR3 memory are removed on the top side.
The exact DQS pins are shown in a dual function in the Logic Signal Connections table in this data sheet. Additional detail is provided in the Signal Descriptions table. The DQS signal from the bus is used to strobe the DDR
data from the memory into input register blocks. Interfaces on the left, right and top edges are designed for DDR
memories that support 10 bits of data.
Figure 2-35. DQS Grouping on the Left, Right and Top Edges
PIO A
PADA "T"
PIO B
PADB "C"
LVDS Pair
PIO A
PADA "T"
PIO B
PADB "C"
PIO A
PADA "T"
PIO B
PADB "C"
LVDS Pair
LVDS Pair
PIO A
DQS
sysIO
Buffer
Delay
PIO B
Assigned
DQS Pin
PADA "T"
LVDS Pair
PADB "C"
PIO A
PADA "T"
PIO B
PADB "C"
PIO A
PADA "T"
PIO B
PADB "C"
LVDS Pair
LVDS Pair
DLL Calibrated DQS Delay Block
Source synchronous interfaces generally require the input clock to be adjusted in order to correctly capture data at
the input register. For most interfaces, a PLL is used for this adjustment. However, in DDR memories the clock
2-36
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
(referred to as DQS) is not free-running so this approach cannot be used. The DQS Delay block provides the
required clock alignment for DDR memory interfaces.
The delay required for the DQS signal is generated by two dedicated DLLs (DDR DLL) on opposite side of the
device. Each DLL creates DQS delays in its half of the device as shown in Figure 2-36. The DDR DLL on the left
side will generate delays for all the DQS Strobe pins on Banks 0, 7 and 6 and DDR DLL on the right will generate
delays for all the DQS pins on Banks 1, 2 and 3. The DDR DLL loop compensates for temperature, voltage and process variations by using the system clock and DLL feedback loop. DDR DLL communicates the required delay to
the DQS delay block using a 7-bit calibration bus (DCNTL[6:0])
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
The DQS signal (selected PIOs only, as shown in Figure 2-35) feeds from the PAD through a DQS control logic
block to a dedicated DQS routing resource. The DQS control logic block consists of DQS Read Control logic block
that generates control signals for the read side and DQS Write Control logic that generates the control signals
required for the write side. A more detailed DQS control diagram is shown in Figure 2-37, which shows how the
DQS control blocks interact with the data paths.
The DQS Read control logic receives the delay generated by the DDR DLL on its side and delays the incoming
DQS signal by 90 degrees. This delayed ECLKDQSR is routed to 10 or 11 DQ pads covered by that DQS signal.
This block also contains a polarity control logic that generates a DDRCLKPOL signal, which controls the polarity of
the clock to the sync registers in the input register blocks. The DQS Read control logic also generates a DDRLAT
signal that is in the input register block to transfer data from the first set of DDR register to the second set of DDR
registers when using the DDRX2 gearbox mode for DDR3 memory interface.
The DQS Write control logic block generates the DQCLK0 and DQCLK1 clocks used to control the output gearing
in the Output register block which generates the DDR data output and the DQS output. They are also used to control the generation of the DQS output through the DQS output register block. In addition to the DCNTL [6:0] input
from the DDR DLL, the DQS Write control block also uses a Dynamic Delay DYN DEL [7:0] attribute which is used
to further delay the DQS to accomplish the write leveling found in DDR3 memory. Write leveling is controlled by the
DDR memory controller implementation. The DYN DELAY can set 128 possible delay step settings. In addition, the
most significant bit will invert the clock for a 180-degree shift of the incoming clock. This will generate the DQSW
signal used to generate the DQS output in the DQS output register block.
Figure 2-36 and Figure 2-37 show how the DQS transition signals that are routed to the PIOs.
Please see TN1180, LatticeECP3 High-Speed I/O Interface for more information on this topic.
2-37
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-36. Edge Clock, DLL Calibration and DQS Local Bus Distribution
Bank 0
DQS
DQS
DQS
Bank 1
DQS
DQS
DQS
DQS
Configuration Bank
DQS
DQS
DQS
DQS Delay Control Bus
ECLK1
ECLK2
Bank 7
DQS
DQS
DDR DLL
(Right)
DQS
DDR DLL
(Left)
Bank 2
DQS
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
DQS
DQS
DQCLK0
DQCLK1
DDRLAT
DDRCLKPOL
ECLKDQSR
DATAVALID
DQS
DQS
DQS
DQS
DQS
DQS Strobe and Transition Detect Logic
I/O Ring
*Includes shared configuration I/Os and dedicated configuration I/Os.
2-38
Bank 3
Bank 6
DQS
DQS
SERDES
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
DCNTL[6:0]
ECLKDQSR
DDRCLKPOL
DDRLAT
DQSW
DQCLK1
DQCLK0
Figure 2-37. DQS Local Bus
DDR Data
Pad
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Data Output Register Block
Data Input Register Block
DQS Output Register Block
DQS Write Control Logic
DQS
Pad
DQS Read Control Logic
DQS Delay Block
DDR DLL
Polarity Control Logic
In a typical DDR Memory interface design, the phase relationship between the incoming delayed DQS strobe and
the internal system clock (during the READ cycle) is unknown. The LatticeECP3 family contains dedicated circuits
to transfer data between these domains. A clock polarity selector is used to prevent set-up and hold violations at
the domain transfer between DQS (delayed) and the system clock. This changes the edge on which the data is registered in the synchronizing registers in the input register block. This requires evaluation at the start of each READ
cycle for the correct clock polarity.
Prior to the READ operation in DDR memories, DQS is in tristate (pulled by termination). The DDR memory device
drives DQS low at the start of the preamble state. A dedicated circuit detects the first DQS rising edge after the preamble state. This signal is used to control the polarity of the clock to the synchronizing registers.
DDR3 Memory Support
LatticeECP3 supports the read and write leveling required for DDR3 memory interfaces.
Read leveling is supported by the use of the DDRCLKPOL and the DDRLAT signals generated in the DQS Read
Control logic block. These signals dynamically control the capture of the data with respect to the DQS at the input
register block.
2-39
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
To accomplish write leveling in DDR3, each DQS group has a slightly different delay that is set by DYN DELAY[7:0]
in the DQS Write Control logic block. The DYN DELAY can set 128 possible delay step settings. In addition, the
most significant bit will invert the clock for a 180-degree shift of the incoming clock.
LatticeECP3 input and output registers can also support DDR gearing that is used to receive and transmit the high
speed DDR data from and to the DDR3 Memory.
LatticeECP3 supports the 1.5V SSTL I/O standard required for the DDR3 memory interface. In addition, it supports
on-chip termination to VTT on the DDR3 memory input pins. For more information, refer to the sysIO section of this
data sheet.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Please see TN1180, LatticeECP3 High-Speed I/O Interface for more information on DDR Memory interface implementation in LatticeECP3.
sysI/O Buffer
Each I/O is associated with a flexible buffer referred to as a sysI/O buffer. These buffers are arranged around the
periphery of the device in groups referred to as banks. The sysI/O buffers allow users to implement the wide variety
of standards that are found in today’s systems including LVDS, BLVDS, HSTL, SSTL Class I & II, LVCMOS, LVTTL,
LVPECL, PCI.
sysI/O Buffer Banks
LatticeECP3 devices have six sysI/O buffer banks: six banks for user I/Os arranged two per side. The banks on the
bottom side are wraparounds of the banks on the lower right and left sides. The seventh sysI/O buffer bank (Configuration Bank) is located adjacent to Bank 2 and has dedicated/shared I/Os for configuration. When a shared pin is
not used for configuration it is available as a user I/O. Each bank is capable of supporting multiple I/O standards.
Each sysI/O bank has its own I/O supply voltage (VCCIO). In addition, each bank, except the Configuration Bank,
has voltage references, VREF1 and VREF2, which allow it to be completely independent from the others. The Configuration Bank top side shares VREF1 and VREF2 from sysI/O bank 1 and right side shares VREF1 and VREF2 from
sysI/O bank 2. Figure 2-38 shows the seven banks and their associated supplies.
In LatticeECP3 devices, single-ended output buffers and ratioed input buffers (LVTTL, LVCMOS and PCI) are powered using VCCIO. LVTTL, LVCMOS33, LVCMOS25 and LVCMOS12 can also be set as fixed threshold inputs independent of VCCIO.
Each bank can support up to two separate VREF voltages, VREF1 and VREF2, that set the threshold for the referenced input buffers. Some dedicated I/O pins in a bank can be configured to be a reference voltage supply pin.
Each I/O is individually configurable based on the bank’s supply and reference voltages.
2-40
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-38. LatticeECP3 Banks
TOP
V
REF1(1)
VREF2(1)
Configuration
Bank
JTAG Bank
Bank 1
GND
V REF2(6)
V CCIO6
Bank 3
V REF1(6)
Bank 6
LEFT
V
CCIO7
Bank 2
V REF2(7)
Bank 7
V REF1(7)
GND
VREF1(2)
V REF2(2)
VCCIO2
GND
RIGHT
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
VCCIO1
GND
VREF1(0)
VREF2(0)
V
CCIO0
GND
Bank 0
V REF1(3)
V REF2(3)
VCCIO3
GND
SERDES
Quads
BOTTOM
LatticeECP3 devices contain two types of sysI/O buffer pairs.
1. Top (Bank 0 and Bank 1) and Bottom sysI/O Buffer Pairs (Single-Ended Outputs Only)
The sysI/O buffer pairs in the top banks of the device consist of two single-ended output drivers and two sets of
single-ended input buffers (both ratioed and referenced). One of the referenced input buffers can also be configured as a differential input. Only the top edge buffers have a programmable PCI clamp.
The two pads in the pair are described as “true” and “comp”, where the true pad is associated with the positive
side of the differential input buffer and the comp (complementary) pad is associated with the negative side of
the differential input buffer.
On the top and bottom sides, there is no support for programmable on-chip input termination, which is required
for DQ and DQS pins for DDR3 interface. This side is ideal for ADDR/CMD signals of DDR3, general purpose
I/O, PCI, TR-LVDS (transition reduced LVDS) or LVDS inputs. Only the top I/O banks support hot socketing
with IDK specified under the Hot Socketing Specifications. The configuration bank is not hot-socketable.
2-41
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
2. Left and Right (Banks 2, 3, 6 and 7) sysI/O Buffer Pairs (50% Differential and 100% Single-Ended Outputs)
The sysI/O buffer pairs in the left and right banks of the device consist of two single-ended output drivers, two
sets of single-ended input buffers (both ratioed and referenced) and one differential output driver. One of the
referenced input buffers can also be configured as a differential input. In these banks the two pads in the pair
are described as “true” and “comp”, where the true pad is associated with the positive side of the differential I/O,
and the comp (complementary) pad is associated with the negative side of the differential I/O.
In addition, programmable on-chip input termination (parallel or differential, static or dynamic) is supported on
these sides, which is required for DDR3 interface. However, there is no support for hot-socketing on these
sides as the clamp is always present.
LVDS, RSDS, PPLVDS and Mini-LVDS differential output drivers are available on 50% of the buffer pairs on the
left and right banks.
3. Configuration Bank sysI/O Buffer Pairs (Single-Ended Outputs, Only on Shared Pins When Not Used by
Configuration)
The sysI/O buffers in the Configuration Bank consist of single-ended output drivers and single-ended input buffers (both ratioed and referenced). The referenced input buffer can also be configured as a differential input.
The two pads in the pair are described as “true” and “comp”, where the true pad is associated with the positive
side of the differential input buffer and the comp (complementary) pad is associated with the negative side of
the differential input buffer.
Programmable PCI clamps are only available on top banks (PCI clamps are used primarily on inputs and bidirectional pads to reduce ringing on the receiving end) can also be used on inputs.
Typical sysI/O I/O Behavior During Power-up
The internal power-on-reset (POR) signal is deactivated when VCC, VCCIO8 and VCCAUX have reached satisfactory
levels. After the POR signal is deactivated, the FPGA core logic becomes active. It is the user’s responsibility to
ensure that all other VCCIO banks are active with valid input logic levels to properly control the output logic states of
all the I/O banks that are critical to the application. For more information about controlling the output logic state with
valid input logic levels during power-up in LatticeECP3 devices, see the list of technical documentation at the end
of this data sheet.
The VCC and VCCAUX supply the power to the FPGA core fabric, whereas the VCCIO supplies power to the I/O buffers. In order to simplify system design while providing consistent and predictable I/O behavior, it is recommended
that the I/O buffers be powered-up prior to the FPGA core fabric. VCCIO supplies should be powered-up before or
together with the VCC and VCCAUX supplies.
Supported sysI/O Standards
The LatticeECP3 sysI/O buffer supports both single-ended and differential standards. Single-ended standards can
be further subdivided into LVCMOS, LVTTL and other standards. The buffers support the LVTTL, LVCMOS 1.2V,
1.5V, 1.8V, 2.5V and 3.3V standards. In the LVCMOS and LVTTL modes, the buffer has individual configuration
options for drive strength, slew rates, bus maintenance (weak pull-up, weak pull-down, or a bus-keeper latch) and
open drain. Other single-ended standards supported include SSTL and HSTL. Differential standards supported
include LVDS, BLVDS, LVPECL, MLVDS, RSDS, Mini-LVDS, PPLVDS (point-to-point LVDS), TRLVDS (Transition
Reduced LVDS), differential SSTL and differential HSTL. Tables 2-13 and 2-14 show the I/O standards (together
with their supply and reference voltages) supported by LatticeECP3 devices. For further information on utilizing the
sysI/O buffer to support a variety of standards please see TN1177, LatticeECP3 sysIO Usage Guide.
2-42
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
On-Chip Programmable Termination
The LatticeECP3 supports a variety of programmable on-chip terminations options, including:
• Dynamically switchable Single Ended Termination for SSTL15 inputs with programmable resistor values of 40,
50, or 60 ohms. This is particularly useful for low power JEDEC compliant DDR3 memory controller implementations. External termination to Vtt should be used for DDR2 memory controller implementation.
• Common mode termination of 80, 100, 120 ohms for differential inputs
Figure 2-39. On-Chip Termination
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Vtt
Control Signal
Z0
+
Vtt*
Zo
-
+
Z0
-
Off-chip
Off-chip
On-Chip
On-Chip
*Vtt must be left floating for this termination
Programmable resistance (40, 50 and 60 Ohms)
Parallel Single-Ended Input
Differential Input
See Table 2-12 for termination options for input modes.
Table 2-12. On-Chip Termination Options for Input Modes
IO_TYPE
TERMINATE to VTT1, 2 DIFFRENTIAL TERMINATION RESISTOR1
LVDS25
þ
80, 100, 120
BLVDS25
þ
80, 100, 120
MLVDS
þ
80, 100, 120
HSTL18_I
40, 50, 60
þ
HSTL18_II
40, 50, 60
þ
HSTL18D_I
40, 50, 60
þ
HSTL18D_II
40, 50, 60
þ
HSTL15_I
40, 50, 60
þ
HSTL15D_I
40, 50, 60
þ
SSTL25_I
40, 50, 60
þ
SSTL25_II
40, 50, 60
þ
SSTL25D_I
40, 50, 60
þ
SSTL25D_II
40, 50, 60
þ
SSTL18_I
40, 50, 60
þ
SSTL18_II
40, 50, 60
þ
SSTL18D_I
40, 50, 60
þ
SSTL18D_II
40, 50, 60
þ
SSTL15
40, 50, 60
þ
SSTL15D
40, 50, 60
þ
1. TERMINATE to VTT and DIFFRENTIAL TERMINATION RESISTOR when turn on can only have
one setting per bank. Only left and right banks have this feature.
Use of TERMINATE to VTT and DIFFRENTIAL TERMINATION RESISTOR are mutually exclusive
in an I/O bank.
On-chip termination tolerance +/- 20%
2. External termination to VTT should be used when implementing DDR2 memory controller.
2-43
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Please see TN1177, LatticeECP3 sysIO Usage Guide for on-chip termination usage and value ranges.
Equalization Filter
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Equalization filtering is available for single-ended inputs on both true and complementary I/Os, and for differential
inputs on the true I/Os on the left, right, and top sides. Equalization is required to compensate for the difficulty of
sampling alternating logic transitions with a relatively slow slew rate. It is considered the most useful for the Input
DDRX2 modes, used in DDR3 memory, LVDS, or TRLVDS signaling. Equalization filter acts as a tunable filter with
settings to determine the level of correction. In the LatticeECP3 devices, there are four settings available: 0 (none),
1, 2 and 3. The default setting is 0. The equalization logic resides in the sysI/O buffers, the two bits of setting is set
uniquely in each input IOLOGIC block. Therefore, each sysI/O can have a unique equalization setting within a
DQS-12 group.
Hot Socketing
LatticeECP3 devices have been carefully designed to ensure predictable behavior during power-up and powerdown. During power-up and power-down sequences, the I/Os remain in tri-state until the power supply voltage is
high enough to ensure reliable operation. In addition, leakage into I/O pins is controlled within specified limits.
Please refer to the Hot Socketing Specifications in the DC and Switching Characteristics in this data sheet.
SERDES and PCS (Physical Coding Sublayer)
LatticeECP3 devices feature up to 16 channels of embedded SERDES/PCS arranged in quads at the bottom of the
devices supporting up to 3.2Gbps data rate. Figure 2-40 shows the position of the quad blocks for the LatticeECP3150 devices. Table 2-14 shows the location of available SERDES Quads for all devices.
The LatticeECP3 SERDES/PCS supports a range of popular serial protocols, including:
• PCI Express 1.1
• Ethernet (XAUI, GbE - 1000 Base CS/SX/LX and SGMII)
• Serial RapidIO
• SMPTE SDI (3G, HD, SD)
• CPRI
• SONET/SDH (STS-3, STS-12, STS-48)
Each quad contains four dedicated SERDES for high speed, full duplex serial data transfer. Each quad also has a
PCS block that interfaces to the SERDES channels and contains protocol specific digital logic to support the standards listed above. The PCS block also contains interface logic to the FPGA fabric. All PCS logic for dedicated protocol support can also be bypassed to allow raw 8-bit or 10-bit interfaces to the FPGA fabric.
Even though the SERDES/PCS blocks are arranged in quads, multiple baud rates can be supported within a quad
with the use of dedicated, per channel 1, 2 and 11 rate dividers. Additionally, multiple quads can be arranged
together to form larger data pipes.
For information on how to use the SERDES/PCS blocks to support specific protocols, as well on how to combine
multiple protocols and baud rates within a device, please refer to TN1176, LatticeECP3 SERDES/PCS Usage
Guide.
2-44
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 2-40. SERDES/PCS Quads (LatticeECP3-150)
sysIO Bank 1
CH0
CH1
CH2
CH3
SERDES/PCS
Quad C
CH0
CH1
CH2
SERDES/PCS
Quad A
CH3
CH0
CH1
CH2
SERDES/PCS
Quad B
CH3
CH0
CH1
CH2
CH3
SERDES/PCS
Quad D
sysIO Bank 3
sysIO Bank 6
sysIO Bank 2
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
sysIO Bank 7
Configuration Bank
sysIO Bank 0
Table 2-13. LatticeECP3 SERDES Standard Support
Data Rate
(Mbps)
Number of
General/Link Width
PCI Express 1.1
2500
x1, x2, x4
8b10b
Gigabit Ethernet
1250, 2500
x1
8b10b
SGMII
1250
x1
8b10b
XAUI
3125
x4
8b10b
Serial RapidIO Type I,
Serial RapidIO Type II,
Serial RapidIO Type III
1250,
2500,
3125
x1, x4
8b10b
614.4,
1228.8,
2457.6,
3072.0
x1
8b10b
1431,
1771,
270,
360,
540
x1
NRZI/Scrambled
Standard
CPRI-1,
CPRI-2,
CPRI-3,
CPRI-4
SD-SDI
(259M, 344M)
Encoding Style
HD-SDI
(292M)
1483.5,
1485
x1
NRZI/Scrambled
3G-SDI
(424M)
2967,
2970
x1
NRZI/Scrambled
SONET-STS-32
155.52
x1
N/A
SONET-STS-122
622.08
x1
N/A
2488
x1
N/A
2
SONET-STS-48
1. For slower rates, the SERDES are bypassed and CML signals are directly connected to the FPGA routing.
2. The SONET protocol is supported in 8-bit SERDES mode. See TN1176 Lattice ECP3 SERDES/PCS Usage Guide for more information.
2-45
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Table 2-14. Available SERDES Quads per LatticeECP3 Devices
Package
ECP3-17
ECP3-35
ECP3-70
ECP3-95
ECP3-150
—
256 ftBGA
1
1
—
—
484 ftBGA
1
1
1
1
672 ftBGA
—
1
2
2
2
1156 ftBGA
—
—
3
3
4
SERDES Block
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
A SERDES receiver channel may receive the serial differential data stream, equalize the signal, perform Clock and
Data Recovery (CDR) and de-serialize the data stream before passing the 8- or 10-bit data to the PCS logic. The
SERDES transmitter channel may receive the parallel 8- or 10-bit data, serialize the data and transmit the serial bit
stream through the differential drivers. Figure 2-41 shows a single-channel SERDES/PCS block. Each SERDES
channel provides a recovered clock and a SERDES transmit clock to the PCS block and to the FPGA core logic.
Each transmit channel, receiver channel, and SERDES PLL shares the same power supply (VCCA). The output
and input buffers of each channel have their own independent power supplies (VCCOB and VCCIB).
Figure 2-41. Simplified Channel Block Diagram for SERDES/PCS Block
SERDES
PCS
FPGA Core
Recovered Clock*
RX_REFCLK
HDINP
Equalizer
HDINN
Recovered Clock
Data
Clock/Data
Recovery Clock
Deserializer
1:8/1:10
Receiver
Polarity
Adjust
Word Alignment
8b10b Decoder
Bypass
CTC
Downsample
FIFO
Bypass
Bypass
Receive Data
Receive Clock
SERDES Transmit Clock*
TX REFCLK
HDOUTP
SERDES Transmit Clock
TX PLL
Serializer
8:1/10:1
HDOUTN
Transmitter
8b10b
Encoder
Polarity
Adjust
Bypass
Bypass
Upsample
FIFO
Transmit Data
Transmit Clock
* 1/8 or 1/10 line rate
PCS
As shown in Figure 2-41, the PCS receives the parallel digital data from the deserializer and selects the polarity,
performs word alignment, decodes (8b/10b), provides Clock Tolerance Compensation and transfers the clock
domain from the recovered clock to the FPGA clock via the Down Sample FIFO.
For the transmit channel, the PCS block receives the parallel data from the FPGA core, encodes it with 8b/10b,
selects the polarity and passes the 8/10 bit data to the transmit SERDES channel.
The PCS also provides bypass modes that allow a direct 8-bit or 10-bit interface from the SERDES to the FPGA
logic. The PCS interface to the FPGA can also be programmed to run at 1/2 speed for a 16-bit or 20-bit interface to
the FPGA logic.
SCI (SERDES Client Interface) Bus
The SERDES Client Interface (SCI) is an IP interface that allows the SERDES/PCS Quad block to be controlled by
registers rather than the configuration memory cells. It is a simple register configuration interface that allows
SERDES/PCS configuration without power cycling the device.
2-46
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
The ispLEVER design tools from Lattice support all modes of the PCS. Most modes are dedicated to applications
associated with a specific industry standard data protocol. Other more general purpose modes allow users to
define their own operation. With ispLEVER, the user can define the mode for each quad in a design.
Popular standards such as 10Gb Ethernet, x4 PCI Express and 4x Serial RapidIO can be implemented using IP
(available through Lattice), a single quad (Four SERDES channels and PCS) and some additional logic from the
core.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
The LatticeECP3 family also supports a wide range of primary and secondary protocols. Within the same quad, the
LatticeECP3 family can support mixed protocols with semi-independent clocking as long as the required clock frequencies are integer x1, x2, or x11 multiples of each other. Table 2-15 lists the allowable combination of primary
and secondary protocol combinations.
Flexible Quad SERDES Architecture
The LatticeECP3 family SERDES architecture is a quad-based architecture. For most SERDES settings and standards, the whole quad (consisting of four SERDES) is treated as a unit. This helps in silicon area savings, better
utilization and overall lower cost.
However, for some specific standards, the LatticeECP3 quad architecture provides flexibility; more than one standard can be supported within the same quad.
Table 2-15 shows the standards can be mixed and matched within the same quad. In general, the SERDES standards whose nominal data rates are either the same or a defined subset of each other, can be supported within the
same quad. In Table 2-15, the Primary Protocol column refers to the standard that determines the reference clock
and PLL settings. The Secondary Protocol column shows the other standard that can be supported within the
same quad.
Furthermore, Table 2-15 also implies that more than two standards in the same quad can be supported, as long as
they conform to the data rate and reference clock requirements. For example, a quad may contain PCI Express 1.1,
SGMII, Serial RapidIO Type I and Serial RapidIO Type II, all in the same quad.
Table 2-15. LatticeECP3 Primary and Secondary Protocol Support
Primary Protocol
Secondary Protocol
PCI Express 1.1
SGMII
PCI Express 1.1
Gigabit Ethernet
PCI Express 1.1
Serial RapidIO Type I
PCI Express 1.1
Serial RapidIO Type II
Serial RapidIO Type I
SGMII
Serial RapidIO Type I
Gigabit Ethernet
Serial RapidIO Type II
SGMII
Serial RapidIO Type II
Gigabit Ethernet
Serial RapidIO Type II
Serial RapidIO Type I
CPRI-3
CPRI-2 and CPRI-1
3G-SDI
HD-SDI and SD-SDI
For further information on SERDES, please see TN1176, LatticeECP3 SERDES/PCS Usage Guide.
IEEE 1149.1-Compliant Boundary Scan Testability
All LatticeECP3 devices have boundary scan cells that are accessed through an IEEE 1149.1 compliant Test
Access Port (TAP). This allows functional testing of the circuit board on which the device is mounted through a
serial scan path that can access all critical logic nodes. Internal registers are linked internally, allowing test data to
be shifted in and loaded directly onto test nodes, or test data to be captured and shifted out for verification. The test
2-47
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
access port consists of dedicated I/Os: TDI, TDO, TCK and TMS. The test access port has its own supply voltage
VCCJ and can operate with LVCMOS3.3, 2.5, 1.8, 1.5 and 1.2 standards.
For more information, please see TN1169, LatticeECP3 sysCONFIG Usage Guide.
Device Configuration
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
All LatticeECP3 devices contain two ports that can be used for device configuration. The Test Access Port (TAP),
which supports bit-wide configuration, and the sysCONFIG port, support dual-byte, byte and serial configuration.
The TAP supports both the IEEE Standard 1149.1 Boundary Scan specification and the IEEE Standard 1532 InSystem Configuration specification. The sysCONFIG port includes seven I/Os used as dedicated pins with the
remaining pins used as dual-use pins. See TN1169, LatticeECP3 sysCONFIG Usage Guide for more information
about using the dual-use pins as general purpose I/Os.
There are various ways to configure a LatticeECP3 device:
1. JTAG
2. Standard Serial Peripheral Interface (SPI and SPIm modes) - interface to boot PROM memory
3. System microprocessor to drive a x8 CPU port (PCM mode)
4. System microprocessor to drive a serial slave SPI port (SSPI mode)
5. Generic byte wide flash with a MachXO™ device, providing control and addressing
On power-up, the FPGA SRAM is ready to be configured using the selected sysCONFIG port. Once a configuration
port is selected, it will remain active throughout that configuration cycle. The IEEE 1149.1 port can be activated any
time after power-up by sending the appropriate command through the TAP port.
LatticeECP3 devices also support the Slave SPI Interface. In this mode, the FPGA behaves like a SPI Flash device
(slave mode) with the SPI port of the FPGA to perform read-write operations.
Enhanced Configuration Options
LatticeECP3 devices have enhanced configuration features such as: decryption support, TransFR™ I/O and dualboot image support.
1. TransFR (Transparent Field Reconfiguration)
TransFR I/O (TFR) is a unique Lattice technology that allows users to update their logic in the field without
interrupting system operation using a single ispVM command. TransFR I/O allows I/O states to be frozen during device configuration. This allows the device to be field updated with a minimum of system disruption and
downtime. See TN1087, Minimizing System Interruption During Configuration Using TransFR Technology for
details.
2. Dual-Boot Image Support
Dual-boot images are supported for applications requiring reliable remote updates of configuration data for the
system FPGA. After the system is running with a basic configuration, a new boot image can be downloaded
remotely and stored in a separate location in the configuration storage device. Any time after the update the
LatticeECP3 can be re-booted from this new configuration file. If there is a problem, such as corrupt data during download or incorrect version number with this new boot image, the LatticeECP3 device can revert back to
the original backup golden configuration and try again. This all can be done without power cycling the system.
For more information, please see TN1169, LatticeECP3 sysCONFIG Usage Guide.
Soft Error Detect (SED) Support
LatticeECP3 devices have dedicated logic to perform Cycle Redundancy Code (CRC) checks. During configuration, the configuration data bitstream can be checked with the CRC logic block. In addition, the LatticeECP3 device
2-48
Architecture
LatticeECP3 Family Data Sheet
Lattice Semiconductor
can also be programmed to utilize a Soft Error Detect (SED) mode that checks for soft errors in configuration
SRAM. The SED operation can be run in the background during user mode. If a soft error occurs, during user
mode (normal operation) the device can be programmed to generate an error signal.
For further information on SED support, please see TN1184, LatticeECP3 Soft Error Detection (SED) Usage
Guide.
External Resistor
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
LatticeECP3 devices require a single external, 10K ohm ±1% value between the XRES pin and ground. Device
configuration will not be completed if this resistor is missing. There is no boundary scan register on the external
resistor pad.
On-Chip Oscillator
Every LatticeECP3 device has an internal CMOS oscillator which is used to derive a Master Clock (MCLK) for configuration. The oscillator and the MCLK run continuously and are available to user logic after configuration is completed. The software default value of the MCLK is nominally 2.5MHz. Table 2-16 lists all the available MCLK
frequencies. When a different Master Clock is selected during the design process, the following sequence takes
place:
1. Device powers up with a nominal Master Clock frequency of 3.1MHz.
2. During configuration, users select a different master clock frequency.
3. The Master Clock frequency changes to the selected frequency once the clock configuration bits are received.
4. If the user does not select a master clock frequency, then the configuration bitstream defaults to the MCLK frequency of 2.5MHz.
This internal CMOS oscillator is available to the user by routing it as an input clock to the clock tree. For further
information on the use of this oscillator for configuration or user mode, please see TN1169, LatticeECP3 sysCONFIG Usage Guide.
Table 2-16. Selectable Master Clock (MCLK) Frequencies During Configuration (Nominal)
MCLK (MHz)
MCLK (MHz)
MCLK (MHz)
1
2.5
10
41
3.1
13
45
4.3
15
51
5.4
20
55
6.9
26
60
8.1
30
130
9.2
34
—
1. Software default MCLK frequency. Hardware default is
3.1MHz.
Density Shifting
The LatticeECP3 family is designed to ensure that different density devices in the same family and in the same
package have the same pinout. Furthermore, the architecture ensures a high success rate when performing design
migration from lower density devices to higher density devices. In many cases, it is also possible to shift a lower utilization design targeted for a high-density device to a lower density device. However, the exact details of the final
resource utilization will impact the likelihood of success in each case. An example is that some user I/Os may
become No Connects in smaller devices in the same packge.
2-49
LatticeECP3 Family Data Sheet
DC and Switching Characteristics
March 2010
Preliminary Data Sheet DS1021
Absolute Maximum Ratings1, 2, 3
Supply Voltage VCC . . . . . . . . . . . . . . . . . . . -0.5 to 1.32V
Supply Voltage VCCAUX . . . . . . . . . . . . . . . . -0.5 to 3.75V
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Supply Voltage VCCJ . . . . . . . . . . . . . . . . . . -0.5 to 3.75V
Output Supply Voltage VCCIO . . . . . . . . . . . -0.5 to 3.75V
Input or I/O Tristate Voltage Applied4 . . . . . . -0.5 to 3.75V
Storage Temperature (Ambient) . . . . . . . . . -65 to 150°C
Junction Temperature (Tj) . . . . . . . . . . . . . . . . . . +125°C
1. Stress above those listed under the “Absolute Maximum Ratings” may cause permanent damage to the device. Functional operation of the
device at these or any other conditions above those indicated in the operational sections of this specification is not implied.
2. Compliance with the Lattice Thermal Management document is required.
3. All voltages referenced to GND.
4. Overshoot and undershoot of -2V to (VIHMAX + 2) volts is permitted for a duration of 10 Bits Wide) Centered at Pin (GDDRX1_RX.SCLK.Centered) Using PCLK
Pin for Clock Input
Data Left, Right and Top Sides & Clock Left, Right and Top Sides
tSUGDDR
Data Setup Before CLK
ECP3-150EA
—
—
—
ps
tHGDDR
Data Hold After CLK
ECP3-150EA
—
—
—
ps
fMAX_GDDR
DDRX1 Clock Frequency
ECP3-150EA
—
—
—
MHz
Generic DDRX1 Inputs with Clock in the Center of Data Window, without DLL (GDDRX1_RX.ECLK.Centered)
tSUGDDR
Data Setup Before CLK
ECP3-70E/95E
515
—
515
—
515
—
ps
tHOGDDR
Data Hold After CLK
ECP3-70E/95E
515
—
515
—
515
—
ps
fMAX_GDDR
DDRX1 Clock Frequency
ECP3-70E/95E
—
250
—
250
—
250
MHz
Generic DDRX1 Inputs with Clock and Data (> 10 Bits Wide) Aligned at Pin (GDDRX1_RX.SCLK.Aligned) using DLLCLKIN Pin for Clock Input
Data Left, Right and Top Sides & Clock Left and Right Sides
tDVACLKGDDR
Data Setup Before CLK
ECP3-150EA
tDVECLKGDDR
Data Hold After CLK
ECP3-150EA
fMAX_GDDR
DDRX1 Clock Frequency
ECP3-150EA
—
—
—
—
—
—
—
UI
—
—
UI
MHz
Generic DDRX1 Inputs with Clock and Data Aligned, with DLL (GDDRX1_RX.ECLK.Aligned)
tDVACLKGDDR
Data Setup Before CLK
ECP3-70E/95E
tDVECLKGDDR
Data Hold After CLK
ECP3-70E/95E 0.765
3-17
—
0.235
—
0.235
—
0.235
UI
—
0.765
—
0.765
—
UI
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 External Switching Characteristics (Continued)1, 2
Over Recommended Commercial Operating Conditions
-8
Parameter
fMAX_GDDR
Description
DDRX1 Clock Frequency
-7
-6
Device
Min.
Max.
Min.
Max.
Min.
ECP3-70E/95E
—
250
—
250
—
Max. Units
250
MHz
Generic DDRX1 Inputs with Clock and Data (10 Bits Wide) Aligned at Pin (GDDRX2_RX.ECLK.Aligned)
Left and Right Side Using DLLCLKIN Pin for Clock Input
tDVACLKGDDR
Data Setup Before CLK (Left and
Right Side)
ECP3-150EA
tDVECLKGDDR
Data Hold After CLK (Left and Right
Side)
ECP3-150EA
fMAX_GDDR
DDRX1 Clock Frequency (Left and
Right Side)
ECP3-150EA
—
—
—
MHz
—
—
—
UI
—
—
—
—
—
UI
—
UI
Top Side Using PCLK Pin for Clock Input
tDVACLKGDDR
Data Setup Before CLK (Top Side)
ECP3-150EA
tDVECLKGDDR
Data Hold After CLK (Top Side)
ECP3-150EA
fMAX_GDDR
DDRX1 Clock Frequency (Top Side)
ECP3-150EA
—
—
—
—
—
—
UI
MHz
Generic DDRX2 Inputs with Clock and Data Edges Aligned, with DLLDEL3 (GDDRX2_RX.ECLK.Aligned)
tDVACLKGDDR
Data Valid After CLK
ECP3-70E/95E
3-18
—
0.235
—
0.235
—
0.235
UI
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 External Switching Characteristics (Continued)1, 2
Over Recommended Commercial Operating Conditions
-8
Parameter
Description
Device
Min.
tDVECLKGDDR
Data Hold After CLK
ECP3-70E/95E 0.765
fMAX_GDDR
DDR/DDRX2 Clock Frequency8
ECP3-70E/95E
—
-7
Max.
Min.
—
500
-6
Max.
Min.
Max. Units
0.765
—
0.765
—
UI
—
420
—
375
MHz
Generic DDRX2 Inputs with Clock and Data ( 10 Bits Wide) Aligned at Pin (GDDRX1_TX.SCLK.Aligned)
Left, Right and Top Sides
tDIBGDDR
Data Hold After CLK
ECP3-150EA
—
—
—
tDIAGDDR
Data Setup Before CLK
ECP3-150EA
—
—
—
fMAX_GDDR
DDRX1 Clock Frequency
ECP3-150EA
—
—
—
Generic DDRX1 Outputs with clock and data edge aligned, without PLL
tDIBGDDR
Data Invalid Before CLK
ECP3-70E/95E
—
330
—
330
—
330
tDIAGDDR
Data Invalid After CLK
ECP3-70E/95E
—
330
—
330
—
330
ps
fMAX_GDDR
DDRX1 Clock Frequency
ECP3-70E/95E
—
250
—
250
—
250
MHz
Generic DDRX1 Output with Clock and Data ( 10 Bits Wide) Aligned at Pin (GDDRX2_TX.ECLK.Aligned)
Left and Right Sides
Data Setup Before CLK
ECP3-150EA
—
—
—
ps
tDIAGDDR
Data Hold After CLK
ECP3-150EA
—
—
—
ps
fMAX_GDDR
DDRX2 Clock Frequency
ECP3-150EA
—
—
—
MHz
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
tDIBGDDR
Generic DDRX2 Outputs with Clock and Data Edges Aligned, Without PLL 90-degree shifted clock output5
(GDDRX2_TX.Aligned)
tDIBGDDR
Data Invalid Before Clock
ECP3-70E/95E
—
200
—
225
—
250
ps
tDIAGDDR
Data Invalid After Clock
ECP3-70E/95E
—
200
—
225
—
250
ps
fMAX_GDDR
DDR/DDRX2 Clock Frequency8
ECP3-70E/95E
—
500
—
420
—
375
MHz
Generic DDRX2 Output with Clock and Data (> 10 Bits Wide) Centered at Pin Using DQSDLL (GDDRX2_TX.DQSDLL.Centered)
Left and Right Sides
tDVBGDDR
Data Valid Before CLK
ECP3-150EA
tDVAGDDR
Data Valid After CLK
ECP3-150EA
fMAX_GDDR
DDRX2 Clock Frequency
ECP3-150EA
—
—
—
—
—
—
—
ns
—
ns
—
ns
Generic DDRX2 Output with Clock and Data (> 10 Bits Wide) Centered at Pin Using PLL (GDDRX2_TX.PLL.Centered)
Left and Right Sides
tDVBGDDR
Data Valid Before CLK
ECP3-150EA
—
—
—
ns
tDVAGDDR
Data Valid After CLK
ECP3-150EA
—
—
—
ns
fMAX_GDDR
DDRX2 Clock Frequency
ECP3-150EA
—
—
—
ns
Generic DDRX2 Outputs with Clock Edge in the Center of Data Window, with PLL 90-degree Shifted Clock Output6
(GDDRX2_TX.PLL.Centered)
tDVBGDDR
Data Valid Before CLK
ECP3-70E/95E
300
—
370
tDVAGDDR
Data Valid After CLK
ECP3-70E/95E
fMAX_GDDR
DDR/DDRX2 Clock Frequency8
ECP3-70E/95E
—
417
300
—
—
500
370
—
417
—
ps
—
420
—
375
MHz
-8
Parameter
Description
Device
Min.
-7
Max.
Min.
—
ps
-6
Max.
Min.
Max. Units
Memory Interface
DDR/DDR2 SDRAM I/O Pin Parameters (Input Data are Strobe Edge Aligned, Output Strobe Edge is Data Centered)4
tDVADQ
Data Valid After DQS (DDR Read)
ECP3-150EA
—
0.225
—
0.225
—
0.225
UI
tDVEDQ
Data Hold After DQS (DDR Read)
ECP3-150EA
0.64
—
0.64
—
0.64
—
UI
tDQVBS
Data Valid Before DQS
ECP3-150EA
0.25
—
0.25
—
0.25
—
UI
tDQVAS
Data Valid After DQS
ECP3-150EA
0.25
—
0.25
—
0.25
—
UI
fMAX_DDR
DDR Clock Frequency
ECP3-150EA
95
200
95
200
95
166
MHz
fMAX_DDR2
DDR2 clock frequency
ECP3-150EA
133
266
133
200
133
166
MHz
tDVADQ
Data Valid After DQS (DDR Read)
ECP3-70E/95E
—
0.225
—
0.225
—
0.225
UI
tDVEDQ
Data Hold After DQS (DDR Read)
ECP3-70E/95E
0.64
—
0.64
—
0.64
—
UI
tDQVBS
Data Valid Before DQS
ECP3-70E/95E
0.25
—
0.25
—
0.25
—
UI
tDQVAS
Data Valid After DQS
ECP3-70E/95E
0.25
—
0.25
—
0.25
—
UI
fMAX_DDR
DDR Clock Frequency
ECP3-70E/95E
95
200
95
200
95
133
MHz
3-20
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 External Switching Characteristics (Continued)1, 2
Over Recommended Commercial Operating Conditions
-8
Parameter
fMAX_DDR2
Description
DDR2 Clock Frequency
-7
-6
Device
Min.
Max.
Min.
Max.
Min.
Max. Units
ECP3-70E/95E
133
266
133
200
133
166
MHz
DDR3 (Using PLL for SCLK) I/O Pin Parameters
Data Valid After DQS (DDR Read)
ECP3-150EA
—
0.225
—
0.225
—
0.225
UI
tDVEDQ
Data Hold After DQS (DDR Read)
ECP3-150EA
0.64
—
0.64
—
0.64
—
UI
tDQVBS
Data Valid Before DQS
ECP3-150EA
0.25
—
0.25
—
0.25
—
UI
tDQVAS
Data Valid After DQS
ECP3-150EA
0.25
—
0.25
—
0.25
—
UI
fMAX_DDR3
DDR3 clock frequency
ECP3-150EA
266
400
266
333
266
300
MHz
1.
2.
3.
4.
5.
6.
7.
8.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
tDVADQ
Commercial timing numbers are shown. Industrial numbers are typically slower and can be extracted from the ispLEVER software.
General I/O timing numbers based on LVCMOS 2.5, 12mA, 0pf load.
Generic DDR timing numbers based on LVDS I/O.
DDR timing numbers based on SSTL25. DDR2 timing numbers based on SSTL18.
DDR3 timing numbers based on SSTL15.
Uses LVDS I/O standard.
The current version of software does not support per bank skew numbers; this will be supported in a future release.
Maximum clock frequencies are tested under best case conditions. System performance may vary upon the user environment.
3-21
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 3-6. Generic DDR/DDR2 (With Clock and Data Edges Aligned)
Transmit Parameters
t DIBGDDR
t DIAGDDR
CLK
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Data (TDAT, TCTL)
t DIAGDDR
t DIBGDDR
Receive Parameters
RDTCLK
Data (RDAT, RCTL)
t DVACLKGDDR
t DVACLKGDDR
t DVECLKGDDR
t DVECLKGDDR
3-22
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 3-7. DDR/DDR2/DDR3 SDRAM
Transmit Parameters
DQS
DQ
t DQVAS
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
t DQVBS
t DQVAS
t DQVBS
Receive Parameters
DQS
DQ
t DVADQ
t DVADQ
t DVEDQ
t DVEDQ
Figure 3-8. Generic DDR/DDR2 Parameters (With Clock Center on Data Window)
Transmit Parameters
CLOCK
DATA
tDVBCKGDDR
t DVACKGDDR
t DVACKGDDR
t DVBCKGDDR
Receive Parameters
CLOCK
DATA
t SUGDDR
t SUGDDR
t HGDDR
t HGDDR
3-23
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 Internal Switching Characteristics1, 2
Over Recommended Commercial Operating Conditions
-8
Parameter
Description
Min.
-7
Max.
Min.
-6
Max.
Min.
Max.
Units.
PFU/PFF Logic Mode Timing
LUT4 delay (A to D inputs to F output)
—
0.147
—
0.163
—
0.179
ns
tLUT6_PFU
LUT6 delay (A to D inputs to OFX output)
—
0.273
—
0.307
—
0.342
ns
tLSR_PFU
Set/Reset to output of PFU (Asynchronus)
—
0.593
—
0.674
—
0.756
ns
tLSRREC_PFU
Asynchronous Set/Reset recovery time for
PFU Logic
—
0.298
tSUM_PFU
Clock to Mux (M0,M1) Input Setup Time
0.134
tHM_PFU
Clock to Mux (M0,M1) Input Hold Time
-0.097
tSUD_PFU
Clock to D input setup time
tHD_PFU
Clock to D input hold time
tCK2Q_PFU
Clock to Q delay, (D-type Register
Configuration)
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
tLUT4_PFU
—
0.345
—
0.391
ns
—
0.144
—
-0.103
—
0.153
—
ns
—
-0.109
—
ns
0.061
—
0.068
—
0.075
—
ns
0.019
—
0.013
—
0.015
—
ns
—
0.243
—
0.273
—
0.303
ns
ns
PFU Dual Port Memory Mode Timing
tCORAM_PFU
Clock to Output (F Port)
—
0.710
—
0.803
—
0.897
tSUDATA_PFU
Data Setup Time
-0.137
—
-0.155
—
-0.174
—
ns
tHDATA_PFU
Data Hold Time
0.188
—
0.217
—
0.246
—
ns
-0.227
—
-0.257
—
-0.286
—
ns
tSUADDR_PFU
Address Setup Time
tHADDR_PFU
Address Hold Time
0.240
—
0.275
—
0.310
—
ns
tSUWREN_PFU
Write/Read Enable Setup Time
-0.055
—
-0.055
—
-0.063
—
ns
tHWREN_PFU
Write/Read Enable Hold Time
0.059
—
0.059
—
0.071
—
ns
PIC Timing
PIO Input/Output Buffer Timing
tIN_PIO
Input Buffer Delay (LVCMOS25)
—
0.423
—
0.466
—
0.508
ns
tOUT_PIO
Output Buffer Delay (LVCMOS25)
—
1.115
—
1.155
—
1.196
ns
—
1.124
—
1.293
—
ns
IOLOGIC Input/Output Timing
tSUI_PIO
Input Register Setup Time (Data Before
Clock)
0.956
tHI_PIO
Input Register Hold Time (Data after Clock)
0.313
—
0.395
—
0.378
—
ns
tCOO_PIO
Output Register Clock to Output Delay4
—
1.455
—
1.564
—
1.674
ns
tSUCE_PIO
Input Register Clock Enable Setup Time
0.220
—
0.185
—
0.150
—
ns
tHCE_PIO
Input Register Clock Enable Hold Time
-0.085
—
-0.072
—
-0.058
—
ns
tSULSR_PIO
Set/Reset Setup Time
0.117
—
0.103
—
0.088
—
ns
tHLSR_PIO
Set/Reset Hold Time
-0.107
—
-0.094
—
-0.081
—
ns
EBR Timing
tCO_EBR
Clock (Read) to output from Address or Data
—
2.78
—
2.89
—
2.99
ns
tCOO_EBR
Clock (Write) to output from EBR output Register
—
0.31
—
0.32
—
0.33
ns
tSUDATA_EBR
Setup Data to EBR Memory
-0.218
—
-0.227
—
-0.237
—
ns
tHDATA_EBR
Hold Data to EBR Memory
0.249
—
0.257
—
0.265
—
ns
tSUADDR_EBR
Setup Address to EBR Memroy
-0.071
—
-0.070
—
-0.068
—
ns
tHADDR_EBR
Hold Address to EBR Memory
0.118
—
0.098
—
0.077
—
ns
tSUWREN_EBR
Setup Write/Read Enable to PFU Memory
-0.107
—
-0.106
—
-0.106
—
ns
3-24
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 Internal Switching Characteristics1, 2 (Continued)
Over Recommended Commercial Operating Conditions
-8
Description
-6
Min.
Max.
Min.
Max.
Min.
Max.
Units.
tHWREN_EBR
Hold Write/Read Enable to PFU Memory
0.141
—
0.145
—
0.149
—
ns
tSUCE_EBR
Clock Enable Setup Time to EBR Output
Register
0.087
0.096
0.104
ns
tHCE_EBR
Clock Enable Hold Time to EBR Output
Register
-0.066
-0.080
-0.094
ns
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Parameter
-7
tSUBE_EBR
Byte Enable Set-Up Time to EBR Output
Register
-0.071
-0.070
-0.068
ns
tHBE_EBR
Byte Enable Hold Time to EBR Output
Register
0.118
0.098
0.077
ns
DSP Block Timing3
tSUI_DSP
Input Register Setup Time
0.32
—
0.36
—
0.39
—
ns
tHI_DSP
Input Register Hold Time
-0.17
—
-0.19
—
-0.21
—
ns
tSUP_DSP
Pipeline Register Setup Time
2.23
—
2.30
—
2.37
—
ns
tHP_DSP
Pipeline Register Hold Time
-1.02
—
-1.09
—
-1.15
—
ns
tSUO_DSP
Output Register Setup Time
3.09
—
3.22
—
3.34
—
ns
-1.67
—
-1.76
—
-1.84
—
ns
tHO_DSP
Output Register Hold Time
tCOI_DSP
Input Register Clock to Output Time
—
3.68
—
4.03
—
4.38
ns
tCOP_DSP
Pipeline Register Clock to Output Time
—
1.30
—
1.47
—
1.64
ns
tCOO_DSP
Output Register Clock to Output Time
—
0.58
—
0.60
—
0.62
ns
tSUOPT_DSP
Opcode Register Setup Time
0.31
—
0.35
—
0.39
—
ns
tHOPT_DSP
Opcode Register Hold Time
-0.20
—
-0.24
—
-0.27
—
ns
tSUDATA_DSP
Cascade_data through ALU to Output
Register Setup Time
1.55
—
1.67
—
1.78
—
ns
tHPDATA_DSP
Cascade_data through ALU to Output
Register Hold Time
-0.44
—
-0.53
—
-0.61
—
ns
1.
2.
3.
4.
Internal parameters are characterized but not tested on every device.
Commercial timing numbers are shown. Industrial timing numbers are typically slower and can be extracted from the ispLEVER software.
DSP slice is configured in Multiply Add/Sub 18x18 mode.
The output register is in Flip-flop mode.
3-25
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Timing Diagrams
Figure 3-9. Read/Write Mode (Normal)
CLKA
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
CSA
WEA
ADA
A0
A1
A0
A0
A1
tSU tH
DIA
D0
D1
tCO_EBR
tCO_EBR
D0
DOA
tCO_EBR
D1
D0
Note: Input data and address are registered at the positive edge of the clock and output data appears after the positive edge of the clock.
Figure 3-10. Read/Write Mode with Input and Output Registers
CLKA
CSA
WEA
ADA
A0
tSU
DIA
A1
A0
A1
A0
tH
D0
D1
tCOO_EBR
DOA (Regs)
Mem(n) data from previous read
output is only updated during a read cycle
3-26
tCOO_EBR
D0
D1
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 3-11. Write Through (SP Read/Write on Port A, Input Registers Only)
CLKA
CSA
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
WEA
Three consecutive writes to A0
ADA
A0
tSU
DIA
A1
tH
D0
D2
D1
tACCESS
DOA
A0
Data from Prev Read
or Write
tACCESS
D0
D3
D4
tACCESS
D1
tACCESS
D2
D3
D4
Note: Input data and address are registered at the positive edge of the clock and output data appears after the positive edge of the clock.
3-27
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 Family Timing Adders1, 2, 3, 4, 5
Over Recommended Commercial Operating Conditions
Buffer Type
Description
-8
-7
-6
Units
LVDS, Emulated, VCCIO = 2.5V
0.03
-0.01
-0.03
ns
LVDS25
LVDS, VCCIO = 2.5V
0.03
0.00
-0.04
ns
BLVDS25
BLVDS, Emulated, VCCIO = 2.5V
0.03
0.00
-0.04
ns
MLVDS25
MLVDS, Emulated, VCCIO = 2.5V
0.03
0.00
-0.04
ns
Input Adjusters
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
LVDS25E
RSDS25
RSDS, VCCIO = 2.5V
0.03
0.00
-0.03
ns
PPLVDS
Point-to-Point LVDS
0.03
-0.01
-0.03
ns
TRLVDS
Transition-Reduced LVDS
0.03
0.00
-0.04
ns
Mini MLVDS
Mini LVDS
0.03
-0.01
-0.03
ns
LVPECL33
LVPECL, Emulated, VCCIO = 3.0V
0.17
0.07
-0.04
ns
HSTL18_I
HSTL_18 class I, VCCIO = 1.8V
0.20
0.17
0.13
ns
HSTL18_II
HSTL_18 class II, VCCIO = 1.8V
0.20
0.17
0.13
ns
HSTL18D_I
Differential HSTL 18 class I
0.20
0.17
0.13
ns
HSTL18D_II
Differential HSTL 18 class II
0.20
0.17
0.13
ns
HSTL15_I
HSTL_15 class I, VCCIO = 1.5V
0.10
0.12
0.13
ns
HSTL15D_I
Differential HSTL 15 class I
0.10
0.12
0.13
ns
SSTL33_I
SSTL_3 class I, VCCIO = 3.0V
0.17
0.23
0.28
ns
SSTL33_II
SSTL_3 class II, VCCIO = 3.0V
0.17
0.23
0.28
ns
SSTL33D_I
Differential SSTL_3 class I
0.17
0.23
0.28
ns
SSTL33D_II
Differential SSTL_3 class II
0.17
0.23
0.28
ns
SSTL25_I
SSTL_2 class I, VCCIO = 2.5V
0.12
0.14
0.16
ns
SSTL25_II
SSTL_2 class II, VCCIO = 2.5V
0.12
0.14
0.16
ns
SSTL25D_I
Differential SSTL_2 class I
0.12
0.14
0.16
ns
SSTL25D_II
Differential SSTL_2 class II
0.12
0.14
0.16
ns
SSTL18_I
SSTL_18 class I, VCCIO = 1.8V
0.08
0.06
0.04
ns
SSTL18_II
SSTL_18 class II, VCCIO = 1.8V
0.08
0.06
0.04
ns
SSTL18D_I
Differential SSTL_18 class I
0.08
0.06
0.04
ns
SSTL18D_II
Differential SSTL_18 class II
0.08
0.06
0.04
ns
SSTL15
SSTL_15, VCCIO = 1.5V
0.087
0.059
0.032
ns
SSTL15D
Differential SSTL_15
0.087
0.025
-0.036
ns
LVTTL33
LVTTL, VCCIO = 3.0V
0.05
0.05
0.05
ns
LVCMOS33
LVCMOS, VCCIO = 3.0V
0.05
0.05
0.05
ns
LVCMOS25
LVCMOS, VCCIO = 2.5V
0.00
0.00
0.00
ns
LVCMOS18
LVCMOS, VCCIO = 1.8V
0.06
0.08
0.11
ns
LVCMOS15
LVCMOS, VCCIO = 1.5V
0.17
0.21
0.25
ns
LVCMOS12
LVCMOS, VCCIO = 1.2V
0.01
0.05
0.08
ns
PCI33
PCI, VCCIO = 3.0V
0.05
0.05
0.05
ns
LVDS25E
LVDS, Emulated, VCCIO = 2.5V
0.15
0.15
0.16
ns
LVDS25
LVDS, VCCIO = 2.5V
0.02
0.08
0.13
ns
BLVDS25
BLVDS, Emulated, VCCIO = 2.5V
0.00
-0.02
-0.04
ns
MLVDS25
MLVDS, Emulated, VCCIO = 2.5V
0.00
-0.01
-0.03
ns
Output Adjusters
3-28
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 Family Timing Adders1, 2, 3, 4, 5 (Continued)
Over Recommended Commercial Operating Conditions
Buffer Type
Description
-8
-7
-6
Units
0.05
0.10
0.16
ns
Point-to-Point LVDS, Emulated, VCCIO = 2.5V
-0.10
-0.05
0.01
ns
LVPECL, Emulated, VCCIO = 3.0V
-0.02
-0.04
-0.06
ns
RSDS25
RSDS, VCCIO = 2.5V
PPLVDS
LVPECL33
HSTL_18 class I 8mA drive, VCCIO = 1.8V
-0.19
-0.16
-0.12
ns
HSTL18_II
HSTL_18 class II, VCCIO = 1.8V
-0.30
-0.28
-0.25
ns
HSTL18D_I
Differential HSTL 18 class I 8mA drive
-0.19
-0.16
-0.12
ns
HSTL18D_II
Differential HSTL 18 class II
-0.30
-0.28
-0.25
ns
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
HSTL18_I
HSTL15_I
HSTL_15 class I 4mA drive, VCCIO = 1.5V
-0.22
-0.19
-0.16
ns
HSTL15D_I
Differential HSTL 15 class I 4mA drive
-0.22
-0.19
-0.16
ns
SSTL33_I
SSTL_3 class I, VCCIO = 3.0V
0.08
0.13
0.19
ns
SSTL33_II
SSTL_3 class II, VCCIO = 3.0V
-0.20
-0.17
-0.14
ns
SSTL33D_I
Differential SSTL_3 class I
0.08
0.13
0.18
ns
SSTL33D_II
Differential SSTL_3 class II
-0.20
-0.17
-0.14
ns
SSTL25_I
SSTL_2 class I 8mA drive, VCCIO = 2.5V
-0.06
-0.02
0.02
ns
SSTL25_II
SSTL_2 class II 16mA drive, VCCIO = 2.5V
-0.19
-0.15
-0.12
ns
SSTL25D_I
Differential SSTL_2 class I 8mA drive
-0.06
-0.02
0.02
ns
SSTL25D_II
Differential SSTL_2 class II 16mA drive
-0.19
-0.15
-0.12
ns
SSTL18_I
SSTL_1.8 class I, VCCIO = 1.8V
-0.14
-0.10
-0.07
ns
SSTL18_II
SSTL_1.8 class II 8mA drive, VCCIO = 1.8V
-0.20
-0.17
-0.14
ns
SSTL18D_I
Differential SSTL_1.8 class I
-0.14
-0.10
-0.07
ns
SSTL18D_II
Differential SSTL_1.8 class II 8mA drive
-0.20
-0.17
-0.14
ns
SSTL15
SSTL_1.5, VCCIO = 1.5V
0.07
0.08
0.08
ns
SSTL15D
Differential SSTL_15
0.07
0.08
0.08
ns
LVTTL33_4mA
LVTTL 4mA drive, VCCIO = 3.0V
0.21
0.23
0.25
ns
LVTTL33_8mA
LVTTL 8mA drive, VCCIO = 3.0V
0.09
0.09
0.10
ns
LVTTL33_12mA
LVTTL 12mA drive, VCCIO = 3.0V
0.02
0.03
0.03
ns
LVTTL33_16mA
LVTTL 16mA drive, VCCIO = 3.0V
0.12
0.13
0.13
ns
LVTTL33_20mA
LVTTL 20mA drive, VCCIO = 3.0V
0.08
0.08
0.09
ns
LVCMOS33_4mA
LVCMOS 3.3 4mA drive, fast slew rate
0.21
0.23
0.25
ns
LVCMOS33_8mA
LVCMOS 3.3 8mA drive, fast slew rate
0.09
0.09
0.10
ns
LVCMOS33_12mA
LVCMOS 3.3 12mA drive, fast slew rate
0.02
0.03
0.03
ns
LVCMOS33_16mA
LVCMOS 3.3 16mA drive, fast slew rate
0.12
0.13
0.13
ns
LVCMOS33_20mA
LVCMOS 3.3 20mA drive, fast slew rate
0.08
0.08
0.09
ns
LVCMOS25_4mA
LVCMOS 2.5 4mA drive, fast slew rate
0.12
0.12
0.12
ns
LVCMOS25_8mA
LVCMOS 2.5 8mA drive, fast slew rate
0.05
0.05
0.05
ns
LVCMOS25_12mA
LVCMOS 2.5 12mA drive, fast slew rate
0.00
0.00
0.00
ns
LVCMOS25_16mA
LVCMOS 2.5 16mA drive, fast slew rate
0.08
0.08
0.08
ns
LVCMOS25_20mA
LVCMOS 2.5 20mA drive, fast slew rate
0.04
0.04
0.04
ns
LVCMOS18_4mA
LVCMOS 1.8 4mA drive, fast slew rate
0.08
0.09
0.09
ns
LVCMOS18_8mA
LVCMOS 1.8 8mA drive, fast slew rate
0.02
0.01
0.01
ns
LVCMOS18_12mA
LVCMOS 1.8 12mA drive, fast slew rate
-0.03
-0.03
-0.03
ns
LVCMOS18_16mA
LVCMOS 1.8 16mA drive, fast slew rate
0.03
0.03
0.03
ns
3-29
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 Family Timing Adders1, 2, 3, 4, 5 (Continued)
Over Recommended Commercial Operating Conditions
-8
-7
-6
Units
LVCMOS15_4mA
Buffer Type
LVCMOS 1.5 4mA drive, fast slew rate
Description
0.09
0.10
0.10
ns
LVCMOS15_8mA
LVCMOS 1.5 8mA drive, fast slew rate
0.01
0.01
0.00
ns
LVCMOS12_2mA
LVCMOS 1.2 2mA drive, fast slew rate
0.08
0.08
0.08
ns
LVCMOS12_6mA
LVCMOS 1.2 6mA drive, fast slew rate
-0.02
-0.02
-0.02
ns
LVCMOS33_4mA
LVCMOS 3.3 4mA drive, slow slew rate
1.64
1.71
1.77
ns
LVCMOS 3.3 8mA drive, slow slew rate
1.39
1.45
1.51
ns
LVCMOS33_12mA
LVCMOS 3.3 12mA drive, slow slew rate
1.21
1.27
1.33
ns
LVCMOS33_16mA
LVCMOS 3.3 16mA drive, slow slew rate
1.43
1.49
1.55
ns
LVCMOS33_20mA
LVCMOS 3.3 20mA drive, slow slew rate
1.23
1.28
1.34
ns
LVCMOS25_4mA
LVCMOS 2.5 4mA drive, slow slew rate
1.66
1.70
1.74
ns
LVCMOS25_8mA
LVCMOS 2.5 8mA drive, slow slew rate
1.39
1.43
1.46
ns
LVCMOS25_12mA
LVCMOS 2.5 12mA drive, slow slew rate
1.20
1.24
1.28
ns
LVCMOS25_16mA
LVCMOS 2.5 16mA drive, slow slew rate
1.42
1.45
1.49
ns
LVCMOS25_20mA
LVCMOS 2.5 20mA drive, slow slew rate
1.22
1.26
1.29
ns
LVCMOS18_4mA
LVCMOS 1.8 4mA drive, slow slew rate
1.61
1.65
1.68
ns
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
LVCMOS33_8mA
LVCMOS18_8mA
LVCMOS 1.8 8mA drive, slow slew rate
1.32
1.36
1.39
ns
LVCMOS18_12mA
LVCMOS 1.8 12mA drive, slow slew rate
1.14
1.17
1.21
ns
LVCMOS18_16mA
LVCMOS 1.8 16mA drive, slow slew rate
1.35
1.38
1.42
ns
LVCMOS15_4mA
LVCMOS 1.5 4mA drive, slow slew rate
1.57
1.60
1.64
ns
LVCMOS15_8mA
LVCMOS 1.5 8mA drive, slow slew rate
0.01
0.01
0.00
ns
LVCMOS12_2mA
LVCMOS 1.2 2mA drive, slow slew rate
1.51
1.54
1.58
ns
LVCMOS12_6mA
LVCMOS 1.2 6mA drive, slow slew rate
-0.02
-0.02
-0.02
ns
PCI33
PCI, VCCIO = 3.0V
0.19
0.21
0.24
ns
1.
2.
3.
4.
5.
Timing adders are characterized but not tested on every device.
LVCMOS timing measured with the load specified in Switching Test Condition table.
All other standards tested according to the appropriate specifications.
Not all I/O standards and drive strengths are supported for all banks. See the Architecture section of this data sheet for details.
Commercial timing numbers are shown. Industrial numbers are typically slower and can be extracted from the ispLEVER software.
3-30
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 Maximum I/O Buffer Speed 1, 2, 3, 4, 5, 6
Over Recommended Operating Conditions
Buffer
Description
Max.
Units
400
MHz
Maximum Input Frequency
LVDS, VCCIO = 2.5V
MLVDS, Emulated, VCCIO = 2.5V
400
MHz
BLVDS25
BLVDS, Emulated, VCCIO = 2.5V
400
MHz
PPLVDS
Point-to-Point LVDS
400
MHz
TRLVDS
Transition-Reduced LVDS
612
MHz
Mini LVDS
Mini LVDS
400
MHz
400
MHz
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
LVDS25
MLVDS25
LVPECL33
LVPECL, Emulated, VCCIO = 3.0V
HSTL18 (all supported classed)
HSTL_18 class I, II, VCCIO = 1.8V
400
MHz
HSTL15
HSTL_15 class I, VCCIO = 1.5V
400
MHz
SSTL33 (all supported classed)
SSTL_3 class I, II, VCCIO = 3.0V
400
MHz
SSTL25 (all supported classed)
SSTL_2 class I, II, VCCIO = 2.5V
400
MHz
SSTL18 (all supported classed)
SSTL_18 class I, II, VCCIO = 1.8V
400
MHz
LVTTL33
LVTTL, VCCIO = 3.0V
166
MHz
LVCMOS33
LVCMOS, VCCIO = 3.0V
166
MHz
LVCMOS25
LVCMOS, VCCIO = 2.5V
166
MHz
LVCMOS18
LVCMOS, VCCIO = 1.8V
166
MHz
LVCMOS15
LVCMOS 1.5, VCCIO = 1.5V
166
MHz
LVCMOS12
LVCMOS 1.2, VCCIO = 1.2V
166
MHz
PCI33
PCI, VCCIO = 3.3V
66
MHz
LVDS25E
LVDS, Emulated, VCCIO = 2.5V
300
MHz
LVDS25
LVDS, VCCIO = 2.5V
612
MHz
MLVDS25
MLVDS, Emulated, VCCIO = 2.5V
300
MHz
RSDS25
RSDS, Emulated, VCCIO = 2.5V
612
MHz
Maximum Output Frequency
BLVDS25
BLVDS, Emulated, VCCIO = 2.5V
300
MHz
PPLVDS
Point-to-point LVDS
612
MHz
LVPECL33
LVPECL, Emulated, VCCIO = 3.0V
612
MHz
Mini-LVDS
Mini LVDS
612
MHz
HSTL18 (all supported classed)
HSTL_18 class I, II, VCCIO = 1.8V
200
MHz
HSTL15 (all supported classed)
HSTL_15 class I, VCCIO = 1.5V
200
MHz
SSTL33 (all supported classed)
SSTL_3 class I, II, VCCIO = 3.0V
233
MHz
SSTL25 (all supported classed)
SSTL_2 class I, II, VCCIO = 2.5V
233
MHz
SSTL18 (all supported classed)
SSTL_18 class I, II, VCCIO = 1.8V
266
MHz
LVTTL33
LVTTL, VCCIO = 3.0V
166
MHz
LVCMOS33 (For all drives)
LVCMOS, 3.3V
166
MHz
LVCMOS25 (For all drives)
LVCMOS, 2.5V
166
MHz
LVCMOS18 (For all drives)
LVCMOS, 1.8V
166
MHz
LVCMOS15 (For all drives)
LVCMOS, 1.5V
166
MHz
LVCMOS12 (For all drives except 2mA)
LVCMOS, VCCIO = 1.2V
166
MHz
LVCMOS12 (2mA drive)
LVCMOS, VCCIO = 1.2V
100
MHz
3-31
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 Maximum I/O Buffer Speed (Continued)1, 2, 3, 4, 5, 6
Over Recommended Operating Conditions
Buffer
PCI33
PCI, VCCIO = 3.3V
Max.
Units
66
MHz
These maximum speeds are characterized but not tested on every device.
Maximum I/O speed for differential output standards emulated with resistors depends on the layout.
LVCMOS timing is measured with the load specified in the Switching Test Conditions table of this document.
All speeds are measured at fast slew.
Actual system operation may vary depending on user logic implementation.
Maximum data rate equals 2 times the clock rate when utilizing DDR.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
1.
2.
3.
4.
5.
6.
Description
3-32
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
sysCLOCK PLL Timing
Over Recommended Operating Conditions
Parameter
Descriptions
Conditions
Clock
Min.
Typ.
Max.
Units
2
—
500
MHz
—
420
MHz
fIN
Input clock frequency (CLKI,
CLKFB)
Edge clock
Primary clock
2
fOUT
Output clock frequency (CLKOP,
CLKOS)
Edge clock
4
—
500
MHz
Primary clock
4
—
420
MHz
MHz
K-Divider output frequency
CLKOK
0.03125
—
250
K2-Divider output frequency
CLKOK2
0.667
—
166
MHz
fVCO
PLL VCO frequency
500
—
1000
MHz
fPFD3
Phase detector input frequency
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
fOUT1
fOUT2
Edge clock
2
—
500
MHz
Primary clock
2
—
420
MHz
ps
AC Characteristics
tPA
tDT
Programmable delay unit
Output clock duty cycle
(CLKOS, at 50% setting)
65
130
260
Edge clock
45
50
55
%
fOUT 250 MHz
Primary clock
45
50
55
%
fOUT > 250MHz
Primary clock
30
50
70
%
tCPA
Coarse phase shift error
(CLKOS, at all settings)
-5
0
+5
% of
period
tOPW
Output clock pulse width high or
low
(CLKOS)
1.8
—
—
ns
—
—
200
p-p
fOUT 420MHz
tOPJIT1
Output clock period jitter
tSK
Input clock to output clock skew
when N/M = integer
tLOCK2
Lock time
tUNLOCK
Reset to PLL unlock time to
ensure fast reset
tHI
Input clock high time
tLO
Input clock low time
tIPJIT
tRST
420MHz > fOUT 100MHz
—
—
250
p-p
fOUT < 100MHz
—
—
0.025
UIPP
—
—
500
p-p
2 to 25 MHz
—
—
200
us
25 to 500 MHz
—
—
50
us
—
—
50
ns
90% to 90%
0.5
—
—
ns
10% to 10%
0.5
—
—
ns
Input clock period jitter
—
—
400
p-p
Reset signal pulse width high,
RESETM,
RESETK
10
—
—
ns
—
—
ns
Reset signal pulse width high,
CNTRST
500
1. Jitter sample is taken over 10,000 samples of the primary PLL output with clean reference clock with no additional I/O toggling.
2. Output clock is valid after tLOCK for PLL reset and dynamic delay adjustment.
3. Period jitter and cycle-to-cycle jitter numbers are guaranteed for fPFD > 4MHz. For fPFD < 4MHz, the jitter numbers may not be met in certain conditions. Please contact the factory for fPFD < 4MHz.
3-33
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
DLL Timing
Over Recommended Operating Conditions
Parameter
fREF
fFB
Min.
Typ.
Max.
Units
Input reference clock frequency (on-chip or
off-chip)
Description
Condition
133
—
500
MHz
Feedback clock frequency (on-chip or off-chip)
133
—
500
MHz
1
fCLKOP
Output clock frequency, CLKOP
133
—
500
MHz
fCLKOS2
Output clock frequency, CLKOS
33.3
tPJIT
Output clock period jitter (clean input)
tDUTY
Output clock duty cycle (at 50% levels, 50% duty Edge Clock
cycle input clock, 50% duty cycle circuit turned
Primary Clock
off, time reference delay mode)
tDUTYTRD
Output clock duty cycle (at 50% levels, arbitrary
duty cycle input clock, 50% duty cycle circuit
enabled, time reference delay mode)
500
MHz
200
ps p-p
40
60
%
30
70
%
Primary Clock < 250MHz
45
55
%
Primary Clock 250MHz
30
70
%
Edge Clock
45
55
%
40
60
%
30
70
%
45
55
%
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
—
—
tDUTYCIR
Output clock duty cycle (at 50% levels, arbitrary Primary Clock < 250MHz
duty cycle input clock, 50% duty cycle circuit
Primary Clock 250MHz
enabled, clock injection removal mode) with DLL
Edge Clock
cascading
tSKEW3
Output clock to clock skew between two outputs
with the same phase setting
—
—
100
ps
tPHASE
Phase error measured at device pads between
off-chip reference clock and feedback clocks
—
—
+/-400
ps
tPWH
Input clock minimum pulse width high (at 80%
level)
550
—
—
ps
tPWL
Input clock minimum pulse width low (at 20%
level)
550
—
—
ps
tINSTB
Input clock period jitter
—
—
500
p-p
tLOCK
DLL lock time
8
—
8200
cycles
tRSWD
Digital reset minimum pulse width (at 80% level)
3
—
—
ns
tDEL
Delay step size
27
45
70
ps
tRANGE1
Max. delay setting for single delay block
(64 taps)
1.9
3.1
4.4
ns
tRANGE4
Max. delay setting for four chained delay blocks
7.6
12.4
17.6
ns
1. CLKOP runs at the same frequency as the input clock.
2. CLKOS minimum frequency is obtained with divide by 4.
3. This is intended to be a “path-matching” design guideline and is not a measurable specification.
3-34
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
SERDES High-Speed Data Transmitter1
Table 3-6. Serial Output Timing and Levels
Symbol
Description
Frequency
Min.
Typ.
Max.
Units
0.25 to 3.125 Gbps
1150
1440
1730
mV, p-p
0.25 to 3.125 Gbps
1080
1350
1620
mV, p-p
0.25 to 3.125 Gbps
1000
1260
1510
mV, p-p
1, 2
0.25 to 3.125 Gbps
840
1130
1420
mV, p-p
VTX-DIFF-P-P-1.04 Differential swing (1.04V setting)1, 2
0.25 to 3.125 Gbps
780
1040
1300
VTX-DIFF-P-P-1.44 Differential swing (1.44V setting)
1, 2
VTX-DIFF-P-P-1.35 Differential swing (1.35V setting)
1, 2
VTX-DIFF-P-P-1.26 Differential swing (1.26V setting)1, 2
mV, p-p
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
VTX-DIFF-P-P-1.13 Differential swing (1.13V setting)
VTX-DIFF-P-P-0.92 Differential swing (0.92V setting)1, 2
0.25 to 3.125 Gbps
690
920
1150
mV, p-p
VTX-DIFF-P-P-0.87 Differential swing (0.87V setting)1, 2
0.25 to 3.125 Gbps
650
870
1090
mV, p-p
VTX-DIFF-P-P-0.78 Differential swing (0.78V setting)1, 2
0.25 to 3.125 Gbps
585
780
975
mV, p-p
VTX-DIFF-P-P-0.64 Differential swing (0.64V setting)1, 2
0.25 to 3.125 Gbps
480
640
800
mV, p-p
VOCM
Output common mode voltage
—
VCCOB
-0.75
VCCOB
-0.60
VCCOB
-0.45
V
TTX-R
Rise time (20% to 80%)
—
145
185
265
ps
TTX-F
Fall time (80% to 20%)
—
145
185
265
ps
ZTX-OI-SE
Output Impedance 50/75/HiZ Ohms
(single ended)
—
-20%
50/75/
Hi Z
+20%
Ohms
RLTX-RL
Return loss (with package)
—
10
TTX-INTRASKEW
Lane-to-lane TX skew within a
SERDES quad block (intra-quad)
—
—
—
200
ps
TTX-INTERSKEW3
Lane-to-lane skew between SERDES
quad blocks (inter-quad)
—
—
—
1UI +200
ps
dB
1. All measurements are with 50 ohm impedance.
2. See TN1176, LatticeECP3 SERDES/PCS Usage Guide for actual binary settings and the min-max range.
3. Inter-quad skew is between all SERDES channels on the device and requires the use of a low skew internal reference clock.
Table 3-7. Channel Output Jitter
Min.
Typ.
Max.
Units
Deterministic
Description
3.125 Gbps
Frequency
—
—
0.17
UI, p-p
Random
3.125 Gbps
—
—
0.25
UI, p-p
Total
3.125 Gbps
—
—
0.35
UI, p-p
Deterministic
2.5Gbps
—
—
0.17
UI, p-p
Random
2.5Gbps
—
—
0.20
UI, p-p
Total
2.5Gbps
—
—
0.35
UI, p-p
Deterministic
1.25 Gbps
—
—
0.10
UI, p-p
Random
1.25 Gbps
—
—
0.22
UI, p-p
Total
1.25 Gbps
—
—
0.24
UI, p-p
Deterministic
622 Mbps
—
—
0.10
UI, p-p
Random
622 Mbps
—
—
0.20
UI, p-p
Total
622 Mbps
—
—
0.24
UI, p-p
Deterministic
250 Mbps
—
—
0.10
UI, p-p
Random
250 Mbps
—
—
0.18
UI, p-p
Total
250 Mbps
—
—
0.24
UI, p-p
Note: Values are measured with PRBS 27-1, all channels operating, FPGA logic active, I/Os around SERDES pins quiet,
reference clock @ 10X mode.
3-35
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
SERDES/PCS Block Latency
Table 3-8 describes the latency of each functional block in the transmitter and receiver. Latency is given in parallel
clock cycles. Figure 3-12 shows the location of each block.
Table 3-8. SERDES/PCS Latency Breakdown
Item
Description
Min.
Avg.
Max.
Fixed
Bypass
Units
Transmit Data Latency1
FPGA Bridge - Gearing disabled with different clocks
1
3
5
—
1
word clk
FPGA Bridge - Gearing disabled with same clocks
—
—
—
3
1
word clk
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
T1
T2
T3
T4
T5
FPGA Bridge - Gearing enabled
1
3
5
—
—
word clk
8b10b Encoder
—
—
—
2
1
word clk
SERDES Bridge transmit
—
—
—
2
1
word clk
Serializer: 8-bit mode
—
—
—
15 + 1
—
UI + ps
Serializer: 10-bit mode
—
—
—
18 + 1
—
UI + ps
Pre-emphasis ON
—
—
—
1 + 2
—
UI + ps
Pre-emphasis OFF
—
—
—
0 + 3
—
UI + ps
Equalization ON
—
—
—
1
—
UI + ps
Equalization OFF
—
—
—
2
—
UI + ps
Receive Data Latency2
R1
R2
Deserializer: 8-bit mode
—
—
—
10 + 3
—
UI + ps
Deserializer: 10-bit mode
—
—
—
12 + 3
—
UI + ps
R3
SERDES Bridge receive
—
—
—
2
—
word clk
R4
Word alignment
3.1
—
4
—
—
word clk
R5
8b10b decoder
—
—
—
1
—
word clk
R6
Clock Tolerance Compensation
7
15
23
1
1
word clk
R7
FPGA Bridge - Gearing disabled with different clocks
1
3
5
—
1
word clk
FPGA Bridge - Gearing disabled with same clocks
—
—
—
3
1
word clk
FPGA Bridge - Gearing enabled
1
3
5
—
—
word clk
1. 1 = -245ps, 2 = +88ps, 3 = +112ps.
2. 1 = +118ps, 2 = +132ps, 3 = +700ps.
Figure 3-12. Transmitter and Receiver Latency Block Diagram
SERDES
SERDES Bridge
PCS
R1
EQ
Deserializer
1:8/1:10
CDR
HDINNi
REFCLK
DEC
Polarity
Adjust
BYPASS
Receiver
R4
R3
R2
FPGA Core
FPGA
EBRD Clock
WA
HDINPi
FPGA Bridge
Recovered Clock
REFCLK
R5
Elastic
Buffer
FIFO
BYPASS
R6
Down
Sample
FIFO
Receive Data
BYPASS
BYPASS
FPGA
Receive Clock
Transmit Clock
TX PLL
T2
T1
FPGA
Transmit Clock
T3
T4
HDOUTPi
Serializer
8:1/10:1
HDOUTNi
Transmitter
Encoder
Polarity
Adjust
Up
Sample
FIFO
BYPASS
BYPASS
BYPASS
3-36
Transmit Data
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
SERDES High Speed Data Receiver
Table 3-9. Serial Input Data Specifications
Symbol
Description
Stream of nontransitions
(CID = Consecutive Identical Digits) @ 10-12 BER
1
Typ.
Max.
—
—
136
2.5G
—
—
144
1.485G
—
—
160
622M
—
—
204
270M
—
—
228
Units
Bits
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
RX-CIDS
Min.
3.125G
155M
—
—
150
—
1760
mV, p-p
0
—
VCCA +0.54
V
—
VCCA
V
—
VCCA +0.2
V
1000
—
Bits
-20%
50/75/HiZ
+20%
Ohms
10
—
—
dB
VRX-DIFF-S
Differential input sensitivity
VRX-IN
Input levels
VRX-CM-DC
Input common mode range (DC coupled)
0.6
VRX-CM-AC
Input common mode range (AC coupled)3
0.1
—
TRX-RELOCK SCDR re-lock time2
ZRX-TERM
Input termination 50/75 Ohm/High Z
RLRX-RL
Return loss (without package)
296
1. This is the number of bits allowed without a transition on the incoming data stream when using DC coupling.
2. This is the typical number of bit times to re-lock to a new phase or frequency within +/- 300 ppm, assuming 8b10b encoded data.
3. AC coupling is used to interface to LVPECL and LVDS. LVDS interfaces are found in laser drivers and Fibre Channel equipment. LVDS interfaces are generally found in 622 Mbps SERDES devices.
4. Up to 1.76V.
Input Data Jitter Tolerance
A receiver’s ability to tolerate incoming signal jitter is very dependent on jitter type. High speed serial interface standards have recognized the dependency on jitter type and have specifications to indicate tolerance levels for different jitter types as they relate to specific protocols. Sinusoidal jitter is considered to be a worst case jitter type.
Table 3-10. Receiver Total Jitter Tolerance Specification
Description
Frequency
Deterministic
Random
Condition
600 mV differential eye
3.125 Gbps
Min.
Typ.
Max.
Units
—
—
0.47
UI, p-p
600 mV differential eye
—
—
0.18
UI, p-p
600 mV differential eye
—
—
0.65
UI, p-p
600 mV differential eye
—
—
0.47
UI, p-p
600 mV differential eye
—
—
0.18
UI, p-p
Total
600 mV differential eye
—
—
0.65
UI, p-p
Deterministic
600 mV differential eye
—
—
0.47
UI, p-p
Total
Deterministic
Random
Random
2.5 Gbps
1.25 Gbps
Total
Deterministic
Random
Total
622 Mbps
600 mV differential eye
—
—
0.18
UI, p-p
600 mV differential eye
—
—
0.65
UI, p-p
600 mV differential eye
—
—
0.47
UI, p-p
600 mV differential eye
—
—
0.18
UI, p-p
600 mV differential eye
—
—
0.65
UI, p-p
Note: Values are measured with CJPAT, all channels operating, FPGA Logic active, I/Os around SERDES pins quiet, voltages are nominal,
room temperature.
3-37
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Table 3-11. Periodic Receiver Jitter Tolerance Specification
Description
Frequency
Condition
Min.
Typ.
Max.
Units
Periodic
2.97 Gbps
600 mV differential eye
—
—
0.24
UI, p-p
Periodic
2.5 Gbps
600 mV differential eye
—
—
0.22
UI, p-p
Periodic
1.485 Gbps
600 mV differential eye
—
—
0.24
UI, p-p
Periodic
622 Mbps
600 mV differential eye
—
—
0.15
UI, p-p
Periodic
155 Mbps
600 mV differential eye
—
—
0.5
UI, p-p
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Note: Values are measured with PRBS 27-1, all channels operating, FPGA Logic active, I/Os around SERDES pins
quiet, voltages are nominal, room temperature.
3-38
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
SERDES External Reference Clock
The external reference clock selection and its interface are a critical part of system applications for this product.
Table 3-12 specifies reference clock requirements, over the full range of operating conditions.
Table 3-12. External Reference Clock Specification (refclkp/refclkn)
Symbol
FREF
Description
Frequency range
4
Frequency tolerance
VREF-IN-SE
Input swing, single-ended clock1
Typ.
Max.
Units
15
—
320
MHz
-1000
—
1000
ppm
200
—
VCCA
mV, p-p
200
—
2*VCCA
mV, p-p
differential
0
—
VCCA + 0.3
V
0.125
—
VCCA
V
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
FREF-PPM
Min.
VREF-IN-DIFF
Input swing, differential clock
VREF-IN
Input levels
VREF-CM-AC
Input common mode range (AC coupled)2
3
DREF
Duty cycle
40
—
60
%
TREF-R
Rise time (20% to 80%)
200
500
1000
ps
TREF-F
Fall time (80% to 20%)
ZREF-IN-TERM-DIFF Differential input termination
CREF-IN-CAP
Input capacitance
200
500
1000
ps
-20%
100/2K
+20%
Ohms
—
—
7
pF
1. The signal swing for a single-ended input clock must be as large as the p-p differential swing of a differential input clock to get the same gain
at the input receiver. Lower swings for the clock may be possible, but will tend to increase jitter.
2. When AC coupled, the input common mode range is determined by:
(Min input level) + (Peak-to-peak input swing)/2 (Input common mode voltage) (Max input level) - (Peak-to-peak input swing)/2
3. Measured at 50% amplitude.
4. Depending on the application, the PLL_LOL_SET and CDR_LOL_SET control registers may be adjusted for other tolerance values as
described in TN1176, LatticeECP3 SERDES/PCS Usage Guide.
Figure 3-13. SERDES External Reference Clock Waveforms
3-39
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
PCI Express Electrical and Timing Characteristics
AC and DC Characteristics
Over Recommended Operating Conditions
Symbol
Description
Test Conditions
Min
Typ
Max
Units
399.88
400
400.12
ps
1
Transmit
Unit interval
VTX-DIFF_P-P
Differential peak-to-peak output voltage
0.8
1.0
1.2
V
VTX-DE-RATIO
De-emphasis differential output voltage
ratio
-3
-3.5
-4
dB
VTX-CM-AC_P
RMS AC peak common-mode output
voltage
—
—
20
mV
VTX-RCV-DETECT
Amount of voltage change allowed during receiver detection
—
—
600
mV
VTX-DC-CM
Tx DC common mode voltage
0
—
VCCOB + 5%
V
ITX-SHORT
Output short circuit current
—
—
90
mA
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
UI
VTX-D+=0.0V
VTX-D-=0.0V
ZTX-DIFF-DC
Differential output impedance
80
100
120
Ohms
RLTX-DIFF
Differential return loss
10
—
—
dB
RLTX-CM
Common mode return loss
6.0
—
—
dB
TTX-RISE
Tx output rise time
20 to 80%
0.125
—
—
UI
TTX-FALL
Tx output fall time
20 to 80%
0.125
—
—
UI
LTX-SKEW
Lane-to-lane static output skew for all
lanes in port/link
—
—
1.3
ns
TTX-EYE
Transmitter eye width
0.75
—
—
UI
TTX-EYE-MEDIAN-TO-MAX-JITTER
Maximum time between jitter median
and maximum deviation from median
—
—
0.125
UI
400
400.12
ps
Receive1, 2
UI
Unit Interval
399.88
3
VRX-DIFF_P-P
Differential peak-to-peak input voltage
0.34
—
1.2
V
VRX-IDLE-DET-DIFF_P-P
Idle detect threshold voltage
65
—
3403
mV
VRX-CM-AC_P
Receiver common mode voltage for AC
coupling
—
—
150
mV
ZRX-DIFF-DC
DC differential input impedance
80
100
120
Ohms
ZRX-DC
DC input impedance
40
50
60
Ohms
ZRX-HIGH-IMP-DC
Power-down DC input impedance
200K
—
—
Ohms
RLRX-DIFF
Differential return loss
10
—
—
dB
RLRX-CM
Common mode return loss
6.0
—
—
dB
TRX-IDLE-DET-DIFF-ENTERTIME
Maximum time required for receiver to
recognize and signal an unexpected idle
on link
—
—
—
ms
1. Values are measured at 2.5 Gbps.
2. Measured with external AC-coupling on the receiver.
3.Not in compliance with PCI Express 1.1 standard.
3-40
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
XAUI/Serial Rapid I/O Type 3 Electrical and Timing Characteristics
AC and DC Characteristics
Table 3-13. Transmit
Over Recommended Operating Conditions
Symbol
Description
Test Conditions
Min.
Typ.
Max.
Units
20%-80%
—
80
—
ps
Differential impedance
80
100
120
Ohms
Output data deterministic jitter
—
—
0.17
UI
Total output data jitter
—
—
0.35
UI
Differential rise/fall time
ZTX_DIFF_DC
JTX_DDJ2, 3, 4
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
TRF
JTX_TJ
1.
2.
3.
4.
1, 2, 3, 4
Total jitter includes both deterministic jitter and random jitter.
Jitter values are measured with each CML output AC coupled into a 50-ohm impedance (100-ohm differential impedance).
Jitter and skew are specified between differential crossings of the 50% threshold of the reference signal.
Values are measured at 2.5 Gbps.
Table 3-14. Receive and Jitter Tolerance
Over Recommended Operating Conditions
Symbol
Description
Test Conditions
Min.
Typ.
Max.
Units
RLRX_DIFF
Differential return loss
From 100 MHz
to 3.125 GHz
10
—
—
dB
RLRX_CM
Common mode return loss
From 100 MHz
to 3.125 GHz
6
—
—
dB
ZRX_DIFF
Differential termination resistance
80
100
120
Ohms
JRX_DJ1, 2, 3
Deterministic jitter tolerance (peak-to-peak)
—
—
0.37
UI
1, 2, 3
JRX_RJ
Random jitter tolerance (peak-to-peak)
—
—
0.18
UI
JRX_SJ1, 2, 3
Sinusoidal jitter tolerance (peak-to-peak)
—
—
0.10
UI
JRX_TJ1, 2, 3
Total jitter tolerance (peak-to-peak)
—
—
0.65
UI
TRX_EYE
Receiver eye opening
0.35
—
—
UI
1.
2.
3.
4.
5.
Total jitter includes deterministic jitter, random jitter and sinusoidal jitter. The sinusoidal jitter tolerance mask is shown in Figure 3-14.
Jitter values are measured with each high-speed input AC coupled into a 50-ohm impedance.
Jitter and skew are specified between differential crossings of the 50% threshold of the reference signal.
Jitter tolerance parameters are characterized when Full Rx Equalization is enabled.
Values are measured at 2.5 Gbps.
3-41
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
SJ Amplitude
Figure 3-14. XAUI Sinusoidal Jitter Tolerance Mask
8.5UI
20dB/dec
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
0.1UI
Data_rate/
1667
20MHz
SJ Frequency
Note: The sinusoidal jitter tolerance is measured with at least 0.37UIpp of Deterministic
jitter (Dj) and the sum of Dj and Rj (random jitter) is at least 0.55UIpp. Therefore, the
sum of Dj, Rj and Sj (sinusoidal jitter) is at least 0.65UIpp (Dj = 0.37, Rj = 0.18, Sj = 0.1).
3-42
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Serial Rapid I/O Type 2 Electrical and Timing Characteristics
AC and DC Characteristics
Table 3-15. Transmit
Symbol
Description
Test Conditions
Min.
Typ.
Max.
Units
20%-80%
—
80
—
ps
Differential impedance
80
100
120
Ohms
Output data deterministic jitter
—
—
0.17
UI
Total output data jitter
—
—
0.35
UI
TRF1
Differential rise/fall time
ZTX_DIFF_DC
JTX_DDJ3, 4, 5
1.
2.
3.
4.
5.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
JTX_TJ
2, 3, 4, 5
Rise and Fall times measured with board trace, connector and approximately 2.5pf load.
Total jitter includes both deterministic jitter and random jitter. The random jitter is the total jitter minus the actual deterministic jitter.
Jitter values are measured with each CML output AC coupled into a 50-ohm impedance (100-ohm differential impedance).
Jitter and skew are specified between differential crossings of the 50% threshold of the reference signal.
Values are measured at 2.5 Gbps.
Table 3-16. Receive and Jitter Tolerance
Symbol
Description
Test Conditions
Min.
Typ.
Max.
Units
RLRX_DIFF
Differential return loss
From 100 MHz to 2.5 GHz
10
—
—
dB
RLRX_CM
Common mode return loss
From 100 MHz to 2.5 GHz
6
—
—
dB
ZRX_DIFF
Differential termination resistance
80
100
120
Ohms
JRX_DJ2, 3, 4, 5
Deterministic jitter tolerance (peak-to-peak)
—
—
0.37
UI
2, 3, 4, 5
JRX_RJ
Random jitter tolerance (peak-to-peak)
—
—
0.18
UI
JRX_SJ2, 3, 4, 5
Sinusoidal jitter tolerance (peak-to-peak)
—
—
0.10
UI
—
—
0.65
UI
0.35
—
—
UI
JRX_TJ1, 2, 3, 4, 5 Total jitter tolerance (peak-to-peak)
TRX_EYE
1.
2.
3.
4.
5.
Receiver eye opening
Total jitter includes deterministic jitter, random jitter and sinusoidal jitter. The sinusoidal jitter tolerance mask is shown in Figure 3-14.
Jitter values are measured with each high-speed input AC coupled into a 50-ohm impedance.
Jitter and skew are specified between differential crossings of the 50% threshold of the reference signal.
Jitter tolerance, Differential Input Sensitivity and Receiver Eye Opening parameters are characterized when Full Rx Equalization is enabled.
Values are measured at 2.5 Gbps.
3-43
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Gigabit Ethernet/Serial Rapid I/O Type 1/SGMII Electrical and Timing
Characteristics
AC and DC Characteristics
Table 3-17. Transmit
Symbol
Description
Test Conditions
Min.
Typ.
Max.
Units
TRF
Differential rise/fall time
—
80
—
ps
ZTX_DIFF_DC
Differential impedance
80
100
120
Ohms
JTX_DDJ3, 4, 5
Output data deterministic jitter
—
—
0.10
UI
JTX_TJ2, 3, 4, 5
Total output data jitter
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
1.
2.
3.
4.
5.
20%-80%
—
—
0.24
UI
Rise and fall times measured with board trace, connector and approximately 2.5pf load.
Total jitter includes both deterministic jitter and random jitter. The random jitter is the total jitter minus the actual deterministic jitter.
Jitter values are measured with each CML output AC coupled into a 50-ohm impedance (100-ohm differential impedance).
Jitter and skew are specified between differential crossings of the 50% threshold of the reference signal.
Values are measured at 1.25 Gbps.
Table 3-18. Receive and Jitter Tolerance
Symbol
Description
Test Conditions
RLRX_DIFF
Differential return loss
From 100 MHz to 1.25 GHz
RLRX_CM
Common mode return loss
From 100 MHz to 1.25 GHz
ZRX_DIFF
Differential termination resistance
Min.
Typ.
Max.
Units
10
—
—
dB
6
—
—
dB
80
100
120
Ohms
JRX_DJ1, 2, 3, 4, 5 Deterministic jitter tolerance (peak-to-peak)
—
—
0.34
UI
JRX_RJ1, 2, 3, 4, 5 Random jitter tolerance (peak-to-peak)
—
—
0.26
UI
1, 2, 3, 4, 5
JRX_SJ
Sinusoidal jitter tolerance (peak-to-peak)
JRX_TJ1, 2, 3, 4, 5 Total jitter tolerance (peak-to-peak)
TRX_EYE
1.
2.
3.
4.
5.
Receiver eye opening
—
—
0.11
UI
—
—
0.71
UI
0.29
—
—
UI
Total jitter includes deterministic jitter, random jitter and sinusoidal jitter. The sinusoidal jitter tolerance mask is shown in Figure 3-14.
Jitter values are measured with each high-speed input AC coupled into a 50-ohm impedance.
Jitter and skew are specified between differential crossings of the 50% threshold of the reference signal.
Jitter tolerance, Differential Input Sensitivity and Receiver Eye Opening parameters are characterized when Full Rx Equalization is enabled.
Values are measured at 1.25 Gbps.
3-44
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
SMPTE SD/HD-SDI/3G-SDI (Serial Digital Interface) Electrical and Timing
Characteristics
AC and DC Characteristics
Table 3-19. Transmit
Symbol
Description
Test Conditions
Min.
Typ.
Max.
Units
BRSDO
Serial data rate
270
—
2975
Mbps
TJALIGNMENT 2
Serial output jitter, alignment
270 Mbps
—
—
0.20
UI
2
Serial output jitter, alignment
1485 Mbps
—
—
0.20
UI
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
TJALIGNMENT
TJALIGNMENT1, 2
Serial output jitter, alignment
2970Mbps
—
—
0.30
UI
TJTIMING
Serial output jitter, timing
270 Mbps
—
—
0.20
UI
TJTIMING
Serial output jitter, timing
1485 Mbps
—
—
1.0
UI
TJTIMING
Serial output jitter, timing
2970 Mbps
—
—
2.0
UI
Notes:
1. Timing jitter is measured in accordance with SMPTE RP 184-1996, SMPTE RP 192-1996 and the applicable serial data transmission standard, SMPTE 259M-1997 or SMPTE 292M (proposed). A color bar test pattern is used.The value of fSCLK is 270 MHz or 360 MHz for
SMPTE 259M, 540 MHz for SMPTE 344M or 1485 MHz for SMPTE 292M serial data rates. See the Timing Jitter Bandpass section.
2. Jitter is defined in accordance with SMPTE RP1 184-1996 as: jitter at an equipment output in the absence of input jitter.
3. All Tx jitter is measured at the output of an industry standard cable driver; connection to the cable driver is via a 50 ohm impedance differential signal from the Lattice SERDES device.
4. The cable driver drives: RL=75 ohm, AC-coupled at 270, 1485, or 2970 Mbps, RREFLVL=RREFPRE=4.75kohm 1%.
Table 3-20. Receive
Symbol
Description
BRSDI
Serial input data rate
CID
Stream of non-transitions
(=Consecutive Identical Digits)
Test Conditions
Min.
Typ.
Max.
Units
270
—
2970
Mbps
7(3G)/26(SMPTE
Triple rates)
@ 10-12 BER
—
—
Bits
Table 3-21. Reference Clock
Symbol
Description
Test Conditions
Min.
Typ.
Max.
Units
FVCLK
Video output clock frequency
27
—
74.25
MHz
DCV
Duty cycle, video clock
45
50
55
%
3-45
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
HDMI (High-Definition Multimedia Interface) Electrical and Timing
Characteristics
AC and DC Characteristics
Table 3-22. Transmit and Receive1, 2
Spec. Compliance
Symbol
Description
Min. Spec.
Max. Spec.
Units
Intra-pair Skew
—
75
ps
Inter-pair Skew
—
800
ps
TMDS Differential Clock Jitter
—
0.25
UI
Termination Resistance
40
60
Ohms
VICM
Input AC Common Mode Voltage (50-ohm Setting)
—
50
mV
TMDS Clock Jitter
Clock Jitter Tolerance
—
0.25
UI
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Transmit
Receive
RT
1. Output buffers must drive a translation device. Max. speed is 2Gbps. If translation device does not modify rise/fall time, the maximum speed
is 1.5Gbps.
2. Input buffers must be AC coupled in order to support the 3.3V common mode. Generally, HDMI inputs are terminated by an external cable
equalizer before data/clock is forwarded to the LatticeECP3 device.
3-46
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 3-15. Test Loads
Test Loads
VDDSD
75Ω
1%
VDDIO
SDO
SDO
IOL
CL
Hi-Z test eqpt. ≥ 5Ωk
(atteunation 0dB)
1.0µF
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
S1
75Ω test eqpt.
(atteunation 0dB)
CMOS
outputs
CL
IOH
VDDSD
S2
CL including probe and jig capacitance, 3pF max.
S1 - open, S2 - closed for VOH measurement.
S1 - closed, S2 - open for VOL measurement.
5.5-30pF*
75Ω
1%
50 test eqpt.
(atteunation 3.5dB)
SDO
SDO
CL
1.0µF
*Risetime compensation.
Timing Jitter Bandpass
Jitter Bandpass
0db
Slopes:
20dB/Decade
Passband Ripple
< ±1dB
Stopband
Rejection
1/10 fSCLK
10Hz
Jitter Frequency
3-47
24.9Ω
1%
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 sysCONFIG Port Timing Specifications
Over Recommended Operating Conditions
Parameter
Description
Min.
Max.
Units
POR, Configuration Initialization, and Wakeup
23
ms
—
6
ms
tVMC
Time from tICFG to the Valid Master MCLK
—
5
µs
tPRGM
PROGRAMN Low Time to Start Configuration
25
—
ns
tPRGMRJ
PROGRAMN Pin Pulse Rejection
—
10
ns
tDPPINIT
Delay Time from PROGRAMN Low to INITN Low
—
37
ns
tDPPDONE
Delay Time from PROGRAMN Low to DONE Low
—
37
ns
tDINIT
PROGRAMN High to INITN High Delay
—
1
ms
tMWC
Additional Wake Master Clock Signals After DONE Pin is High
100
500
cycles
tCZ
MCLK From Active To Low To High-Z
—
300
ns
5
—
ns
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Time from the Application of VCC, VCCAUX or VCCIO8* (Whichever Master mode
is the Last to Cross the POR Trip Point) to the Rising Edge of
Slave mode
INITN
—
tICFG
All Configuration Modes
tSUCDI
Data Setup Time to CCLK/MCLK
tHCDI
Data Hold Time to CCLK/MCLK
1
—
ns
tCODO
CCLK/MCLK to DOUT in Flowthrough Mode
—
12
ns
5
—
ns
Slave Serial
tSSCH
CCLK Minimum High Pulse
tSSCL
CCLK Minimum Low Pulse
fCCLK
CCLK Frequency
5
—
ns
Without encryption
—
33
MHz
With encryption
—
20
MHz
Master and Slave Parallel
tSUCS
CSN[1:0] Setup Time to CCLK/MCLK
7
—
ns
tHCS
CSN[1:0] Hold Time to CCLK/MCLK
1
—
ns
tSUWD
WRITEN Setup Time to CCLK/MCLK
7
—
ns
tHWD
WRITEN Hold Time to CCLK/MCLK
1
—
ns
tDCB
CCLK/MCLK to BUSY Delay Time
—
12
ns
tCORD
CCLK to Out for Read Data
—
12
ns
tBSCH
CCLK Minimum High Pulse
6
—
ns
tBSCL
CCLK Minimum Low Pulse
6
—
ns
tBSCYC
Byte Slave Cycle Time
fCCLK
CCLK/MCLK Frequency
30
—
ns
Without encryption
—
33
MHz
With encryption
—
20
MHz
Master and Slave SPI
tCFGX
INITN High to MCLK Low
—
80
ns
0.2
2
µs
15
tCSSPI
INITN High to CSSPIN Low
tSOCDO
MCLK Low to Output Valid
—
tCSPID
CSSPIN[0:1] Low to First MCLK Edge Setup Time
0.3
fCCLK
CCLK Frequency
ns
µs
Without encryption
—
33
MHz
With encryption
—
20
MHz
tSSCH
CCLK Minimum High Pulse
5
—
ns
tSSCL
CCLK Minimum Low Pulse
5
—
ns
tHLCH
HOLDN Low Setup Time (Relative to CCLK)
5
—
ns
3-48
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 sysCONFIG Port Timing Specifications (Continued)
Over Recommended Operating Conditions
Parameter
tCHHH
Description
HOLDN Low Hold Time (Relative to CCLK)
Min.
Max.
Units
5
—
ns
5
—
ns
Master and Slave SPI (Continued)
HOLDN High Hold Time (Relative to CCLK)
tHHCH
HOLDN High Setup Time (Relative to CCLK)
5
—
ns
tHLQZ
HOLDN to Output High-Z
—
9
ns
tHHQX
HOLDN to Output Low-Z
—
9
ns
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
tCHHL
Parameter
Master Clock Frequency
Min.
Max.
Units
Selected value - 15%
Selected value + 15%
MHz
40
60
%
Duty Cycle
Figure 3-16. sysCONFIG Parallel Port Read Cycle
tBSCL
tBSCYC
tBSCH
CCLK
t SUCS
tHCS
tSUWD
t HWD
CS1N
CSN
WRITEN
tDCB
BUSY
t CORD
D[0:7]
Byte 0
Byte 1
*n = last byte of read cycle.
3-49
Byte 2
Byte n*
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 3-17. sysCONFIG Parallel Port Write Cycle
tBSCYC
tBSCL
tBSCH
CCLK 1
t SUCS
tHCS
CS1N
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
CSN
t SUWD
t HWD
WRITEN
tDCB
BUSY
t HCBDI
tSUCBDI
D[0:7]
Byte 0
Byte 1
Byte 2
Byte n
1. In Master Parallel Mode the FPGA provides CCLK (MCLK). In Slave Parallel Mode the external device provides CCLK.
Figure 3-18. sysCONFIG Master Serial Port Timing
CCLK (output)
t HMCDI
tSUMCDI
DIN
t CODO
DOUT
Figure 3-19. sysCONFIG Slave Serial Port Timing
tSSCL
tSSCH
CCLK (input)
tHSCDI
t SUSCDI
DIN
t CODO
DOUT
3-50
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 3-20. Power-On-Reset (POR) Timing
VCC / VCCAUX /
VCCIO81
tICFG
INITN
DONE
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
t VMC
CCLK 2
CFG[2:0] 3
Valid
1. Time taken from VCC, VCCAUX or VCCIO8, whichever is the last to cross the POR trip point.
2. Device is in a Master Mode (SPI, SPIm).
3. The CFG pins are normally static (hard wired).
Figure 3-21. sysCONFIG Port Timing
3-51
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 3-22. Configuration from PROGRAMN Timing
tPRGMRJ
PROGRAMN
t DINIT
tDPPINIT
DONE
t DINITD
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
INITN
CCLK
CFG[2:0] 1
Valid
t IODISS
USER I/O
1. The CFG pins are normally static (hard wired)
Figure 3-23. Wake-Up Timing
PROGRAMN
INITN
DONE
Wake-Up
tMWC
CCLK
tIOENSS
USER I/O
3-52
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Figure 3-24. Master SPI Configuration Waveforms
Capture CR0
Capture CFGx
VCC
PROGRAMN
DONE
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
INITN
CSSPIN
0
1
2
3
…
7
8
9 10
… 31 32 33 34
… 127 128
CCLK
SISPI
Opcode
Address
Ignore
SOSPI
3-53
Valid Bitstream
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
JTAG Port Timing Specifications
Over Recommended Operating Conditions
Symbol
Max
Units
TCK clock frequency
Parameter
—
25
MHz
tBTCP
TCK [BSCAN] clock pulse width
40
—
ns
tBTCPH
TCK [BSCAN] clock pulse width high
20
—
ns
tBTCPL
TCK [BSCAN] clock pulse width low
20
—
ns
tBTS
TCK [BSCAN] setup time
10
—
ns
tBTH
TCK [BSCAN] hold time
8
—
ns
tBTRF
TCK [BSCAN] rise/fall time
50
—
mV/ns
—
10
ns
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Min
fMAX
tBTCO
TAP controller falling edge of clock to valid output
tBTCODIS
TAP controller falling edge of clock to valid disable
—
10
ns
tBTCOEN
TAP controller falling edge of clock to valid enable
—
10
ns
tBTCRS
BSCAN test capture register setup time
8
—
ns
tBTCRH
BSCAN test capture register hold time
25
—
ns
tBUTCO
BSCAN test update register, falling edge of clock to valid output
—
25
ns
tBTUODIS
BSCAN test update register, falling edge of clock to valid disable
—
25
ns
tBTUPOEN
BSCAN test update register, falling edge of clock to valid enable
—
25
ns
Figure 3-25. JTAG Port Timing Waveforms
TMS
TDI
tBTS
tBTCPH
tBTH
tBTCP
tBTCPL
TCK
tBTCO
tBTCOEN
TDO
Valid Data
tBTCRS
Data to be
captured
from I/O
tBTCODIS
Valid Data
tBTCRH
Data Captured
tBTUPOEN
tBUTCO
Data to be
driven out
to I/O
Valid Data
3-54
tBTUODIS
Valid Data
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Switching Test Conditions
Figure 3-26 shows the output test load that is used for AC testing. The specific values for resistance, capacitance,
voltage, and other test conditions are shown in Table 3-23.
Figure 3-26. Output Test Load, LVTTL and LVCMOS Standards
VT
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
R1
DUT
Test Poi nt
R2
CL*
*CL Includes Test Fixture and Probe Capacitance
Table 3-23. Test Fixture Required Components, Non-Terminated Interfaces
Test Condition
LVTTL and other LVCMOS settings (L -> H, H -> L)
R1
LVCMOS 2.5 I/O (Z -> H)
R2
LVCMOS 2.5 I/O (Z -> L)
1M
LVCMOS 2.5 I/O (H -> Z)
LVCMOS 2.5 I/O (L -> Z)
100
CL
0pF
VT
LVCMOS 3.3 = 1.5V
—
LVCMOS 2.5 = VCCIO/2
—
LVCMOS 1.8 = VCCIO/2
—
LVCMOS 1.5 = VCCIO/2
—
LVCMOS 1.2 = VCCIO/2
—
1M
0pF
VCCIO/2
—
VCCIO
0pF
VCCIO/2
100
0pF
—
VOH - 0.10
0pF
VOL + 0.10
VCCIO
Note: Output test conditions for all other interfaces are determined by the respective standards.
3-55
Timing Ref.
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
sysI/O Differential Electrical Characteristics
Transition Reduced LVDS (TRLVDS DC Specification)
Over Recommended Operating Conditions
Symbol
Description
Min.
Nom.
3.3
VCCO
Driver supply voltage (+/- 5%)
3.14
VID
Input differential voltage
150
VICM
Input common mode voltage
VCCO
Termination supply voltage
Max.
3
3.3
3.47
V
1200
mV
3.265
V
3.47
V
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
3.14
Units
RT
Termination resistance (off-chip)
45
50
55
Ohms
Note: LatticeECP3 only supports the TRLVDS receiver.
VCCO = 3.3V
RT
Transmitter
RT
Z0
Receiver
Current
Source
Mini LVDS
Over Recommended Operating Conditions
Parameter Symbol
ZO
RT
Description
Single-ended PCB trace impedance
Min.
Typ.
Max.
Units
30
50
75
ohms
Differential termination resistance
50
100
150
ohms
VOD
Output voltage, differential, |VOP - VOM|
300
—
600
mV
VOS
Output voltage, common mode, |VOP + VOM|/2
1
1.2
1.4
V
VOD
Change in VOD, between H and L
—
—
50
mV
VID
Change in VOS, between H and L
—
—
50
mV
VTHD
Input voltage, differential, |VINP - VINM|
200
—
600
mV
VCM
Input voltage, common mode, |VINP + VINM|/2 0.3+(VTHD/2)
—
2.1-(VTHD/2)
TR, TF
Output rise and fall times, 20% to 80%
—
—
550
ps
TODUTY
Output clock duty cycle
40
—
60
%
Note: Data is for 6mA differential current drive. Other differential driver current options are available.
3-56
DC and Switching Characteristics
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Point-to-Point LVDS (PPLVDS)
Over Recommended Operating Conditions
Description
Output driver supply (+/- 5%)
Min.
Typ.
Max.
Units
3.14
3.3
3.47
V
2.25
2.5
2.75
V
Input differential voltage
100
400
mV
Input common mode voltage
0.2
2.3
V
Output differential voltage
130
400
mV
Output common mode voltage
0.5
1.4
V
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
0.8
RSDS
Over Recommended Operating Conditions
Parameter Symbol
Description
Min.
Typ.
Max.
Units
VOD
Output voltage, differential, RT = 100 ohms
100
200
600
mV
VOS
Output voltage, common mode
0.5
1.2
1.5
V
IRSDS
Differential driver output current
1
2
6
mA
VTHD
Input voltage differential
100
—
—
mV
VCM
Input common mode voltage
0.3
—
1.5
V
TR, TF
Output rise and fall times, 20% to 80%
—
500
—
ps
TODUTY
Output clock duty cycle
35
50
65
%
Note: Data is for 2mA drive. Other differential driver current options are available.
3-57
LatticeECP3 Family Data Sheet
Pinout Information
March 2010
Preliminary Data Sheet DS1021
Signal Descriptions
Signal Name
I/O
Description
General Purpose
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
[Edge] indicates the edge of the device on which the pad is located. Valid
edge designations are L (Left), B (Bottom), R (Right), T (Top).
[Row/Column Number] indicates the PFU row or the column of the device on
which the PIC exists. When Edge is T (Top) or B (Bottom), only need to specify Column Number. When Edge is L (Left) or R (Right), only need to specify
Row Number.
P[Edge] [Row/Column Number]_[A/B]
I/O
[A/B] indicates the PIO within the PIC to which the pad is connected. Some of
these user-programmable pins are shared with special function pins. These
pins, when not used as special purpose pins, can be programmed as I/Os for
user logic. During configuration the user-programmable I/Os are tri-stated
with an internal pull-up resistor enabled. If any pin is not used (or not bonded
to a package pin), it is also tri-stated with an internal pull-up resistor enabled
after configuration.
P[Edge][Row Number]E_[A/B/C/D]
I
These general purpose signals are input-only pins and are located near the
PLLs.
GSRN
I
Global RESET signal (active low). Any I/O pin can be GSRN.
NC
—
No connect.
RESERVED
—
This pin is reserved and should not be connected to anything on the board.
GND
—
Ground. Dedicated pins.
VCC
—
Power supply pins for core logic. Dedicated pins.
VCCAUX
—
Auxiliary power supply pin. This dedicated pin powers all the differential and
referenced input buffers.
VCCIOx
—
Dedicated power supply pins for I/O bank x.
VCCA
—
SERDES, transmit, receive, PLL and reference clock buffer power supply.
VCCPLL_[LOC]
—
General purpose PLL supply pins where LOC=L (left) or R (right).
VREF1_x, VREF2_x
—
Reference supply pins for I/O bank x. Pre-determined pins in each bank are
assigned as VREF inputs. When not used, they may be used as I/O pins.
VTTx
—
Power supply for on-chip termination of I/Os (Required for DDR3 and LVDS at
1.25Gbps).
XRES1
—
10K ohm +/-1% resistor must be connected between this pad and ground.
PLL, DLL and Clock Functions
[LOC][num]_GPLL[T, C]_IN_[index]
I
General Purpose PLL (GPLL) input pads: LUM, LLM, RUM, RLM, num = row
from center, T = true and C = complement, index A,B,C...at each side.
[LOC][num]_GPLL[T, C]_FB_[index]
I
Optional feedback GPLL input pads: LUM, LLM, RUM, RLM, num = row from
center, T = true and C = complement, index A,B,C...at each side.
[LOC]0_GDLLT_IN_[index]
I/O
General Purpose DLL (GDLL) input pads where LOC=RUM or LUM, T is True
Complement, index is A or B.
[LOC]0_GDLLT_FB_[index]
I/O
Optional feedback GDLL input pads where LOC=RUM or LUM, T is True
Complement, index is A or B.
PCLK[T, C][n:0]_[3:0]
I
Primary Clock pads, T = true and C = complement, n per side, indexed by
bank and 0, 1, 2, 3 within bank.
© 2010 Lattice Semiconductor Corp. All Lattice trademarks, registered trademarks, patents, and disclaimers are as listed at www.latticesemi.com/legal. All other brand
or product names are trademarks or registered trademarks of their respective holders. The specifications and information herein are subject to change without notice.
www.latticesemi.com
4-1
DS1021 Pinout Information_01.2
Pinout Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Signal Descriptions (Cont.)
Signal Name
I/O
Description
[LOC]DQS[num]
I/O
DQ input/output pads: T (top), R (right), B (bottom), L (left), DQS, num = ball
function number.
[LOC]DQ[num]
I/O
DQ input/output pads: T (top), R (right), B (bottom), L (left), DQ, associated
DQS number.
Test and Programming (Dedicated Pins)
TMS
I
Test Mode Select input, used to control the 1149.1 state machine. Pull-up is
enabled during configuration.
Test Clock input pin, used to clock the 1149.1 state machine. No pull-up
enabled.
TDI
I
Test Data in pin. Used to load data into device using 1149.1 state machine.
After power-up, this TAP port can be activated for configuration by sending
appropriate command. (Note: once a configuration port is selected it is
locked. Another configuration port cannot be selected until the power-up
sequence). Pull-up is enabled during configuration.
TDO
O
Output pin. Test Data Out pin used to shift data out of a device using 1149.1.
VCCJ
—
Power supply pin for JTAG Test Access Port.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
I
TCK
Configuration Pads (Used During sysCONFIG)
CFG[2:0]
INITN
I
Mode pins used to specify configuration mode values latched on rising edge
of INITN. During configuration, a pull-up is enabled. These are dedicated
pins.
I/O
Open Drain pin. Indicates the FPGA is ready to be configured. During configuration, a pull-up is enabled. It is a dedicated pin.
I
Initiates configuration sequence when asserted low. This pin always has an
active pull-up. It is a dedicated pin.
DONE
I/O
Open Drain pin. Indicates that the configuration sequence is complete, and
the startup sequence is in progress. It is a dedicated pin.
CCLK
I
Input Configuration Clock for configuring an FPGA in Slave SPI, Serial, and
CPU modes. It is a dedicated pin.
MCLK
I/O
Output Configuration Clock for configuring an FPGA in SPI, SPIm, and Master configuration modes.
BUSY/SISPI
O
Parallel configuration mode busy indicator. SPI/SPIm mode data output.
CSN/SN/OEN
I/O
Parallel configuration mode active-low chip select. Slave SPI chip select.
Parallel burst Flash output enable.
PROGRAMN
CS1N/HOLDN/RDY
I
Parallel configuration mode active-low chip select. Slave SPI hold input.
WRITEN
I
Write enable for parallel configuration modes.
DOUT/CSON/CSSPI1N
O
Serial data output. Chip select output. SPI/SPIm mode chip select.
sysCONFIG Port Data I/O for Parallel mode. Open drain during configuration.
D[0]/SPIFASTN
I/O
sysCONFIG Port Data I/O for SPI or SPIm. When using the SPI or SPIm
mode, this pin should either be tied high or low, must not be left floating. Open
drain during configuration.
D1
I/O
Parallel configuration I/O. Open drain during configuration.
D2
I/O
Parallel configuration I/O. Open drain during configuration.
D3/SI
I/O
Parallel configuration I/O. Slave SPI data input. Open drain during configuration.
D4/SO
I/O
Parallel configuration I/O. Slave SPI data output. Open drain during configuration.
D5
I/O
Parallel configuration I/O. Open drain during configuration.
I/O
Parallel configuration I/O. SPI/SPIm data input. Open drain during configuration.
D6/SPID1
4-2
Pinout Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Signal Descriptions (Cont.)
Signal Name
D7/SPID0
DI/CSSPI0N/CEN
I/O
Description
I/O
Parallel configuration I/O. SPI/SPIm data input. Open drain during configuration.
I/O
Serial data input for slave serial mode. SPI/SPIm mode chip select.
2
Dedicated SERDES Signals
PCS[Index]_HDINNm
I
High-speed input, negative channel m
PCS[Index]_HDOUTNm
O
High-speed output, negative channel m
I
Negative Reference Clock Input
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
PCS[Index]_REFCLKN
PCS[Index]_HDINPm
I
High-speed input, positive channel m
PCS[Index]_HDOUTPm
O
High-speed output, positive channel m
PCS[Index]_REFCLKP
I
Positive Reference Clock Input
PCS[Index]_VCCOBm
—
Output buffer power supply, channel m (1.2V/1.5)
PCS[Index]_VCCIBm
—
Input buffer power supply, channel m (1.2V/1.5V)
1. When placing switching I/Os around these critical pins that are designed to supply the device with the proper reference or supply voltage,
care must be given.
2. m defines the associated channel in the quad.
4-3
Pinout Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
PICs and DDR Data (DQ) Pins Associated with the DDR Strobe (DQS) Pin
PICs Associated with
DQS Strobe
PIO Within PIC
DDR Strobe (DQS) and
Data (DQ) Pins
For Left and Right Edges of the Device
P[Edge] [n-3]
DQ
B
DQ
A
DQ
B
DQ
A
DQ
B
DQ
A
[Edge]DQSn
B
DQ
A
DQ
B
DQ
A
DQ
B
DQ
A
DQ
B
DQ
A
DQ
B
DQ
A
DQ
B
DQ
A
[Edge]DQSn
B
DQ
A
DQ
B
DQ
A
DQ
B
DQ
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
P[Edge] [n-2]
A
P[Edge] [n-1]
P[Edge] [n]
P[Edge] [n+1]
P[Edge] [n+2]
For Top Edge of the Device
P[Edge] [n-3]
P[Edge] [n-2]
P[Edge] [n-1]
P[Edge] [n]
P[Edge] [n+1]
P[Edge] [n+2]
Note: “n” is a row PIC number.
4-4
Pinout Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Pin Information Summary
Pin Information Summary
Pin Type
ECP3-17EA
ECP3-35EA
ECP3-70E/EA
256
484
256
484
672
484
672
1156
ftBGA fpBGA ftBGA fpBGA fpBGA fpBGA fpBGA fpBGA
36
26
14
24
14
36
36
36
48
78
6
12
6
24
24
24
34
36
Bank 3
18
44
16
54
59
54
59
86
Bank 6
20
44
18
63
61
63
67
86
Bank 7
19
32
19
36
42
36
48
54
Bank 8
24
24
24
24
24
24
24
24
Bank 0
0
0
0
0
0
0
0
0
Bank 1
0
0
0
0
0
0
0
0
Bank 2
2
2
2
4
4
4
8
8
Bank 3
0
0
2
4
4
4
12
12
Bank 6
0
0
2
4
4
4
12
12
Bank 7
4
4
4
4
4
4
8
8
Bank 8
0
0
0
0
0
0
0
0
Bank 0
0
0
0
0
0
0
0
0
Bank 1
0
0
0
0
0
0
0
0
Bank 2
0
0
0
0
0
0
0
0
General Purpose Outputs per
Bank 3
Bank
Bank 6
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Bank 7
0
0
0
0
0
0
0
0
Bank 8
0
0
0
0
0
0
0
0
26
Bank 1
Bank 2
42
48
42
60
86
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
General Purpose
Inputs/Outputs per Bank
Bank 0
General Purpose Inputs per
Bank
Total Single-Ended User I/O
133
222
133
295
310
295
380
490
VCC
6
16
6
16
32
16
32
32
VCCAUX
4
8
4
8
12
8
12
16
VTT
4
4
4
4
4
4
4
8
VCCA
4
4
4
4
8
4
8
16
VCCPLL
2
4
2
4
4
4
4
4
Bank 0
2
2
2
2
4
2
4
4
Bank 1
2
2
2
2
4
2
4
4
Bank 2
2
2
2
2
4
2
4
4
Bank 3
2
2
2
2
4
2
4
4
Bank 6
2
2
2
2
4
2
4
4
Bank 7
2
2
2
2
4
2
4
4
Bank 8
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
VCCIO
VCCJ
TAP
4
4
4
4
4
4
4
4
GND, GNDIO
50
98
50
98
139
98
139
233
NC
0
73
0
0
96
0
0
238
Reserved1
0
2
0
2
2
2
2
2
SERDES
26
26
26
26
26
26
52
78
Miscellaneous Pins
8
8
8
8
8
8
8
8
Total Bonded Pins
256
484
256
484
672
484
672
1156
4-5
Pinout Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Pin Information Summary (Cont.)
Pin Information Summary
Pin Type
Emulated Differential I/O per
Bank
ECP3-17EA
256 ftBGA
ECP3-35EA
484 fpBGA
256 ftBGA
484 fpBGA
672 fpBGA
Bank 0
13
18
13
21
24
Bank 1
7
12
7
18
18
Bank 2
2
4
1
8
8
Bank 3
4
13
5
20
19
5
13
6
22
20
6
10
6
11
13
Bank 8
12
12
12
12
12
Bank 0
0
0
0
0
0
Bank 1
0
0
0
0
0
Bank 2
2
3
3
6
6
Highspeed Differential I/O per
Bank 3
Bank
Bank 6
5
9
4
9
12
5
9
4
11
12
Bank 7
5
8
5
9
10
Bank 8
0
0
0
0
0
Bank 0
26/13
36/18
26/13
42/21
48/24
Bank 1
14/7
24/12
14/7
36/18
36/18
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Bank 6
Bank 7
Total Single Ended/ Total
Differential I/O per Bank
DDR Groups Bonded per
Bank
Bank 2
8/4
14/7
8/4
28/14
28/14
Bank 3
18/9
44/22
18/9
58/29
63/31
Bank 6
20/10
44/22
20/10
67/33
65/32
Bank 7
23/11
36/18
23/11
40/20
46/23
Bank 8
24/12
24/12
24/12
24/12
24/12
Bank 0
2
3
2
3
4
Bank 1
1
2
1
3
3
Bank 2
0
1
0
2
2
Bank 3
1
3
1
3
4
Bank 6
1
3
1
4
4
Bank 7
1
2
1
3
3
Configuration Bank 8
0
0
0
0
0
1
1
1
1
1
SERDES Quads
1. These pins must remain floating on the board.
4-6
Pinout Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Pin Information Summary (Cont.)
Pin Information Summary
Pin Type
ECP3-70E
484 fpBGA 672 fpBGA
ECP3-70EA
1156
fpBGA
484 fpBGA 672 fpBGA
Bank 0
21
30
43
21
Bank 1
18
24
39
18
24
39
Bank 2
10
15
16
8
12
13
Bank 3
23
27
39
20
23
33
Bank 6
26
30
39
22
25
33
Bank 7
14
20
22
11
16
18
Bank 8
12
12
12
12
12
12
Bank 0
0
0
0
0
0
0
Bank 1
0
0
0
0
0
0
Bank 2
High-Speed Differential
Bank 3
I/O per Bank
Bank 6
4
6
6
6
9
9
6
8
10
9
12
16
7
9
10
11
14
16
13
43
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Emulated Differential
I/O per Bank
30
1156
fpBGA
Total Single-Ended/
Total Differential I/O
per Bank
DDR Groups Bonded
per Bank
SERDES Quads
Bank 7
6
8
9
9
12
Bank 8
0
0
0
0
0
0
Bank 0
42/21
60/30
86/43
42/21
60/30
86/43
Bank 1
36/18
48/24
78/39
36/18
48/24
78/39
Bank 2
28/14
42/21
44/22
28/14
42/21
44/22
Bank 3
58/29
71/35
98/49
58/29
71/35
98/49
Bank 6
67/33
79/38
98/49
67/33
78/39
98/49
Bank 7
40/20
56/28
62/31
40/20
56/28
62/31
Bank 8
24/12
24/12
24/12
24/12
24/12
24/12
Bank 0
3
5
7
3
5
7
Bank 1
3
4
7
3
4
7
Bank 2
2
3
3
2
3
3
Bank 3
3
4
5
3
4
5
Bank 6
4
4
5
4
4
5
Bank 7
3
4
4
3
4
4
Configuration Bank 8
0
0
0
0
0
0
1
2
3
1
2
3
4-7
Pinout Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Pin Information Summary (Cont.)
Pin Information Summary
ECP3-150EA
484
fpBGA
672
fpBGA
1156
fpBGA
672
fpBGA
1156
fpBGA
Bank 0
42
60
86
60
94
Bank 1
36
48
78
48
86
Bank 2
24
34
36
34
58
Bank 3
54
59
86
59
104
Bank 6
63
67
86
67
104
Pin Type
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
General Purpose
Inputs/Outputs per bank
ECP3-95E/EA
Bank 7
36
48
54
48
76
Bank 8
24
24
24
24
24
Bank 0
0
0
0
0
0
Bank 1
0
0
0
0
0
Bank 2
4
8
8
8
8
Bank 3
4
12
12
12
12
Bank 6
4
12
12
12
12
Bank 7
4
8
8
8
8
Bank 8
0
0
0
0
0
Bank 0
0
0
0
0
0
Bank 1
0
0
0
0
0
Bank 2
0
0
0
0
0
General Purpose Outputs per
Bank 3
Bank
Bank 6
0
0
0
0
0
0
0
0
0
0
Bank 7
0
0
0
0
0
Bank 8
0
0
0
0
0
Total Single-Ended User I/O
295
380
490
380
586
VCC
16
32
32
32
32
VCCAUX
8
12
16
12
16
VTT
4
4
8
4
8
VCCA
4
8
16
8
16
VCCPLL
4
4
4
4
4
Bank 0
2
4
4
4
4
Bank 1
2
4
4
4
4
Bank 2
2
4
4
4
4
Bank 3
2
4
4
4
4
Bank 6
2
4
4
4
4
Bank 7
2
4
4
4
4
Bank 8
2
2
2
2
2
1
1
1
1
1
General Purpose Inputs per
Bank
VCCIO
VCCJ
TAP
4
4
4
4
4
GND, GNDIO
98
139
233
139
233
NC
0
0
238
0
116
Reserved1
2
2
2
2
2
SERDES
26
52
78
52
104
Miscellaneous Pins
8
8
8
8
8
Total Bonded Pins
484
672
1156
672
1156
4-8
Pinout Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Pin Information Summary (Cont.)
Pin Information Summary
ECP3-95EA
ECP3-150EA
672
fpBGA
1156
fpBGA
484
fpBGA
672
fpBGA
1156
fpBGA
672
fpBGA
1156
fpBGA
Bank 0
21
30
43
21
30
43
30
47
Bank 1
18
24
39
18
24
39
24
43
Bank 2
10
15
16
8
12
13
12
18
Bank 3
23
27
39
20
23
33
23
37
Bank 6
26
30
39
22
25
33
25
37
Bank 7
14
20
22
11
16
18
16
24
Bank 8
12
12
12
12
12
12
12
12
Bank 0
0
0
0
0
0
0
0
0
Bank 1
0
0
0
0
0
0
0
0
Bank 2
4
6
6
6
9
9
9
15
Bank 3
6
8
10
9
12
16
12
21
Bank 6
7
9
10
11
14
16
14
21
Bank 7
6
8
9
9
12
13
12
18
Pin Type
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Emulated
Differential I/O
per Bank
ECP3-95E
484
fpBGA
Highspeed
Differential I/O
per Bank
Bank 8
0
0
0
0
0
0
0
0
Bank 0
42/21
60/30
86/43
42/21
60/30
86/43
60/30
94/47
Bank 1
36/18
48/24
78/39
36/18
48/24
78/39
48/24
86/43
Total Single Ended/ Bank 2
Bank 3
Total Differential
I/O per Bank
Bank 6
28/14
42/21
44/22
28/14
42/21
44/22
42/21
66/33
58/29
71/35
98/49
58/29
71/35
98/49
71/35
116/58
67/33
78/39
98/49
67/33
78/39
98/49
78/39
116/58
Bank 7
40/20
56/28
62/31
40/20
56/28
62/31
56/28
84/42
Bank 8
24/12
24/12
24/12
24/12
24/12
24/12
24/12
24/12
Bank 0
3
5
7
3
5
7
5
7
Bank 1
3
4
7
3
4
7
4
7
Bank 2
2
3
3
2
3
3
3
4
Bank 3
3
4
5
3
4
5
4
7
Bank 6
4
4
5
4
4
5
4
7
Bank 7
3
4
4
3
4
4
4
6
Configuration
Bank8
0
0
0
0
0
0
0
0
1
2
3
1
2
3
2
4
DDR Groups
Bonded
per Bank
SERDES Quads
1.These pins must remain floating on the board.
4-9
Pinout Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Logic Signal Connections
Package pinout information can be found under “Data Sheets” on the LatticeECP3 product pages on the Lattice
website at www.latticesemi.com/products/fpga/ecp3 and in the Lattice ispLEVER Design Planner software. To create pinout information from within Design Planner, select View -> Package View. Then select Select File ->
Export and choose a type of output file. See Design Planner help for more information.
Thermal Management
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Thermal management is recommended as part of any sound FPGA design methodology. To assess the thermal
characteristics of a system, Lattice specifies a maximum allowable junction temperature in all device data sheets.
Designers must complete a thermal analysis of their specific design to ensure that the device and package do not
exceed the junction temperature limits. Refer to the Thermal Management document to find the device/package
specific thermal values.
For Further Information
For further information regarding Thermal Management, refer to the following:
• Thermal Management document
• TN1181, Power Consumption and Management for LatticeECP3 Devices
• Power Calculator tool included with the Lattice ispLEVER design tool, or as a standalone download from
www.latticesemi.com/software
4-10
LatticeECP3 Family Data Sheet
Ordering Information
March 2010
Preliminary Data Sheet DS1021
LatticeECP3 Part Number Description
LFE3 – XXX XX – X XXXXXX X
Device Family
ECP3 (LatticeECP3 FPGA + SERDES)
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Grade
C = Commercial
I = Industrial
Package
FTN256 = 256-ball Lead-Free ftBGA
FN484 = 484-ball Lead-Free fpBGA
FN672 = 672-ball Lead-Free fpBGA
FN1156 = 1156-ball Lead-Free fpBGA
Logic Capacity
17 = 17K LUTs
35 = 33K LUTs
70 = 67K LUTs
95 = 92K LUTs
150 = 149K LUTs
Supply Voltage
E or EA = 1.2V
Speed
6 = Slowest
7
8 = Fastest
Ordering Information
LatticeECP3 devices have top-side markings, for commercial and industrial grades, as shown below:
Commercial
Industrial
ECP3
ECP3
LFE3-95E
7FN672C
LFE3-95E
7FN672I
Datecode
Datecode
© 2010 Lattice Semiconductor Corp. All Lattice trademarks, registered trademarks, patents, and disclaimers are as listed at www.latticesemi.com/legal. All other brand
or product names are trademarks or registered trademarks of their respective holders. The specifications and information herein are subject to change without notice.
www.latticesemi.com
5-1
DS1021 Order Info_01.2
Ordering Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
LatticeECP3 Devices, Lead-Free Packaging
The following devices may have associated errata. Specific devices with associated errata will be notated with a
footnote.
Commercial
Part Number
Voltage
Grade
Package
Pins
Temp.
LUTs (K)
LFE3-17EA-6FTN256C
1.2V
-6
Lead-Free ftBGA
256
COM
17
LFE3-17EA-7FTN256C
1.2V
-7
Lead-Free ftBGA
256
COM
17
1.2V
-8
Lead-Free ftBGA
256
COM
17
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
LFE3-17EA-8FTN256C
LFE3-17EA-6FN484C
1.2V
-6
Lead-Free fpBGA
484
COM
17
LFE3-17EA-7FN484C
1.2V
-7
Lead-Free fpBGA
484
COM
17
LFE3-17EA-8FN484C
1.2V
-8
Lead-Free fpBGA
484
COM
17
Voltage
Grade
Package
Pins
Temp.
LUTs (K)
LFE3-35EA-6FTN256C
Part Number
1.2V
-6
Lead-Free ftBGA
256
COM
33
LFE3-35EA-7FTN256C
1.2V
-7
Lead-Free ftBGA
256
COM
33
LFE3-35EA-8FTN256C
1.2V
-8
Lead-Free ftBGA
256
COM
33
LFE3-35EA-6FN484C
1.2V
-6
Lead-Free fpBGA
484
COM
33
LFE3-35EA-7FN484C
1.2V
-7
Lead-Free fpBGA
484
COM
33
LFE3-35EA-8FN484C
1.2V
-8
Lead-Free fpBGA
484
COM
33
LFE3-35EA-6FN672C
1.2V
-6
Lead-Free fpBGA
672
COM
33
LFE3-35EA-7FN672C
1.2V
-7
Lead-Free fpBGA
672
COM
33
LFE3-35EA-8FN672C
1.2V
-8
Lead-Free fpBGA
672
COM
33
Voltage
Grade
Package
Pins
Temp.
LUTs (K)
LFE3-70EA-6FN484C
Part Number
1.2V
-6
Lead-Free fpBGA
484
COM
67
LFE3-70EA-7FN484C
1.2V
-7
Lead-Free fpBGA
484
COM
67
LFE3-70EA-8FN484C
1.2V
-8
Lead-Free fpBGA
484
COM
67
LFE3-70EA-6FN672C
1.2V
-6
Lead-Free fpBGA
672
COM
67
LFE3-70EA-7FN672C
1.2V
-7
Lead-Free fpBGA
672
COM
67
LFE3-70EA-8FN672C
1.2V
-8
Lead-Free fpBGA
672
COM
67
LFE3-70EA-6FN1156C
1.2V
-6
Lead-Free fpBGA
1156
COM
67
LFE3-70EA-7FN1156C
1.2V
-7
Lead-Free fpBGA
1156
COM
67
LFE3-70EA-8FN1156C
1.2V
-8
Lead-Free fpBGA
1156
COM
67
5-2
Ordering Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Part Number
Grade
Package
Pins
Temp.
LUTs (K)
1.2V
-6
Lead-Free fpBGA
484
COM
67
LFE3-70E-7FN484C
1
1.2V
-7
Lead-Free fpBGA
484
COM
67
LFE3-70E-8FN484C
1
1.2V
-8
Lead-Free fpBGA
484
COM
67
LFE3-70E-6FN672C
1
1.2V
-6
Lead-Free fpBGA
672
COM
67
LFE3-70E-7FN672C
1
1.2V
-7
Lead-Free fpBGA
672
COM
67
LFE3-70E-8FN672C1
1.2V
-8
Lead-Free fpBGA
672
COM
67
LFE3-70E-6FN1156C1
1.2V
-6
Lead-Free fpBGA
1156
COM
67
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Voltage
LFE3-70E-6FN484C1
1
1.2V
-7
Lead-Free fpBGA
1156
COM
67
LFE3-70E-8FN1156C1
1.2V
-8
Lead-Free fpBGA
1156
COM
67
LFE3-70E-7FN1156C
1.This device has associated errata. View www.latticesemi.com/documents/ds1021.zip for a description of the errata.
Voltage
Grade
Package
Pins
Temp.
LUTs (K)
LFE3-95EA-6FN484C
Part Number
1.2V
-6
Lead-Free fpBGA
484
COM
92
LFE3-95EA-7FN484C
1.2V
-7
Lead-Free fpBGA
484
COM
92
LFE3-95EA-8FN484C
1.2V
-8
Lead-Free fpBGA
484
COM
92
LFE3-95EA-6FN672C
1.2V
-6
Lead-Free fpBGA
672
COM
92
LFE3-95EA-7FN672C
1.2V
-7
Lead-Free fpBGA
672
COM
92
LFE3-95EA-8FN672C
1.2V
-8
Lead-Free fpBGA
672
COM
92
LFE3-95EA-6FN1156C
1.2V
-6
Lead-Free fpBGA
1156
COM
92
LFE3-95EA-7FN1156C
1.2V
-7
Lead-Free fpBGA
1156
COM
92
LFE3-95EA-8FN1156C
1.2V
-8
Lead-Free fpBGA
1156
COM
92
Part Number
Voltage
Grade
Package
Pins
Temp.
LUTs (K)
1
1.2V
-6
Lead-Free fpBGA
484
COM
92
LFE3-95E-7FN484C1
1.2V
-7
Lead-Free fpBGA
484
COM
92
1
1.2V
-8
Lead-Free fpBGA
484
COM
92
LFE3-95E-6FN672C1
1.2V
-6
Lead-Free fpBGA
672
COM
92
1
1.2V
-7
Lead-Free fpBGA
672
COM
92
LFE3-95E-8FN672C1
1.2V
-8
Lead-Free fpBGA
672
COM
92
1
1.2V
-6
Lead-Free fpBGA
1156
COM
92
LFE3-95E-7FN1156C1
1.2V
-7
Lead-Free fpBGA
1156
COM
92
1
1.2V
-8
Lead-Free fpBGA
1156
COM
92
LFE3-95E-6FN484C
LFE3-95E-8FN484C
LFE3-95E-7FN672C
LFE3-95E-6FN1156C
LFE3-95E-8FN1156C
1.This device has associated errata. View www.latticesemi.com/documents/ds1021.zip for a description of the errata.
Part Number
Voltage
Grade
Package
Pins
Temp.
LUTs (K)
LFE3-150EA-6FN672C
1.2V
-6
Lead-Free fpBGA
672
COM
149
LFE3-150EA-7FN672C
1.2V
-7
Lead-Free fpBGA
672
COM
149
LFE3-150EA-8FN672C
1.2V
-8
Lead-Free fpBGA
672
COM
149
LFE3-150EA-6FN1156C
1.2V
-6
Lead-Free fpBGA
1156
COM
149
LFE3-150EA-7FN1156C
1.2V
-7
Lead-Free fpBGA
1156
COM
149
LFE3-150EA-8FN1156C
1.2V
-8
Lead-Free fpBGA
1156
COM
149
5-3
Ordering Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Part Number
Grade
Package
Pins
Temp.
LUTs (K)
1.2V
-6
Lead-Free fpBGA
672
COM
149
LFE3-150EA-7FN672CTW*
1.2V
-7
Lead-Free fpBGA
672
COM
149
LFE3-150EA-8FN672CTW*
1.2V
-8
Lead-Free fpBGA
672
COM
149
LFE3-150EA-6FN1156CTW*
1.2V
-6
Lead-Free fpBGA
1156
COM
149
LFE3-150EA-7FN1156CTW*
1.2V
-7
Lead-Free fpBGA
1156
COM
149
LFE3-150EA-8FN1156CTW*
1.2V
-8
Lead-Free fpBGA
1156
COM
149
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Voltage
LFE3-150EA-6FN672CTW*
*Note: Specifications for the LFE3-150EA-spFNpkgCTW and LFE3-150EA-spFNpkgITW devices, (where sp is the speed and
pkg is the package), are the same as the LFE3-150EA-spFNpkgC and LFE3-150EA-spFNpkgI devices respectively, except as
specified below.
• The CTC (Clock Tolerance Circuit) inside the SERDES hard PCS in the TW device is not functional but it can be
bypassed and implemented in soft IP.
• The SERDES XRES pin on the TW device passes CDM testing at 250V.
5-4
Ordering Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Industrial
The following devices may have associated errata. Specific devices with associated errata will be notated with a
footnote.
Part Number
LFE3-17EA-6FTN256I
Voltage
Grade
Package
Pins
Temp.
LUTs (K)
1.2V
-6
Lead-Free ftBGA
256
IND
17
1.2V
-7
Lead-Free ftBGA
256
IND
17
1.2V
-8
Lead-Free ftBGA
256
IND
17
LFE3-17EA-6FN484I
1.2V
-6
Lead-Free fpBGA
484
IND
17
LFE3-17EA-7FN484I
1.2V
-7
Lead-Free fpBGA
484
IND
17
LFE3-17EA-8FN484I
1.2V
-8
Lead-Free fpBGA
484
IND
17
LUTs (K)
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
LFE3-17EA-7FTN256I
LFE3-17EA-8FTN256I
Voltage
Grade
Package
Pins
Temp.
LFE3-35EA-6FTN256I
Part Number
1.2V
-6
Lead-Free ftBGA
256
IND
33
LFE3-35EA-7FTN256I
1.2V
-7
Lead-Free ftBGA
256
IND
33
LFE3-35EA-8FTN256I
1.2V
-8
Lead-Free ftBGA
256
IND
33
LFE3-35EA-6FN484I
1.2V
-6
Lead-Free fpBGA
484
IND
33
LFE3-35EA-7FN484I
1.2V
-7
Lead-Free fpBGA
484
IND
33
LFE3-35EA-8FN484I
1.2V
-8
Lead-Free fpBGA
484
IND
33
LFE3-35EA-6FN672I
1.2V
-6
Lead-Free fpBGA
672
IND
33
LFE3-35EA-7FN672I
1.2V
-7
Lead-Free fpBGA
672
IND
33
LFE3-35EA-8FN672I
1.2V
-7
Lead-Free fpBGA
672
IND
33
Voltage
Grade
Package
Pins
Temp.
LUTs (K)
LFE3-70EA-6FN484I
1.2V
-6
Lead-Free fpBGA
484
IND
67
LFE3-70EA-7FN484I
1.2V
-7
Lead-Free fpBGA
484
IND
67
LFE3-70EA-8FN484I
1.2V
-8
Lead-Free fpBGA
484
IND
67
LFE3-70EA-6FN672I
1.2V
-6
Lead-Free fpBGA
672
IND
67
LFE3-70EA-7FN672I
1.2V
-7
Lead-Free fpBGA
672
IND
67
LFE3-70EA-8FN672I
1.2V
-8
Lead-Free fpBGA
672
IND
67
Part Number
LFE3-70EA-6FN1156I
1.2V
-6
Lead-Free fpBGA
1156
IND
67
LFE3-70EA-7FN1156I
1.2V
-7
Lead-Free fpBGA
1156
IND
67
LFE3-70EA-8FN1156I
1.2V
-8
Lead-Free fpBGA
1156
IND
67
5-5
Ordering Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Part Number
Grade
Package
Pins
Temp.
LUTs (K)
1.2V
-6
Lead-Free fpBGA
484
IND
67
LFE3-70E-7FN484I1
1.2V
-7
Lead-Free fpBGA
484
IND
67
1
LFE3-70E-8FN484I
1.2V
-8
Lead-Free fpBGA
484
IND
67
LFE3-70E-6FN672I1
1.2V
-6
Lead-Free fpBGA
672
IND
67
1
LFE3-70E-7FN672I
1.2V
-7
Lead-Free fpBGA
672
IND
67
LFE3-70E-8FN672I1
1.2V
-8
Lead-Free fpBGA
672
IND
67
LFE3-70E-6FN1156I1
1.2V
-6
Lead-Free fpBGA
1156
IND
67
1
1.2V
-7
Lead-Free fpBGA
1156
IND
67
LFE3-70E-8FN1156I1
1.2V
-8
Lead-Free fpBGA
1156
IND
67
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Voltage
LFE3-70E-6FN484I1
LFE3-70E-7FN1156I
1.This device has associated errata. View www.latticesemi.com/documents/ds1021.zip for a description of the errata.
Voltage
Grade
Package
Pins
Temp.
LUTs (K)
LFE3-95EA-6FN484I
Part Number
1.2V
-6
Lead-Free fpBGA
484
IND
92
LFE3-95EA-7FN484I
1.2V
-7
Lead-Free fpBGA
484
IND
92
LFE3-95EA-8FN484I
1.2V
-8
Lead-Free fpBGA
484
IND
92
LFE3-95EA-6FN672I
1.2V
-6
Lead-Free fpBGA
672
IND
92
LFE3-95EA-7FN672I
1.2V
-7
Lead-Free fpBGA
672
IND
92
LFE3-95EA-8FN672I
1.2V
-8
Lead-Free fpBGA
672
IND
92
LFE3-95EA-6FN1156I
1.2V
-6
Lead-Free fpBGA
1156
IND
92
LFE3-95EA-7FN1156I
1.2V
-7
Lead-Free fpBGA
1156
IND
92
LFE3-95EA-8FN1156I
1.2V
-8
Lead-Free fpBGA
1156
IND
92
Part Number
Voltage
Grade
Package
Pins
Temp.
LUTs (K)
1
1.2V
-6
Lead-Free fpBGA
484
IND
92
LFE3-95E-7FN484I1
1.2V
-7
Lead-Free fpBGA
484
IND
92
1
LFE3-95E-8FN484I
1.2V
-8
Lead-Free fpBGA
484
IND
92
LFE3-95E-6FN672I1
1.2V
-6
Lead-Free fpBGA
672
IND
92
1
LFE3-95E-7FN672I
1.2V
-7
Lead-Free fpBGA
672
IND
92
LFE3-95E-8FN672I1
1.2V
-8
Lead-Free fpBGA
672
IND
92
1
1.2V
-6
Lead-Free fpBGA
1156
IND
92
LFE3-95E-7FN1156I1
1.2V
-7
Lead-Free fpBGA
1156
IND
92
1
1.2V
-8
Lead-Free fpBGA
1156
IND
92
LFE3-95E-6FN484I
LFE3-95E-6FN1156I
LFE3-95E-8FN1156I
1.This device has associated errata. View www.latticesemi.com/documents/ds1021.zip for a description of the errata.
5-6
Ordering Information
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Part Number
Grade
Package
Pins
Temp.
LUTs (K)
1.2V
-6
Lead-Free fpBGA
672
IND
149
LFE3-150EA-7FN672I
1.2V
-7
Lead-Free fpBGA
672
IND
149
LFE3-150EA-8FN672I
1.2V
-8
Lead-Free fpBGA
672
IND
149
LFE3-150EA-6FN1156I
1.2V
-6
Lead-Free fpBGA
1156
IND
149
LFE3-150EA-7FN1156I
1.2V
-7
Lead-Free fpBGA
1156
IND
149
LFE3-150EA-8FN1156I
1.2V
-8
Lead-Free fpBGA
1156
IND
149
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Voltage
LFE3-150EA-6FN672I
Part Number
Voltage
Grade
Package
Pins
Temp.
LUTs (K)
LFE3-150EA-6FN672ITW*
1.2V
-6
Lead-Free fpBGA
672
IND
149
LFE3-150EA-7FN672ITW*
1.2V
-7
Lead-Free fpBGA
672
IND
149
LFE3-150EA-8FN672ITW*
1.2V
-8
Lead-Free fpBGA
672
IND
149
LFE3-150EA-6FN1156ITW*
1.2V
-6
Lead-Free fpBGA
1156
IND
149
LFE3-150EA-7FN1156ITW*
1.2V
-7
Lead-Free fpBGA
1156
IND
149
LFE3-150EA-8FN1156ITW*
1.2V
-8
Lead-Free fpBGA
1156
IND
149
*Note: Specifications for the LFE3-150EA-spFNpkgCTW and LFE3-150EA-spFNpkgITW devices, (where sp is the speed and
pkg is the package), are the same as the LFE3-150EA-spFNpkgC and LFE3-150EA-spFNpkgI devices respectively, except as
specified below.
• The CTC (Clock Tolerance Circuit) inside the SERDES hard PCS in the TW device is not functional but it can be
bypassed and implemented in soft IP.
• The SERDES XRES pin on the TW device passes CDM testing at 250V.
5-7
LatticeECP3 Family Data Sheet
Supplemental Information
February 2009
Preliminary Data Sheet DS1021
For Further Information
A variety of technical notes for the LatticeECP3 family are available on the Lattice website at www.latticesemi.com.
• TN1169, LatticeECP3 sysCONFIG Usage Guide
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
• TN1176, LatticeECP3 SERDES/PCS Usage Guide
• TN1177, LatticeECP3 sysIO Usage Guide
• TN1178, LatticeECP3 sysCLOCK PLL/DLL Design and Usage Guide
• TN1179, LatticeECP3 Memory Usage Guide
• TN1180, LatticeECP3 High-Speed I/O Interface
• TN1181, Power Consumption and Management for LatticeECP3 Devices
• TN1182, LatticeECP3 sysDSP Usage Guide
• TN1184, LatticeECP3 Soft Error Detection (SED) Usage Guide
• TN1189, LatticeECP3 Hardware Checklist
For further information on interface standards refer to the following websites:
• JEDEC Standards (LVTTL, LVCMOS, SSTL, HSTL): www.jedec.org
• PCI: www.pcisig.com
© 2009 Lattice Semiconductor Corp. All Lattice trademarks, registered trademarks, patents, and disclaimers are as listed at www.latticesemi.com/legal. All other brand
or product names are trademarks or registered trademarks of their respective holders. The specifications and information herein are subject to change without notice.
www.latticesemi.com
6-1
DS1021 Further Info_01.0
LatticeECP3 Family Data Sheet
Revision History
March 2010
Preliminary Data Sheet DS1021
Date
Version
Section
February 2009
01.0
—
May 2009
01.1
All
Removed references to Parallel burst mode Flash.
Features - Changed 250 Mbps to 230 Mbps in Embedded SERDES bulleted section and added a footnote to indicate 230 Mbps applies to
8b10b and 10b12b applications.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Introduction
Change Summary
Initial release.
Updated data for ECP3-17 in LatticeECP3 Family Selection Guide table.
Changed embedded memory from 552 to 700 Kbits in LatticeECP3
Family Selection Guide table.
Architecture
Updated description for CLKFB in General Purpose PLL Diagram.
Corrected Primary Clock Sources text section.
Corrected Secondary Clock/Control Sources text section.
Corrected Secondary Clock Regions table.
Corrected note below Detailed sysDSP Slice Diagram.
Corrected Clock, Clock Enable, and Reset Resources text section.
Corrected ECP3-17 EBR number in Embedded SRAM in the
LatticeECP3 Family table.
Added On-Chip Termination Options for Input Modes table.
Updated Available SERDES Quads per LatticeECP3 Devices table.
Updated Simplified Channel Block Diagram for SERDES/PCS Block
diagram.
Updated Device Configuration text section.
Corrected software default value of MCLK to be 2.5 MHz.
DC and Switching
Characteristics
Updated VCCOB Min/Max data in Recommended Operating Conditions
table.
Corrected footnote 2 in sysIO Recommended Operating Conditions
table.
Added added footnote 7 for tSKEW_PRIB to External Switching Characteristics table.
Added 2-to-1 Gearing text section and table.
Updated External Reference Clock Specification (refclkp/refclkn) table.
LatticeECP3 sysCONFIG Port Timing Specifications - updated tDINIT
information.
Added sysCONFIG Port Timing waveform.
Serial Input Data Specifications table, delete Typ data for VRX-DIFF-S.
Added footnote 4 to sysCLOCK PLL Timing table for tPFD.
Added SERDES/PCS Block Latency Breakdown table.
External Reference Clock Specifications table, added footnote 4, add
symbol name vREF-IN-DIFF.
Added SERDES External Reference Clock Waveforms.
Updated Serial Output Timing and Levels table.
Pin-to-pin performance table, changed "typically 3% slower" to "typically
slower".
© 2010 Lattice Semiconductor Corp. All Lattice trademarks, registered trademarks, patents, and disclaimers are as listed at www.latticesemi.com/legal. All other brand
or product names are trademarks or registered trademarks of their respective holders. The specifications and information herein are subject to change without notice.
www.latticesemi.com
7-1
DS1021 Revision History
Revision History
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Date
Version
Section
May 2009
(cont.)
01.1
(cont.)
DC and Switching
Characteristics
(cont.)
Change Summary
Updated timing information
Updated SERDES minimum frequency.
Added data to the following tables: External Switching Characteristics,
Internal Switching Characteristics, Family Timing Adders, Maximum I/O
Buffer Speed, DLL Timing, High Speed Data Transmitter, Channel Output Jitter, Typical Building Block Function Performance, Register-toRegister Performance, and Power Supply Requirements.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Updated Serial Input Data Specifications table.
Updated Transmit table, Serial Rapid I/O Type 2 Electrical and Timing
Characteristics section.
Pinout Information
Updated Signal Description tables.
Updated Pin Information Summary tables and added footnote 1.
July 2009
01.2
Multiple
Architecture
Changed references of “multi-boot” to “dual-boot” throughout the data
sheet.
Updated On-Chip Programmable Termination bullets.
Updated On-Chip Termination Options for Input Modes table.
Updated On-Chip Termination figure.
DC and Switching
Characterisitcs
Changed min/max data for FREF_PPM and added footnote 4 in
SERDES External Reference Clock Specification table.
Updated SERDES minimum frequency.
Pinout Information
Corrected MCLK to be I/O and CCLK to be I in Signal Descriptions table
Corrected truncated numbers for VCCIB and VCCOB in Recommended
Operating Conditions table.
August 2009
01.3
DC and Switching
Characterisitcs
September 2009
01.4
Architecture
Corrected link in sysMEM Memory Block section.
Updated information for On-Chip Programmable Termination and modified corresponding figure.
Added footnote 2 to On-Chip Programmable Termination Options for
Input Modes table.
Corrected Per Quadrant Primary Clock Selection figure.
DC and Switching
Characterisitcs
Modified -8 Timing data for 1024x18 True-Dual Port RAM (Read-BeforeWrite, EBR Output Registers)
Added ESD Performance table.
LatticeECP3 External Switching Characteristics table - updated data for
tDIBGDDR, tW_PRI, tW_EDGE and tSKEW_EDGE_DQS.
LatticeECP3 Internal Switching Characteristics table - updated data for
tCOO_PIO and added footnote #4.
sysCLOCK PLL Timing table - updated data for fOUT.
External Reference Clock Specification (refclkp/refclkn) table - updated
data for VREF-IN-SE and VREF-IN-DIFF.
LatticeECP3 sysCONFIG Port Timing Specifications table - updated
data for tMWC.
Added TRLVDS DC Specification table and diagram.
Updated Mini LVDS table.
November 2009
01.5
Introduction
Updated Embedded SERDES features.
Architecture
Updated Figure 2-4, General Purpose PLL Diagram.
Added SONET/SDH to Embedded SERDES protocols.
Updated SONET/SDH to SERDES and PCS protocols.
7-2
Revision History
LatticeECP3 Family Data Sheet
Lattice Semiconductor
Date
Version
Section
November 2009
(cont.)
01.5
(cont.)
Architecture
(cont.)
DC and Switching
Characterisitcs
Change Summary
Updated Table 2-13, SERDES Standard Support to include SONET/
SDH and updated footnote 2.
Added footnote to ESD Performance table.
Updated SERDES Power Supply Requirements table and footnotes.
Updated Maximum I/O Buffer Speed table.
Updated Pin-to-Pin Peformance table.
Updated sysCLOCK PLL Timing table.
SE
C
E
U DA L
R
a
T
R A tt
EN S ic
e
T HE EC
IN E P
FO T 3
F
R O EA
M R
A
TI
O
N
Updated DLL timing table.
Updated High-Speed Data Transmitter tables.
Updated High-Speed Data Receiver table.
Updated footnote for Receiver Total Jitter Tolerance Specification table.
Updated Periodic Receiver Jitter Tolerance Specification table.
Updated SERDES External Reference Clock Specification table.
Updated PCI Express Electrical and Timing AC and DC Characteristics.
Deleted Reference Clock table for PCI Express Electrical and Timing
AC and DC Characteristics.
Updated SMPTE AC/DC Characteristics Transmit table.
Updated Mini LVDS table.
Updated RSDS table.
Added Supply Current (Standby) table for EA devices.
Updated Internal Switching Characteristics table.
Updated Register-to-Register Performance table.
Added HDMI Electrical and Timing Characteristics data.
Updated Family Timing Adders table.
Updated sysCONFIG Port Timing Specifications table.
Updated Recommended Operating Conditions table.
Updated Hot Socket Specifications table.
Updated Single-Ended DC table.
Updated TRLVDS table and figure.
Updated Serial Data Input Specifications table.
Updated HDMI Transmit and Receive table.
Ordering Information
March 2010
01.6
Architecture
Added LFE3-150EA “TW” devices and footnotes to the Commercial and
Industrial tables.
Added Read-Before-Write information.
DC and Switching
Characteristics
Added footnote #6 to Maximum I/O Buffer Speed table.
Pinout Information
Added pin information for the LatticeECP3-70EA and LatticeECP395EA devices.
Ordering Information
Corrected minimum operating conditions for input and output differential
voltages in the Point-to-Point LVDS table.
Added ordering part numbers for the LatticeECP3-70EA and
LatticeECP3-95EA devices.
Removed dual mark information.
7-3
工商网监
湘ICP备2023018690号
