LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
LM97593 Dual ADC / Digital Tuner / AGC
Check for Samples: LM97593
FEATURES
DESCRIPTION
•
•
The LM97593 Dual ADC / Digital Tuner / AGC IC is a
two channel digital downconverter (DDC) with
integrated 12-bit analog-to-digital converters (ADCs)
and automatic gain control (AGC). The LM97593
further enhances TI’s Diversity Receiver Chipset
(DRCS) by integrating a wide-bandwidth dual ADC
core with the DDC. The complete DRCS includes one
LM97593 Dual ADC / Digital Tuner / AGC and two
CLC5526 digitally controlled variable gain amplifiers
(DVGAs). This system allows direct IF sampling of
signals up to 300MHz for enhanced receiver
performance and reduced system costs. A block
diagram
for
a
DRCS-based
narrowband
communications system is shown in Figure 1.
1
2
•
•
•
•
•
•
•
•
•
100% Software Compatible with the CLC5903
Pin Compatible with the CLC5903 Except for
the Analog Input and Reference Section
123 dB Dynamic Range with CLC5526 DVGA
(200kHz)
On-chip Precision Reference
User Programmable AGC with Enhanced
Power Detector
Channel Filters Include a Fourth Order CIC
Followed by 21-tap and 63-tap Symmetric FIRs
Flexible Output Formats
Serial and Parallel Output Ports
JTAG Boundary Scan
8-bit Microprocessor Interface
128 pin PQFP
APPLICATIONS
•
•
•
•
•
Cellular Basestations
GSM / GPRS / EDGE / GSM Phase 2 Receivers
Satellite Receivers
Wireless Local Loop Receivers
Digital Communications
KEY SPECIFICATIONS
•
•
•
•
•
•
•
Internal ADC Resolution: 12 Bits
Sample Rate: 65 MSPS
SNR (fIN = 250MHz, 11-bit, Nyquist): 62 dBFS
(typ)
SNR (fIN = 250MHz, 200kHz): 83 dBFS (typ)
SFDR (fIN = 250MHz, 11-bit, Nyquist): 68 dBFS
(typ)
Full Power Bandwidth: 650 MHz (typ)
Power Consumption: (65MSPS) 560 mW (typ)
The LM97593 offers high dynamic range digital tuning
and filtering based on hard-wired digital signal
processing (DSP) technology. Each channel has
independent tuning, phase offset, filter coefficients,
and gain settings. Channel filtering is performed by a
series of three filters. The first is a 4-stage Cascaded
Integrator Comb (CIC) filter with a programmable
decimation ratio from 8 to 2048. Next there are two
symmetric FIR filters, a 21-tap and a 63-tap, both with
independent programmable coefficients. The first FIR
filter decimates the data by 2, the second FIR
decimates by either 2 or 4. Channel filter bandwidth
at 52MSPS ranges from ±650kHz down to ±1.3kHz.
At 65MSPS, the maximum bandwidth increases to
±812kHz.
The LM97593’s AGC controller monitors the ADC
output and controls the ADC input signal level by
adjusting the DVGA setting. AGC threshold,
deadband+hysteresis, and the loop time constant are
user defined. Total dynamic range of greater than
123dB full-scale signal to noise in a 200kHz
bandwidth can be achieved with the Diversity
Receiver Chipset.
1
2
Please be aware that an important notice concerning availability, standard warranty, and use in critical applications of
Texas Instruments semiconductor products and disclaimers thereto appears at the end of this data sheet.
All trademarks are the property of their respective owners.
PRODUCTION DATA information is current as of publication date.
Products conform to specifications per the terms of the Texas
Instruments standard warranty. Production processing does not
necessarily include testing of all parameters.
Copyright © 2007–2013, Texas Instruments Incorporated
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
Block Diagram 1
LM97593
CLC5526
(x2)
SCK_IN
LC
12
SerialOutA/B
ADC
DVGA
IF A
SerialOutB
SCK
8
Dual Digital
Tuner/AGC
12
ADC
DVGA
IF B
SFS
RDY
ParallelOutput[15:0]
LC
ParallelOutputEnable
CLK
ParallelSelect[2:0]
Figure 1. Diversity Receiver Chipset Block Diagram
AGND
VD
NC
VD
DGND
DGND
REFSEL/DCS
DRGND
1
2
3
4
5
6
7
VA
VINAAGND
VA
AGND
8
9
10
11
12
VINA+
13
VCOMA
14
VRNA
VA
AGND
VRPA
15
16
17
18
19
VRNB
VRPB
VCOMB
VREF
AGND
20
21
22
23
24
VINB+
VA
25
VA
AGND
VD
PD
VINB-
26
27
28
29
30
VD
DGND
NC
NC
DGND
31
32
33
34
35
36
53
54
117
116
115
114
113
NC
VDR
AGAIN[2]
AGAIN[1]
AGAIN[0]
~ASTROBE
VD18
~SCAN_EN
~TRST
D18GND
TCK
TMS
TDI
TDO
VD18
POUT_SEL[0]
POUT_SEL[1]
POUT_SEL[2]
~POUT_EN
DRGND
POUT[0]
POUT[1]
VDR
POUT[2]
POUT[3]
POUT[4]
102
D18GND
VDR
101
100
NC
SCK_IN
99
98
DRGND
97
POUT[5]
96
POUT[6]
95
POUT[7]
POUT[8]
94
93
VDR
POUT[9]
POUT[10]
POUT[11]
DRGND
POUT[12]
POUT[13]
VD18
POUT[14]
DRGND
POUT[15]
AOUT
SFS
VDR
SCK
BOUT
RDY
D18GND
92
104
103
91
105
63
64
90
62
89
106
88
61
87
107
86
60
85
108
84
59
83
109
82
110
58
81
57
80
111
79
56
78
112
77
55
65
VDR
52
D18GND
D[6]
51
76
D[7]
LM97593VH
Dual Digital Tuner / AGC / ADC
(Top View)
50
75
DRGND
118
VD18
D[0]
~CE
49
74
~RD
119
73
~WR
120
48
D[1]
A[0]
47
72
A[1]
121
D[2]
A[2]
122
46
D[3]
A[3]
45
71
A[4]
123
D[4]
D18GND
A[5]
44
70
VD18
A[6]
124
D[5]
A[7]
43
69
~SI
DRGND
125
68
~MR
42
67
D18GND
126
NC
~BSTROBE
127
41
DRGND
BGAIN[0]
128
40
66
BGAIN[1]
39
D18GND
VDR
BGAIN[2]
37
38
VDR
CK
Connection Diagram
Figure 2. LM97593VH PQFP Pinout
2
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
Block Diagram 2
AGAIN[2:0]
ASTROBE
12
VINA
ADC_A
MUX
A
14
Channel A
Tuning,
Channel Filters, and
AGC
Output
Formatter
SCK_IN
AOUT/BOUT
BOUT
SCK
SFS
RDY
VINB
12
MUX
B
ADC_B
14
Channel B
Tuning,
Channel Filters, and
AGC
POUT[15:0]
PSEL[2:0]
POUT_EN
BSTROBE
BGAIN[2:0]
TEST_REG
Figure 3. LM97593 Block Diagram
PIN DESCRIPTIONS AND EQUIVALENT CIRCUITS
Pin No.
Symbol
Equivalent Circuit
Description
13
27
VINA−
VINB−
Analog Input
Negative differential input signal for the 'A' channel
Negative differential input signal for the 'B' channel
14
26
VINA+
VINB+
Analog Input
Positive differential input signal for the 'A' channel
Positive differential input signal for the 'B' channel
21
VREF
Control / Analog Input
Reference Select Pin / External Reference Voltage Input
Input differential full scale swing = 2 * VREF
VREF = VA to VA - 0.3V: Reference Voltage = 1.0 V (Internal)
VREF = 0.8V to 1.5V:
Reference Voltage = VREF (External)
15
24
VCOMA
VCOMB
Analog Output
Common Mode reference voltage for the 'A' channel
Common Mode reference voltage for the 'B' channel
These pins may be loaded to 1 mA for use as temperature stable 1.5V
references.
16
23
VRPA
VRPB
Analog Output
Upper reference voltage for the 'A' channel
Upper reference voltage for the 'B' channel
17
22
VRNA
VRNB
Analog Output
Lower reference voltage for the 'A' channel
Lower reference voltage for the 'B' channel
REFSEL/DCS
Control Input
This is a three-state pin. VCOM = VCOMA or VCOMB.
REFSEL/DCS = AGND: the internal reference is enabled and duty cycle
correction is applied to the ADC input clock (CK).
REFSEL/DCS = VCOM: the internal reference is enabled and no duty
cycle correction is applied to the ADC input clock (CK).
REFSEL/DCS = VA: DCS is on, the internal reference is disabled. Apply
A 0.8-1.2V external reference to the VREF pin.
30
PD
Input
POWER DOWN, when high both ADCs are powered down, when low,
both ADCs are enabled
45
MR
Input
MASTER RESET, Active low
Resets all registers within the chip. ASTROBE and BSTROBE are
asserted during MR.
ANALOG I/O
8
DIGITAL I/O
82
78
AOUT
BOUT
Output
SERIAL OUTPUT DATA, Active high
The 2's complement serial output data is transmitted on these pins, MSB
first. The output bits change on the rising edge of SCK (falling edge if
SCK_POL=1) and should be captured on the falling edge of SCK (rising if
SCK_POL=1). These pins are tri-stated at power up and are enabled by
the SOUT_EN control register bit. See Figure 13 and Figure 75 timing
diagrams. In Debug Mode AOUT=DEBUG[1], BOUT=DEBUG[0].
127:125
40:42
AGAIN[2:0]
BGAIN[2:0]
Output
OUTPUT DATA TO DVGA, Active high
3 bit bus that sets the gain of the DVGA determined by the AGC circuit.
124
43
ASTROBE
BSTROBE
Output
DVGA STROBE, Active low
Strobes the data into the DVGA. See Figure 7 and Figure 82 timing
diagrams.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
3
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
PIN DESCRIPTIONS AND EQUIVALENT CIRCUITS (continued)
Pin No.
80
Symbol
SCK
99
SCK_IN
81
SFS
84, 86:88,
90, 91,
93:97,
104:106,
108, 109
POUT[15:0]
Equivalent Circuit
Description
Output
SERIAL DATA CLOCK, Active high or low
The serial data is clocked out of the chip by this clock. The active edge of
the clock is user programmable. This pin is tri-stated at power up and is
enabled by the SOUT_EN control register bit. See Figure 13 and
Figure 75 timing diagrams. In Debug Mode outputs an appropiate clock
for the debug data. If RATE=0 the input CK duty cycle will be reflected to
SCK.
Input
SERIAL DATA CLOCK INPUT, Active high or low
Data bits from a serial daisy-chain slave are clocked into a serial daisychain master on the falling edge of SCK_IN (rising if SCK_POL=1 on the
slave). Tie low if not used.
Output
SERIAL FRAME STROBE, Active high or low
The serial word strobe. This strobe delineates the words within the serial
output streams. This strobe is a pulse at the beginning of each serial
word (PACKED=0) or each serial word I/Q pair (PACKED=1). The polarity
of this signal is user programmable. This pin is tri-stated at power up and
is enabled by the SOUT_EN control register bit. See Figure 13 and
Figure 75 timing diagrams. In Debug Mode SFS=DEBUG[2].
Output
PARALLEL OUTPUT DATA, Active high
The output data is transmitted on these pins in parallel format. The
POUT_SEL[2:0] pins select one of eight 16-bit output words. The
POUT_EN pin enables these outputs. POUT[15] is the MSB. In Debug
Mode POUT[15:0]=DEBUG[19:4].
112:114
POUT_SEL[2:0]
Input
PARALLEL OUTPUT DATA SELECT, Active high
The 16-bit output word is selected with these 3 pins. Not used in Debug
Mode. For a serial daisy-chain master, POUT_SEL[2:0] become inputs
from the slave: POUT_SEL[2]=SFSSLAVE, POUT_SEL[1]=BOUTSLAVE,
and POUT_SEL[0]=AOUTSLAVE. Tie low if not used.
111
POUT_EN
Input
PARALLEL OUTPUT ENABLE. Active low
This pin enables the chip to output the selected output word on the
POUT[15:0] pins. Not used in Debug Mode. Tie high if not used.
77
RDY
Output
READY FLAG, Active high or low
The chip asserts this signal to identify the beginning of an output sample
period (OSP). The polarity of this signal is user programmable. This
signal is typically used as an interrupt to a DSP chip, but can also be
used as a start pulse to dedicated circuitry. This pin is active regardless
of the state of SOUT_EN. In Debug Mode RDY=DEBUG[3].
37
CK
Input
INPUT CLOCK. Active high
The clock input to the chip. The The VINA and VINB analog input signals
are sampled on the rising edge of this signal. SI is clocked into the chip
on the rising edge of CK.
46
SI
Input
SYNC IN. Active low
The sync input to the chip. The decimation counters, dither, and NCO
phase can be synchronized by SI. This sync is clocked into the chip on
the rising edge of CK. Tie this pin high if external sync is not required. All
sample data is flushed by SI. To properly initialize the DVGA ASTROBE
and BSTROBE are asserted during SI.
62, 63,
69:73, 75
D[7:0]
Input/Output
DATA BUS. Active high
This is the 8 bit control data I/O bus. Control register data is loaded into
the chip or read from the chip through these pins. The chip will only drive
output data on these pins when CE is low, RD is low, and WR is high.
48, 50, 52:57 A[7:0]
Input
ADDRESS BUS. Active high
These pins are used to address the control registers within the chip. Each
of the control registers within the chip are assigned a unique address. A
control register can be written to or read from by setting A[7:0] to the
register’s address and setting CE, RD, and WR appropriately.
59
RD
Input
READ ENABLE. Active low
This pin enables the chip to output the contents of the selected register
on the D[7:0] pins when CE is also low.
58
WR
Input
WRITE ENABLE. Active low
This pin enables the chip to write the value on the D[7:0] pins into the
selected register when CE is also low. This pin can also function as RD/
CE if RD is held low. See Figure 15 for details.
4
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
PIN DESCRIPTIONS AND EQUIVALENT CIRCUITS (continued)
Pin No.
Symbol
Equivalent Circuit
Description
60
CE
Input
CHIP ENABLE. Active low
This control strobe enables the read or write operation. The contents of
the register selected by A[7:0] will be output on D[7:0] when RD is low
and CE is low. If WR is low and CE is low, then the selected register will
be loaded with the contents of D[7:0].
116
TDO
Output
TEST DATA OUT. Active high
117
TDI
Input
TEST DATA IN. Active high with pull-up
118
TMS
Input
TEST MODE SELECT. Active high with pull-up
119
TCK
Input
TEST CLOCK. Active high. Tie low if JTAG is not used.
121
TRST
Input
TEST RESET. Active low with pull-up
Asynchronous reset for TAP controller. Tie low or to MR if JTAG is not
used.
122
SCAN_EN
Input
SCAN ENABLE. Active low with pull-up
Enables access to internal scan registers. Tie high. Used for
manufacturing test only!
DDC Output Driver Power
I/O Power Supply, 3.3V nominal. Quantity 8.
1, 47, 61, 68,
83, 89, 98,
DRGND
110
DDC Output Driver Ground
I/O Ground Return. Quantity 8.
49, 74, 85,
115, 123
VD18
DDC Core Power
DSP Digital Core Power Supply, 1.8V nominal. Quantity 5.
49, 74, 85,
115, 123
VD18
DDC Core Power
DSP Digital Core Power Supply, 1.8V nominal. Quantity 5.
44, 51, 65,
66, 76, 103,
120
D18GND
DDC Core Ground
DSP Digital Core Ground Return. Quantity 7.
4, 6, 31, 34
VD
ADC Digital Power
ADC Digital Logic Power Supply, 3.3V nominal. Quantity 4.
5, 7, 32, 33
DGND
ADC Digital Ground
ADC Digital Logic Ground Return. Quantity 4.
Digital Power Supplies
38, 39, 64,
79, 92, 102,
107, 128
VDR
Analog Power Supplies
10, 11, 19,
25, 29
VA
ADC Analog Power
ADC Analog Power Supply, 3.3V nominal. Quantity 5.
2, 9, 12, 18,
20, 28
AGND
ADC Analog Ground
ADC Analog Ground Return. Quantity 6.
NC
Not Connected. These pins should be left floating.
Unconnected Pins
3, 35, 36, 67,
NC
100, 101
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
5
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
These devices have limited built-in ESD protection. The leads should be shorted together or the device placed in conductive foam
during storage or handling to prevent electrostatic damage to the MOS gates.
Absolute Maximum Ratings (1) (2) (3)
ADC Analog, Digital and IO Supply Voltages (VA, VD and VDR)
−0.3V to 4.2V
Difference between VA, VD, and VDR
≤ 100 mV
Positive Core Supply Voltage (VD18)
−0.3V to 2.35V
−0.3V to (VDR +0.3V)
Voltage on Any Input or Output Pin
(Not to exceed 4.2V)
Input Current at Any Pin other than Supply Pins
Package Input Current
(4)
±5 mA
(4)
±50 mA
Max Junction Temp (TJ)
+125°C
Thermal Resistance (θJA)
39°C/W
Package Dissipation at TA = 25°C (5)
ESD Susceptibility (6)
3.2W
Human Body Model (1.5kΩ, 100pF)
Machine Model (0Ω, 200pF)
Charge Device Model
(3)
(4)
(5)
(6)
200 V
750 V
−65°C to +150°C
Storage Temperature
(1)
(2)
2000 V
All voltages are measured with respect to GND = AGND = DGND = DRGND = 0V, unless otherwise specified.
Absolute Maximum Ratings indicate limits beyond which damage to the device may occur. Operating Ratings indicate conditions for
which the device is ensured to be functional, but do not ensure specific performance limits. For ensured specifications and test
conditions, see the Electrical Characteristics. The ensured specifications apply only for the test conditions listed. Some performance
characteristics may degrade when the device is not operated under the listed test conditions. Operation of the device beyond the
maximum Operating Ratings is not recommended.
If Military/Aerospace specified devices are required, please contact the Texas Instruments Sales Office/Distributors for availability and
specifications.
When the input voltage at any pin exceeds the power supplies (that is, VIN < AGND, or VIN > VA), the current at that pin should be
limited to ±25 mA. The ±50 mA maximum package input current rating limits the number of pins that can safely exceed the power
supplies with an input current of ±25 mA to two.
The maximum allowable power dissipation is dictated by TJ,max, the junction-to-ambient thermal resistance, (θJA), and the ambient
temperature, (TA), and can be calculated using the formula PD,max = (TJ,max - TA )/θJA. The values for maximum power dissipation listed
above will be reached only when the device is operated in a severe fault condition (e.g. when input or output pins are driven beyond the
power supply voltages, or the power supply polarity is reversed). Such conditions should always be avoided.
Human Body Model is 100 pF discharged through a 1.5 kΩ resistor. Machine Model is 220 pF discharged through 0 Ω.
Operating Ratings (1) (2)
Soldering process must comply with Texas Instrument's Reflow Temperature Profile specifications. Refer to http://www.ti.com/packaging (3)
−40°C ≤ TA ≤ +85°C
Operating Temperature Range
ADC Analog, Digital and IO Supply Voltages (VA, VD and VDR)
+3.0V to +3.6V
Digital Core Supply Voltage (VD18)
+1.6V to +2.0V
Difference Between AGND, DGND, DRGND and D18GND
≤ 100 mV
Voltage on Any Input or Output Pin
0V to +3.3V
VCM
1.0V to 2.0V
Clock Duty Cycle
30% to 70 %
(1)
(2)
(3)
Absolute Maximum Ratings indicate limits beyond which damage to the device may occur. Operating Ratings indicate conditions for
which the device is ensured to be functional, but do not ensure specific performance limits. For ensured specifications and test
conditions, see the Electrical Characteristics. The ensured specifications apply only for the test conditions listed. Some performance
characteristics may degrade when the device is not operated under the listed test conditions. Operation of the device beyond the
maximum Operating Ratings is not recommended.
All voltages are measured with respect to GND = AGND = DGND = DRGND = 0V, unless otherwise specified.
Reflow temperature profiles are different for lead-free and non-lead-free packages.
Reliability Information
Transistor Count
6
1.3 million
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
LM97593 Electrical Characteristics
Unless otherwise specified, the following specifications apply: AGND = DGND = DRGND = D18GND = 0V, VA = VD = VDR =
+3.3V, VD18 = +1.8V, Internal VREF = +1.0V, fCLK = 65 MHz, VCM = VCOM, tR = tF = 1 ns, CL = 5 pF/pin. The ADC’s 11 most
significant bits observed at the mixer output debug tap with NCO = 0Hz. Typical values are for TA = 25°C. Boldface limits
apply for TMIN ≤ TA ≤ TMAX. All other limits apply for TA = 25°C. (1) (2) (3)
Symbol
Parameter
Conditions
Typical
(4)
Limits
Units
(Limits)
11
Bits (min)
2
LSB (max)
STATIC CONVERTER CHARACTERISTICS
Resolution with No Missing Codes
Integral Non Linearity (5)
INL
Ramp, End Point
±0.7
DNL
Differential Non Linearity
Ramp, End Point
±0.3
VOFF
Offset Error
−40°C to +85°C
-4.1
-2
LSB (min)
0.85
LSB (max)
-0.85
LSB (min)
LSB
REFERENCE AND ANALOG INPUT CHARACTERISTICS
1.0
V (min)
2.0
V (max)
VCM
Common Mode Input Voltage
1.5
VCOMA
VCOMB
Reference Output Voltage
1.5
V
CIN
VIN Input Capacitance (each pin to GND)
(VIN= 1.5Vdc ±0.5V) (6)
CK LOW
8
pF
CK HIGH
7
VREF
External Reference Voltage (7)
1.0
Reference Input Resistance
1
pF
0.8
V (min)
1.2
V (max)
MΩ
(1)
The inputs are protected as shown below. Input voltage magnitudes above VA or below GND will not damage this device, provided
current is limited per Note (4).
(2)
(3)
(4)
To ensure accuracy, it is required that |VA–VD| ≤ 100 mV and separate bypass capacitors are used at each power supply pin.
With the test condition for VREF = +1.0V (2VP-P differential input), the 12-Bit LSB is 488 µV.
Typical figures are at TA = 25°C and represent most likely parametric norms at the time of product characterization. The typical
specifications are not ensured. Test Limits are specified to TI's AOQL (Average Outgoing Quality Level).
Integral Non Linearity is defined as the deviation of the analog value, expressed in LSBs, from the straight line that passes through
positive and negative full-scale.
The input capacitance is the sum of the package/pin capacitance and the sample and hold circuit capacitance.
Optimum performance will be obtained by keeping the reference input in the 0.8V to 1.2V range. The LM4051CIM3-ADJ (SOT-23
package) is recommended for external reference applications.
(5)
(6)
(7)
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
7
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
LM97593 Electrical Characteristics (continued)
Unless otherwise specified, the following specifications apply: AGND = DGND = DRGND = D18GND = 0V, VA = VD = VDR =
+3.3V, VD18 = +1.8V, Internal VREF = +1.0V, fCLK = 65 MHz, VCM = VCOM, tR = tF = 1 ns, CL = 5 pF/pin. The ADC’s 11 most
significant bits observed at the mixer output debug tap with NCO = 0Hz. Typical values are for TA = 25°C. Boldface limits
apply for TMIN ≤ TA ≤ TMAX. All other limits apply for TA = 25°C.(1)(2)(3)
Symbol
Parameter
Typical
Conditions
(4)
Limits
Units
(Limits)
DYNAMIC CONVERTER CHARACTERISTICS
FPBW
Full Power Bandwidth
SNR
Signal-to-Noise Ratio
SINAD
ENOB
THD
H2
fIN = 20MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -9dBFS
fIN = 20MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -9dBFS
Signal-to-Noise and Distortion
fIN = 20MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -9dBFS
Effective Number of Bits
(Relative to Full Scale)
fIN = 20MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -9dBFS
Total Harmonic Distortion
fIN = 20MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -9dBFS
Second Harmonic Distortion
H3
fIN = 20MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -9dBFS
Third Harmonic Distortion
SFDR
IMD
fIN = 20MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -3dBFS
fIN = 249MHz, VIN = -9dBFS
Spurious Free Dynamic Range
650
MHz
66.2
dBFS
63.7
63.9
dBFS
62.2
dBFS (min)
62.8
dBF
62.0
dBFS
63.4
60.4
10.6
dBFS (min)
Bits
10.0
Bits
10.3
Bits (min)
-77.1
dBc
-57.9
-64.6
dBc
-54.1
-82.7
dBc
-59.9
-67.3
dBc (max)
dBc
-59.3
dBc (max)
-91.7
dBc
-69.0
dBc
-72.5
-56.8
dBc (min)
79.7
dBFS
68.0
dBFS (min)
76.3
67.0
dBFS
f1IN = 246MHz, VIN = -15dBFS
f2IN = 250MHz, VIN = -15dBFS
(f1IN + f2IN = -9dBFS)
-77.5
-3 dBFS reference, -50dBFS target
±1.3
Channel - Channel Offset Match
10 MHz -1dBFS driven
±0.2
%FS
Channel - Channel Gain Match
10 MHz -1dBFS driven
±0.4
%FS
249MHz -3dBFS driven channel, 50Ω
termination measured channel
60(+42)
CIC Decimation = 8, NCO = 11.1MHz
(248.9MHz @65MSPS aliases to 11.1
MHz), fIN = 249MHz at -3dBFS, signal
observed at F1 In Debug tap
68.6
66.1
dBFS
68.5
66.1
dBFS
75.1
70.3
dBc
Intermodulation Distortion
Dynamic Gain Error
dBFS
±2
dB
INTERCHANNEL CHARACTERISTICS
Crosstalk (AGC fixed at 0dB gain, with AGC
operating the crosstalk will improve at the
output)
dBc
(8)
CIC OUTPUT CHARACTERISTICS
SNR
Signal-to-Noise Ratio
SINAD
Signal-to-Noise and Distortion
SFDR
Spurious Free Dynamic Range
DDC OUTPUT CHARACTERISTICS
SNR
Signal-to-Noise Ratio
SINAD
Signal-to-Noise and Distortion
SFDR
Spurious Free Dynamic Range
(8)
8
GSM Filter set, 200kHz channel BW,
NCO = 11.1MHz (248.9MHz
@52MSPS aliases to 11.1 MHz), fS =
52MSPS, fIN = 249MHz at -9dBFS
82 (+42)
(8)
dBFS
73
dBc
90
dBc
(+x) indicates the additional dynamic range provided by the AGC. The DVGA in front of the LM97593 provides 42 dB of gain adjustment.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
LM97593 Electrical Characteristics (continued)
Unless otherwise specified, the following specifications apply: AGND = DGND = DRGND = D18GND = 0V, VA = VD = VDR =
+3.3V, VD18 = +1.8V, Internal VREF = +1.0V, fCLK = 65 MHz, VCM = VCOM, tR = tF = 1 ns, CL = 5 pF/pin. The ADC’s 11 most
significant bits observed at the mixer output debug tap with NCO = 0Hz. Typical values are for TA = 25°C. Boldface limits
apply for TMIN ≤ TA ≤ TMAX. All other limits apply for TA = 25°C.(1)(2)(3)
Symbol
Parameter
SNR
Signal-to-Noise Ratio
SINAD
Signal-to-Noise and Distortion
SFDR
Spurious Free Dynamic Range
SNR
Signal-to-Noise Ratio
SINAD
Signal-to-Noise and Distortion
SFDR
Spurious Free Dynamic Range
SNR
Signal-to-Noise Ratio
SINAD
Signal-to-Noise and Distortion
SFDR
Spurious Free Dynamic Range
(9)
Typical
Conditions
(4)
GSM Filter set, 200kHz channel BW,
NCO = 11.1MHz (248.9MHz
@52MSPS aliases to 11.1 MHz), fS =
52MSPS, fIN = 249MHz at -3dBFS
GSM Filter set, 200kHz channel BW,
NCO = 11.1MHz (248.9MHz
@65MSPS aliases to 11.1 MHz), fS =
65MSPS, fIN = 249MHz at -9dBFS
GSM Filter set, 200kHz channel BW,
NCO = 11.1MHz (248.9MHz
@65MSPS aliases to 11.1 MHz), fS =
65MSPS, fIN = 249MHz at -3dBFS
Limits
76 (+42)
Units
(Limits)
(9)
dBFS
74
dBc
90
dBc
79 (+42)
(9)
dBFS
71
dBc
81
dBc
74 (+42)
(9)
dBFS
71
dBc
80
dBc
(+x) indicates the additional dynamic range provided by the AGC. The DVGA in front of the LM97593 provides 42 dB of gain adjustment.
DC and Logic Electrical Characteristics
Unless otherwise specified, the following specifications apply: AGND = DGND = DRGND = D18GND = 0V, VA = VD = VDR =
+3.3V, VD18 = +1.8V, Internal VREF = +1.0V, fCLK = 65 MHz, VCM = VCOM, tR = tF = TBD ns, CL = 5 pF/pin. CIC Decimation = 48,
F2 Decimation = 2. Typical values are for TA = 25°C. Boldface limits apply for TMIN ≤ TA ≤ TMAX. All other limits apply for TA
= 25°C.
Symbol
Parameter
Typical
Conditions
(1)
Limits
Units
(Limits)
VIL
Voltage input low
0.7
V (max)
VIH
Voltage input high
2.3
V (min)
IOZ
Input current
20
µA
VOL
Voltage output low (IOL = 7mA)
0.4
V (max)
VOH
Voltage output high (IOH = -7mA)
2.4
V (min)
CIN
Input capacitance
5.0
pF
121
mA (max)
POWER SUPPLY CHARACTERISTICS
IA
ADC Analog Supply Current
65MSPS
96
IA
ADC Analog Supply Current
52MSPS
84
ID
ADC Digital Supply Current
65MSPS
24
ID
ADC Digital Supply Current
52MSPS
20
IDR
Digital Output Supply Current
(2)
65MSPS
14
IDR
Digital Output Supply Current
(2)
52MSPS
10
ID18
Digital Core Supply Current
65MSPS
67
ID18
Digital Core Supply Current
52MSPS
53
PD65
Total Power Dissipation
GSM Set, 65MSPS
560
PD52
Total Power Dissipation
GSM Set, 52MSPS
485
PSRR
(1)
(2)
Power Supply Rejection Ratio
Rejection of Full-Scale Error with VA =
3.0V vs. 3.6V
mA
28
mA (max)
18
mA (max)
mA
mA
78
mA (max)
793
mW (max)
mA
mW
dB
Typical figures are at TA = 25°C and represent most likely parametric norms at the time of product characterization. The typical
specifications are not ensured. Test Limits are specified to TI's AOQL (Average Outgoing Quality Level).
IDR is the current consumed by the switching of the output drivers and is primarily determined by load capacitance on the output pins,
the supply voltage, VDR, and the rate at which the outputs are switching (which is signal dependent). IDR=VDR(C0 x f0 + C1 x f1 +....C11 x
f11) where VDR is the output driver power supply voltage, Cn is total capacitance on the output pin, and fn is the average frequency at
which that pin is toggling.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
9
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
AC Electrical Characteristics
Unless otherwise specified, the following specifications apply: AGND = DGND = DRGND = D18GND = 0V, VA = VD = VDR =
+3.3V (±10%), VD18 = +1.8V (±10%), Internal VREF = +1.0V, fCLK = 65 MHz, VCM = VCOM, tR = tF = 1 ns, CL = 5 pF/pin. CIC
Decimation = 48, F2 Decimation = 2. Typical values are for TA = 25°C. Boldface limits apply for TMIN ≤ TA ≤ TMAX. All other
limits apply for TA = 25°C. (1)
Symbol
Parameter (CL=50pF)
Min
Typical
(2)
Max
Units
MHz
Clock Input
FCK
Clock (CK) Frequency (Figure 6)
20
65
tCKDC
CK duty cycle, DCS off (Figure 6)
40
60
%
tRF
CK rise and fall times (VIL to VIH) (Figure 6)
2
ns
NCO Tuning Resolution
0.02
Hz
NCO Phase Resolution
0.005
o
Control Interface
tMRA
MR Active Time (Figure 4)
4
CK periods
tMRIC
MR Inactive to first Control Port Access (Figure 4)
10
CK periods
tMRSU
MR Setup Time to CK (Figure 4)
6
ns
tMRH
MR Hold Time from CK (Figure 4)
2
ns
tSISU
SI Setup Time to CK (Figure 5)
6
ns
tSIH
SI Hold Time from CK (Figure 5)
2
ns
tSIW
SI Pulse Width (Figure 5)
4
CK periods
DVGA Interface
tSTIW
A|BSTROBE Inactive Pulse Width (Figure 7)
tGSTB
A|BGAIN setup before A|BSTROBE (Figure 7)
2
CK periods
6
ns
Parallel Output Interface
tOENV
POUT_EN Active to POUT[15:0] Valid (Figure 9)
12
ns
tOENT
POUT_EN Inactive to POUT[15:0] Tri-State (Figure 9)
10
ns
tSELV
PSEL[2:0] to POUT[15:0] Valid (Figure 10)
13
ns
tPOV
RDY to POUT[15:0] New Value Valid (Note 5) (Figure 11)
7
ns
tDBG
SCK to POUT[15:0], RDY, SFS, AOUT, BOUT Valid (Figure 12)
4
ns
ns
Serial Interface
tSFSV
SCK to SFS Valid (Note 3) (Figure 13)
-2
1.6
3.5
tOV
SCK to A|BOUT Valid (Note 4) (Figure 13)
-2
1.7
3.5
tRDYW
RDY Pulse Width (Figure 13)
tDCMSU
PSEL[2:0] Setup Time to SCK_IN (Figure 8)
tDCMH
PSEL[2:0] Hold Time from SCK_IN (Figure 8)
tRDYV
SCK to RDY valid (Figure 13)
-3
1.8
ns
2
CK periods
3
1.4
ns
0.5
-0.9
ns
4
ns
JTAG Interface
tJPCO
Propagation Delay TCK to TDO (Figure 14)
25
ns
tJSCO
Propagation Delay TCK to Data Out (Figure 14)
35
ns
tJPDZ
Disable Time TCK to TDO (Figure 14)
25
ns
tJSDZ
Disable Time TCK to Data Out (Figure 14)
35
ns
tJPEN
Enable Time TCK to TDO (Figure 14)
0
25
ns
tJSEN
Enable Time TCK to Data Out (Figure 14)
0
35
ns
tJSSU
Setup Time Data to TCK (Figure 14)
10
ns
tJPSU
Setup Time TDI, TMS to TCK (Figure 14)
10
ns
tJSH
Hold Time Data to TCK (Figure 14)
45
ns
tJPH
Hold Time TCK to TDI, TMS (Figure 14)
45
ns
(1)
(2)
10
Timing specifications are tested at TTL logic levels, VIL = 0.4V for a falling edge and VIH = 2.4V for a rising edge.
Typical figures are at TA = 25°C and represent most likely parametric norms at the time of product characterization. The typical
specifications are not ensured. Test Limits are specified to TI's AOQL (Average Outgoing Quality Level).
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
AC Electrical Characteristics (continued)
Unless otherwise specified, the following specifications apply: AGND = DGND = DRGND = D18GND = 0V, VA = VD = VDR =
+3.3V (±10%), VD18 = +1.8V (±10%), Internal VREF = +1.0V, fCLK = 65 MHz, VCM = VCOM, tR = tF = 1 ns, CL = 5 pF/pin. CIC
Decimation = 48, F2 Decimation = 2. Typical values are for TA = 25°C. Boldface limits apply for TMIN ≤ TA ≤ TMAX. All other
limits apply for TA = 25°C.(1)
Symbol
Parameter (CL=50pF)
Min
Typical
Max
(2)
Units
tJCH
TCK Pulse Width High (Figure 14)
50
ns
tJCL
TCK Pulse Width Low (Figure 14)
40
ns
JTAGFMAX
TCK Maximum Frequency (Figure 14)
10
MHz
Microprocessor Interface
tCSU
Control Setup before the controlling signal goes low (Figure 15)
5
ns
tCHD
tCSPW
Control hold after the controlling signal goes high (Figure 15)
5
ns
Controlling strobe pulse width (Write) (Figure 15)
30
tCDLY
Control output delay controlling signal low to D (Read) (Figure 15)
30
ns
tCZ
Control tri-state delay after controlling signal high (Figure 15)
20
ns
ns
DDC Timing Diagrams
CK
tMRH
tMRSU
tMRA
MR
tMRIC
RD
or WR
Figure 4. LM97593 Master Reset Timing
CK
tSIH
tSISU
tSIW
SI
Figure 5. LM97593 Synchronization Input (SI) Timing
1
FCK
tCKDC
tCKDC
tRF
VIL
CK
VIH
Figure 6. LM97593 Clock Timing
tSTIW
A|BSTROBE
tGSTB
A|BGAIN[2:0]
Figure 7. LM97593 DVGA Interface Timing
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
11
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
Figure 8. LM97593 Dual Chip Mode Timing
POUT_EN
tOENV
tOENT
POUT[15:0]
Figure 9. LM97593 Parallel Output Enable Timing
n
POUT_SEL[2:0]
n+1
n+2
tSELV
POUT[15:0]
output (n)
output (n+1)
output (n+2)
Figure 10. LM97593 Parallel Output Select Timing
RDY
RDY_POL = 0
RDY
RDY_POL = 1
tPOV
POUT[15:0]
old output
new output
Figure 11. LM97593 Parallel Output Data Ready Timing
SCK_IN
tDBG
POUT[15:0],
AOUT, BOUT,
RDY and SFS
Figure 12. LM97593 Debug Mode Timing
12
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
Figure 13. LM97593 Serial Port Timing
TCK
tJPCO
tJPDZ
tJPEN
TDO
tJCH
tJCL
TCK
tJPSU
tJPH
TDI, TDS
TCK
tJSCO
tJSDZ
tJSEN
Data Out
TCK
tJSSU
tJSH
Data In
Figure 14. LM97593 JTAG Port Timing
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
13
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
Figure 15. LM97593 Control I/O Timing
14
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
ADC Typical Performance Characteristics DNL, INL
Unless otherwise specified, the following specifications apply for AGND = DGND = D18GND = DRGND = 0V, VA = VD = VDR
= +3.3V, VD18 = +1.8V, PD = 0V, Internal VREF = +1.0V, fCLK = 65 MHz, fIN = 12 MHz, AIN = 0dBFS, CL = 10 pF/pin, Duty Cycle
Stabilizer On. Boldface limits apply for TJ = TMIN to TMAX: all other limits TJ = 25°C
DNL
INL
Figure 16.
Figure 17.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
15
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
ADC Typical Performance Characteristics
Unless otherwise specified, the following specifications apply for AGND = DGND = D18GND = DRGND = 0V, VA = VD = VDR
= +3.3V, VD18 = +1.8V, PD = 0V, Internal VREF = +1.0V, fCLK = 65 MHz, fIN = 249 MHz, AIN = -9dBFS, CL = 10 pF/pin, Duty
Cycle Stabilizer On. Boldface limits apply for TJ = TMIN to TMAX: all other limits TJ = 25°C
16
SNR, SINAD, SFDR vs. VSUPPLY
Distortion vs. VSUPPLY
Figure 18.
Figure 19.
Gain Tracking Error vs. VSUPPLY
SNR, SINAD, SFDR vs. Temperature
Figure 20.
Figure 21.
Distortion vs. Temperature
Gain Tracking Error vs. Temperature
Figure 22.
Figure 23.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
ADC Typical Performance Characteristics (continued)
Unless otherwise specified, the following specifications apply for AGND = DGND = D18GND = DRGND = 0V, VA = VD = VDR
= +3.3V, VD18 = +1.8V, PD = 0V, Internal VREF = +1.0V, fCLK = 65 MHz, fIN = 249 MHz, AIN = -9dBFS, CL = 10 pF/pin, Duty
Cycle Stabilizer On. Boldface limits apply for TJ = TMIN to TMAX: all other limits TJ = 25°C
SNR, SINAD, SFDR vs. fIN
(fCLK = 52MSPS, AIN = -9dBFS)
Distortion vs. fIN
(fCLK = 52MSPS, AIN = -9dBFS)
Figure 24.
Figure 25.
SNR, SINAD, SFDR vs. fIN
(fCLK = 52MSPS, AIN = -3dBFS)
Distortion vs. fIN
(fCLK = 52MSPS, AIN = -3dBFS)
Figure 26.
Figure 27.
SNR, SINAD, SFDR vs. fIN
(fCLK = 65MSPS, AIN = -9dBFS)
Distortion vs. fIN
(fCLK = 65MSPS, AIN = -9dBFS)
Figure 28.
Figure 29.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
17
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
ADC Typical Performance Characteristics (continued)
Unless otherwise specified, the following specifications apply for AGND = DGND = D18GND = DRGND = 0V, VA = VD = VDR
= +3.3V, VD18 = +1.8V, PD = 0V, Internal VREF = +1.0V, fCLK = 65 MHz, fIN = 249 MHz, AIN = -9dBFS, CL = 10 pF/pin, Duty
Cycle Stabilizer On. Boldface limits apply for TJ = TMIN to TMAX: all other limits TJ = 25°C
18
Distortion vs. fIN
(fCLK = 65MSPS, AIN = -3dBFS)
Distortion vs. fIN
(fCLK = 65MSPS, AIN = -3dBFS)
Figure 30.
Figure 31.
Spectral Response @ 20MHz Input
(fCLK = 52MSPS, AIN = -9dBFS)
Spectral Response @ 20MHz Input
(fCLK = 52MSPS, AIN = -3dBFS)
Figure 32.
Figure 33.
Spectral Response @ 20MHz Input
(fCLK = 65MSPS, AIN = -9dBFS)
Spectral Response @ 20MHz Input
(fCLK = 65MSPS, AIN = -3dBFS)
Figure 34.
Figure 35.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
ADC Typical Performance Characteristics (continued)
Unless otherwise specified, the following specifications apply for AGND = DGND = D18GND = DRGND = 0V, VA = VD = VDR
= +3.3V, VD18 = +1.8V, PD = 0V, Internal VREF = +1.0V, fCLK = 65 MHz, fIN = 249 MHz, AIN = -9dBFS, CL = 10 pF/pin, Duty
Cycle Stabilizer On. Boldface limits apply for TJ = TMIN to TMAX: all other limits TJ = 25°C
Spectral Response @ 250MHz Input
(fCLK = 52MSPS, AIN = -9dBFS)
Spectral Response @ 250MHz Input
(fCLK = 52MSPS, AIN = -3dBFS)
Figure 36.
Figure 37.
Spectral Response @ 250MHz Input
(fCLK = 65MSPS, AIN = -9dBFS)
Spectral Response @ 250MHz Input
(fCLK = 65MSPS, AIN = -3dBFS)
Figure 38.
Figure 39.
IMD: f1IN = 246MHz, f2IN = 250MHz
(fCLK = 65MSPS, A1IN,2IN = -15dBFS)
Figure 40.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
19
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
CIC Output Typical Performance Characteristics
Unless otherwise specified, the following specifications apply for AGND = DGND = D18GND = DRGND = 0V, VA = VD = VDR
= +3.3V, VD18 = +1.8V, PD = 0V, Internal VREF = +1.0V, fCLK = 65 MHz, fIN = 249 MHz, AIN = -9dBFS, CL = 10 pF/pin, Duty
Cycle Stabilizer On. Boldface limits apply for TJ = TMIN to TMAX: all other limits TJ = 25°C
20
Spectral Response @ 20MHz Input
(fCLK = 52MSPS, AIN = -9dBFS)
Spectral Response @ 20MHz Input
(fCLK = 52MSPS, AIN = -3dBFS)
Figure 41.
Figure 42.
Spectral Response @ 20MHz Input
(fCLK = 65MSPS, AIN = -9dBFS)
Spectral Response @ 20MHz Input
(fCLK = 65MSPS, AIN = -3dBFS)
Figure 43.
Figure 44.
Spectral Response @ 250MHz Input
(fCLK = 52MSPS, AIN = -9dBFS)
Spectral Response @ 250MHz Input
(fCLK = 52MSPS, AIN = -3dBFS)
Figure 45.
Figure 46.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
CIC Output Typical Performance Characteristics (continued)
Unless otherwise specified, the following specifications apply for AGND = DGND = D18GND = DRGND = 0V, VA = VD = VDR
= +3.3V, VD18 = +1.8V, PD = 0V, Internal VREF = +1.0V, fCLK = 65 MHz, fIN = 249 MHz, AIN = -9dBFS, CL = 10 pF/pin, Duty
Cycle Stabilizer On. Boldface limits apply for TJ = TMIN to TMAX: all other limits TJ = 25°C
Spectral Response @ 250MHz Input
(fCLK = 65MSPS, AIN = -9dBFS)
Spectral Response @ 250MHz Input
(fCLK = 65MSPS, AIN = -3dBFS)
Figure 47.
Figure 48.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
21
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
DDC Output Typical Performance Characteristics
Unless otherwise specified, the following specifications apply for AGND = DGND = D18GND = DRGND = 0V, VA = VD = VDR
= +3.3V, VD18 = +1.8V, PD = 0V, CIC decimation = 8, Internal VREF = +1.0V, fCLK = 65 MHz, fIN = 249 MHz, AIN = -9dBFS, CL =
10 pF/pin, Duty Cycle Stabilizer On. Boldface limits apply for TJ = TMIN to TMAX: all other limits TJ = 25°C
22
Spectral Response @ 20MHz Input
(fCLK = 52MSPS, AIN = -9dBFS)
Spectral Response @ 20MHz Input
(fCLK = 52MSPS, AIN = -3dBFS)
Figure 49.
Figure 50.
Spectral Response @ 20MHz Input
(fCLK = 65MSPS, AIN = -9dBFS)
Spectral Response @ 20MHz Input
(fCLK = 65MSPS, AIN = -3dBFS)
Figure 51.
Figure 52.
Spectral Response @ 250MHz Input
(fCLK = 52MSPS, AIN = -9dBFS)
Spectral Response @ 250MHz Input
(fCLK = 52MSPS, AIN = -3dBFS)
Figure 53.
Figure 54.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
DDC Output Typical Performance Characteristics (continued)
Unless otherwise specified, the following specifications apply for AGND = DGND = D18GND = DRGND = 0V, VA = VD = VDR
= +3.3V, VD18 = +1.8V, PD = 0V, CIC decimation = 8, Internal VREF = +1.0V, fCLK = 65 MHz, fIN = 249 MHz, AIN = -9dBFS, CL =
10 pF/pin, Duty Cycle Stabilizer On. Boldface limits apply for TJ = TMIN to TMAX: all other limits TJ = 25°C
Spectral Response @ 250MHz Input
(fCLK = 65MSPS, AIN = -9dBFS)
Spectral Response @ 250MHz Input
(fCLK = 65MSPS, AIN = -3dBFS)
Figure 55.
Figure 56.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
23
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
FUNCTIONAL DESCRIPTION
Figure 57. LM97593 Dual ADC / Digital Tuner / AGC Block Diagram with Control Register Associations
The LM97593 contains two identical 12-bit ADCs driving the digital down-conversion (DDC) circuitry shown in the
block diagram in Figure 57.
ADC
The ADCs operate off of a +3.3V supply and use a pipeline architecture with error correction circuitry to help
ensure maximum performance. The differential analog input signal is digitized to 12 bits. The user has the choice
of using an internal 1.0 Volt or an external reference. Any external reference is buffered on-chip to ease the task
of driving that pin.
The clock frequency is rated up to 65 MHz. The analog input for both channels is acquired at the rising edge of
the clock. The digital data for a given sample is delayed by the pipeline for 7 clock cycles before it reaches the
input to the DDC circuit. A logic high on the power down (PD) pin reduces the converter power consumption to
50 mW. The DDC power can be further reduced by gating off the clock as described in section POWER
MANAGEMENT.
DDC
Each independent DDC channel down converts the sub-sampled IF to baseband, decimates the signal rate by a
programmable factor ranging from 32 to 16384, provides channel filtering, and outputs quadrature symbols.
A crossbar switch enables either of the two inputs or a test register to be routed to either DDC channel. Flexible
channel filtering is provided by the two programmable decimating FIR filters. The final filter outputs can be
converted to a 12-bit floating point format or standard two’s complement format. The output data is available at
both serial and parallel ports.
The LM97593’s DDC maintains over 100 dB of spurious free dynamic range and over 100 dB of out-of-band
rejection. This allows considerable latitude in channel filter partitioning between the analog and digital domains.
24
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
The frequencies, phase offsets, and phase dither of the two sine/cosine numerically controlled oscillators (NCOs)
can be independently specified. Two sets of coefficient memories and a crossbar switch allow shared or
independent filter coefficients and bandwidth for each channel. Both channels share the same decimation ratio
and input/output formats.
Each channel has its own AGC circuit for use with narrowband radio channels where most of the channel filtering
precedes the ADC. The AGC closes the loop around the DVGA, compressing the dynamic range of the signal
into the ADC. AGC gain compensation in the LM97593 removes the DVGA gain steps at the output. The time
alignment of this gain compensation circuit can be adjusted. The AGC can be configured to operate continuously
or set to a fixed gain. The two AGC circuits operate independently but share the same programmed parameters
and control signals.
The chip receives configuration and control information over a microprocessor-compatible bus consisting of an 8bit data I/O port, an 8-bit address port, a chip enable strobe, a read strobe, and a write strobe. The chip’s control
registers (8 bits each) are memory mapped into the 8-bit address space of the control port. Page select bits allow
access to the overlaid A and B set of FIR coefficients.
JTAG boundary scan and on-chip diagnostic circuits are provided to simplify system debug and test.
The LM97593 supports 3.3V I/O even though the core logic voltage is 1.8V. The LM97593 outputs swing to the
3.3V rail so they can be directly connected to 5V TTL inputs if desired.
ADC Application Information
ADC OPERATING CONDITIONS
We recommend that the following conditions be observed for operation:
3.0V ≤ VA ≤ 3.6V
VD = VA = VDR
VD18 = 1.8V
10 MHz ≤ fCLK ≤ 65 MHz
1.0 V internal reference
VCM = 1.5V (from VCOMA and VCOMB)
Analog Inputs
There is one reference input pin, VREF, which is used to select an internal reference, or to supply an external
reference. The ADC has two analog signal input pairs, VIN A+ and VIN A- for one converter and VIN B+ and VIN Bfor the other converter. Each pair of pins forms a differential input pair.
Reference Pins
The ADC is designed to operate with an internal 1.0V reference or an external 1.0V reference, but performs well
with external reference voltages in the range of 0.8V to 1.2V. Lower reference voltages will decrease the signalto-noise ratio (SNR) of the ADC. Increasing the reference voltage (and the input signal swing) beyond 1.2V may
degrade THD for a full-scale input, especially at higher input frequencies.
It is important that all grounds associated with the reference voltage and the analog input signal make connection
to the ground plane at a single, quiet point to minimize the effects of noise currents in the ground path.
The six Reference Bypass Pins (VRPA, VCOMA, VRNA, VRPB, VCOMB and VRNB) are made available for bypass
purposes. All these pins should each be bypassed to ground with a 0.1 µF capacitor. A 10 µF capacitor should
be placed between the VRPA and VRNA pins and between the VRPB and VRNB pins, as shown in Figure 86. This
configuration is necessary to avoid reference oscillation, which could result in reduced
SFDR and/or SNR.
Smaller capacitor values than those specified will allow faster recovery from the power down mode, but may
result in degraded noise performance. Loading any of these pins other than VCOMA and VCOMB may result in
performance degradation.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
25
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
The nominal voltages for the reference bypass pins are as follows:
VCOM = 1.5 V
VRP = VCOM + VREF / 2
VRN = VCOM − VREF / 2
User choice of an on-chip or external reference voltage is provided. The internal 1.0 Volt reference is in use
when the the VREF pin is connected to VA. If a voltage in the range of 0.8V to 1.2V is applied to the VREF pin, that
is used for the voltage reference. When an external reference is used, the VREF pin should be bypassed to
ground with a 0.1 µF capacitor close to the reference input pin. There is no need to bypass the VREF pin when
the internal reference is used.
Signal Inputs
The signal inputs are VIN A+ and VINA− for one ADC and VINB+ and VINB− for the other ADC . The input signal,
VIN, is defined as
VIN A = (VINA+) – (VINA−)
(1)
for the "A" converter and
VIN B = (VINB+) – (VINB−)
(2)
for the "B" converter.
Figure 58 shows the expected input signal range. Note that the common mode input voltage, VCM, should be in
the range of 1.0V to 2.0V.
The peaks of the individual input signals should never exceed 2.6V.
The ADC performs best with a differential input signal with each input centered around a common mode voltage,
VCM. The peak-to-peak voltage swing at each analog input pin should not exceed the value of the reference
voltage or the output data will be clipped.
The two input signals should be exactly 180° out of phase from each other and of the same amplitude. For single
frequency inputs, angular errors result in a reduction of the effective full scale input. For complex waveforms,
however, angular errors will result in distortion.
Figure 58. Expected Input Signal Range
For single frequency sine waves the full scale error in LSBs can be described as approximately
EFS = 4096 ( 1 - sin (90° + dev))
where
•
dev is the angular difference in degrees between the two signals having a 180° relative phase relationship to
each other (see Figure 59)
(3)
Drive the analog inputs with a source impedance less than 100Ω.
26
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
Figure 59. Angular Errors Between the Two Input Signals Will Reduce the Output Level or Cause
Distortion
For differential operation, each analog input pin of the differential pair should have a peak-to-peak voltage equal
to the reference voltage, VREF, be 180 degrees out of phase with each other and be centered around VCM.
Single-Ended Operation
Performance with differential input signals is better than with single-ended signals. For this reason, single-ended
operation is not recommended. However, if single ended-operation is required and the resulting performance
degradation is acceptable, one of the analog inputs should be connected to the d.c. mid point voltage of the
driven input. The peak-to-peak input signal at the driven input pin should be twice the reference voltage to
maximize SNR and SINAD performance (Figure 58b). For example, set VREF to 1.0V, bias VIN− to 1.5V and drive
VIN+ with a signal range of 0.5V to 2.5V.
Because very large input signal swings can degrade distortion performance, better performance with a singleended input can be obtained by reducing the reference voltage when maintaining a full-range output.
Driving the Analog Inputs
The VIN+ and the VIN− inputs of the ADC consist of an analog switch followed by a switched-capacitor amplifier.
As the internal sampling switch opens and closes, current pulses occur at the analog input pins, resulting in
voltage spikes at the signal input pins. As the driving source attempts to counteract these voltage spikes, it may
add noise to the signal at the ADC analog input. C1, C2, and C3 as shown in Figure 86 improve the ADC
performance by filtering these voltage spikes. These components should be placed close to the ADC inputs
because the input pins of the ADC are the most sensitive part of the system and this is the last opportunity to
filter that input.
For Nyquist applications the RC pole should be at the ADC sample rate. The ADC input capacitance in the
sample mode should be considered when setting the RC pole. For wideband undersampling applications, the RC
pole should be set at about 1.5 to 2 times the maximum input frequency to maintain a linear delay response. The
values of the RC shown in Figure 86 are suitable for applications with input frequencies up to approximately
70MHz.
Input Common Mode Voltage
The input common mode voltage, VCM, should be in the range of 1.0V to 2.0V and be a value such that the peak
excursions of the analog signal do not go more negative than ground or more positive than 2.6V. See Reference
Pins.
DIGITAL INPUTS
Digital TTL/CMOS compatible inputs consist of CK, REFSEL/DCS.
CLK
The CLK signal controls the timing of the sampling process. Drive the clock input with a stable, low jitter clock
signal in the range of 10 MHz to 65 MHz. The higher the input frequency, the more critical it is to have a low jitter
clock. The trace carrying the clock signal should be as short as possible and should not cross any other signal
line, analog or digital, not even at 90°.
The CLK signal also drives an internal state machine. If the CLK is interrupted, or its frequency too low, the
charge on internal capacitors can dissipate to the point where the accuracy of the output data will degrade. This
is what limits the lowest sample rate.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
27
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
The clock line should be terminated at its source in the characteristic impedance of that line. Take care to
maintain a constant clock line impedance throughout the length of the line. Refer to Application Note AN-905
(literature number SNLA035) for information on setting characteristic impedance.
It is highly desirable that the the source driving the ADC CLK pin only drive that pin. However, if that source is
used to drive other things, each driven pin should be a.c. terminated with a series RC to ground such that the
resistor value is equal to the characteristic impedance of the clock line and the capacitor value is
where
•
•
•
tPD is the signal propagation time down the clock line
"L" is the line length
ZO is the characteristic impedance of the clock line
(4)
This termination should be as close as possible to the ADC clock pin but beyond it as seen from the clock
source. Typical tPD is about 150 ps/inch (60 ps/cm) on FR-4 board material. The units of "L" and tPD should be
the same (inches or centimeters).
The duty cycle of the clock signal can affect the performance of the A/D Converter. Because achieving a precise
duty cycle is difficult, the LM97593 has a Duty Cycle Stabilizer which can be enabled using the REFSEL/DCS
pin. It is designed to maintain performance over a clock duty cycle range of 30% to 70% at 65 MSPS.
REFSEL/DCS
This pin is used in conjunction with VREF (pin 21) to select the reference source and turn the Duty Cycle
Stabilizer (DCS) on or off.
When REFSEL/DCS is LOW and VREF is HIGH, the internal 1.0V reference is selected and DCS is On.
When REFSEL/DCS is HIGH, an external reference voltage in the range of 0.8V to 1.2V should be applied to the
VREF input. DCS is On.
With REFSEL/DCS pin connected to VCOMA or VCOMB, the internal 1.0V reference is selected and DCS is Off.
When enabled, duty cycle stabilization can compensate for clock inputs with duty cycles ranging from 30% to
70% and generate a stable internal clock, improving the performance of the part.
Table 1. VREF, REFSEL/DCS Pin Functions
REFSEL/DCS (pin 8)
VREF (pin 21)
Reference
DCS
Logic Low
Logic High
Internal 1.0 V
ON
Logic High
0.8 to 1.2V
External
ON
VCOMA or VCOMB
Logic High
Internal 1.0V
OFF
VCOMA or VCOMB
0.8 to 1.2V
External
OFF
PD
The PD pin, when high, holds the ADC in a power-down mode to conserve power when the converter is not
being used. The output data pins are undefined and the data in the pipeline is corrupted while in the power down
mode.
The Power Down Mode Exit Cycle time is determined by the value of the components on pins 15, 16, 17, 22, 23
and 24. These capacitors lose their charge in the Power Down mode and must be recharged by on-chip circuitry
before conversions can be accurate. Smaller capacitor values allow slightly faster recovery from the power down
mode, but can result in a reduction in SNR, SINAD and ENOB performance.
28
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
DDC Application Information
CONTROL INTERFACE
The LM97593 is configured by writing control information into 237 control registers within the chip. The contents
of these control registers and how to use them are described in Control Register Addresses and Defaults. The
registers are written to or read from using the D[7:0], A[7:0], CE, RD and WR pins. This interface is designed to
allow the LM97593 to appear to an external processor as a memory mapped peripheral. See Figure 15 for
details.
The control interface is asynchronous with respect to the system clock, CK. This allows the registers to be
written or read at any time. In some cases this might cause an invalid operation since the interface is not
internally synchronized. In order to assure correct operation, SI must be asserted after the control registers are
written.
The D[7:0], A[7:0], WR, RD and CE pins should not be driven above the positive supply voltage.
Master Reset
A master reset pin, MR, is provided to initialize the LM97593 to a known condition and should be strobed after
power up. This signal will clear all sample data and all user programmed data (filter coefficients and AGC
settings). All outputs will be disabled (tri-stated). ASTROBE and BSTROBE will be asserted to initialize the
DVGA values. Control Register Addresses and Defaults describes the control register default values.
Synchronizing Multiple LM97593 Chips
A system containing two or more LM97593 chips will need to be synchronized if coherent operation is desired.
To synchronize multiple LM97593 chips, connect all of the sync input pins together so they can be driven by a
common sync strobe. Synchronization occurs on the first rising edge of CK after SI goes high. When SI is
asserted all sample data is immediately cleared, the numerically controlled oscillator (NCO) phase offset is
initialized, the NCO dither generators are reset, and the CIC decimation ratio is initialized. Only the configuration
data loaded into the microprocessor interface remains unaffected.
SI may be held low as long as desired after a minimum of 4 CK periods.
Input Source
The input crossbar switch allows either VINA, VINB, or a test register to be routed to the channel A or channel B
AGC/ DDC. The AGC outputs, AGAIN and BGAIN, are not switched. If VINA and VINB are exchanged the AGC
loop will be open and the AGC will not function properly.
Selecting the test register as the input source allows the AGC or DDC operation to be verified with a known
input. See Test Register for further discussion.
DOWN CONVERTERS
A detailed block diagram of each DDC channel is shown in Figure 60. Each down converter uses a complex
NCO and mixer to quadrature downconvert a signal to baseband. The “FLOAT TO FIXED CONVERTER” treats
the 15-bit mixer output as a mantissa and the AGC output, EXP, as a 3-bit exponent. It performs a bit shift on the
data based on the value of EXP. This bit shifting is used to expand the compressed dynamic range resulting from
the DVGA operation. The DVGA gain is adjusted in 6dB steps which are equivalent to each digital bit shift.
Digitally compensating for the DVGA gain steps in the LM97593 causes the DDC output to be linear with respect
to the DVGA input. The AGC operation will be completely transparent at the LM97593 output.
The exponent (EXP) can be forced to its maximum value by setting the EXP_INH bit. If xin(n) is the DDC input,
the signal after the “FLOAT TO FIXED CONVERTER” is
x3(n) = xin(n)*cos(ωn)*2EXP
(5)
for the I component. Changing the ‘cos’ to ‘sin’ in this equation will provide the Q component.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
29
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
DEC_BY_4
F2_COEF
F1_COEF
GAIN_A
DEC
CIC FILTER
DECIMATION BY
8 TO 2k
EXP
15
17
FREQ_A
17
SIN
22
21
Data @ FCK/N
21
F2 FILTER
DECIMATE BY
2 OR 4
SAT & ROUND
21
F1 FILTER
DECIMATE BY 2
Data @ FCK
= FS (FSAMPLE)
SAT & ROUND
14
22
SHIFT UP
ROUND
MUXA
FLOAT
TO FIXED
CONVERTER
EXPONENT
15
I
SAT
3
SCALE
EXP
(from AGC)
SHIFT UP
EXP_INH
The “FLOAT TO FIXED CONVERTER” circuit expands the dynamic range compression performed by the DVGA.
Signals from this point onward extend across the full dynamic range of the signals applied to the DVGA input.
This allows the AGC to operate continuously through a burst without producing artifacts in the signal due to the
settling response of the decimation filters after a 6dB DVGA gain adjustment. For example, if the DVGA input
signal were to increase causing the ADC output level to cross the AGC threshold level, the gain of the DVGA
would change by -6dB. The 6dB step is allowed to propagate through the ADC and mixers and is compensated
out just before the filtering. The accuracy of the compensation is dependent on timing and the accuracy of the
DVGA gain step. The LM97593 allows the timing of the gain compensation to be adjusted in the EXT_DELAY
register; see the end of section AGC for more information. The AGC requires 21 bits (14-bit internal bus output +
7-bit shift) to represent the full linear dynamic range of the signal. The output word must be set to either 24-bit or
32-bit to take advantage of the entire dynamic range available. The LM97593 can also be configured to output a
floating point format with up to 138dB of numerical resolution using only 12 output bits.
TO
OUTPUT
CIRCUIT
Q
21
Data @ FCK/N*2
COS
Data @ FCK/N*2*F2_DEC
= OFS (Output FSAMPLE)
NCO
PHASE_A
N = DEC + 1
Figure 60. LM97593 Down Converter, Channel A (Channel B is identical)
The “SHIFT UP” circuit will be discussed in the section Four Stage CIC Filter.
A 4-stage cascaded-integrator-comb (CIC) filter and a two-stage decimate by 4 or 8 finite impulse response (FIR)
filter are used to lowpass filter and isolate the desired signal. The CIC filter reduces the sample rate by a
programmable factor ranging from 8 to 2048 (decimation ratio). The CIC outputs are followed by a gain stage
and then followed by a two-stage decimate by 4 or 8 filter. The gain circuit allows the user to boost the gain of
weak signals by up to 42 dB in 6 dB steps. It also rounds the signal to 21 bits and saturates at plus or minus full
scale.
The first stage of the two stage filter is a 21-tap, symmetric decimate by 2 FIR filter (F1) with programmable 16
bit tap weights. The coefficients of the first 11 taps are downloaded to the chip as 16 bit words. Since the filter is
a symmetric configuration only the first 11 coefficients must be loaded. First Programmable FIR Filter (F1)
provides a generic set of coefficients that compensate for the rolloff of the CIC filter and provide a passband flat
to 0.01dB with 70 dB of out of band rejection. A second coefficient set is provided that has a narrower output
passband and greater out-of-band rejection. The second set of coefficients is ideal for systems such as GSM
where far-image rejection is more important than adjacent channel rejection.
The second stage is a 63 tap decimate by 2 or 4 programmable FIR filter (F2) also with 16 bit tap weights. Filter
coefficients for a flat response from -0.4FS to +0.4FS of the output sample rate with 80dB of out of band rejection
are provided in Second Programmable FIR Filter (F2). A second set of F2 coefficients is also provided to
enhance performance for GSM systems. The user can also design and download their own final filter to
customize the channel’s spectral response. Typical uses of programmable filter F2 include matched (root-raised
cosine) filtering, or filtering to generate oversampled outputs with greater out of band rejection. The 63 tap
symmetrical filter is downloaded into the chip as 32 words, 16 bits each. Saturation to plus or minus full scale is
performed at the output of F1 and F2 to clip the signal rather than allow it to roll over.
30
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
Figure 61. Complex NCO Output Phase Dither Disabled
Example of NCO spurs due to phase truncation
(Before Phase Dithering)
The LM97593 provides two sets of coefficient memory for both F1 and F2. These coefficient memories can be
independently routed to channel A, channel B, or both channel A and B with a crossbar switch. The coefficients
can be switched on the fly but some time will be required before valid output data is available.
The Numerically Controlled Oscillator
The tuning frequency of each down converter is specified as a 32 bit word (.02Hz resolution at CK=52MHz) and
the phase offset is specified as a 16 bit word (.005o). These two parameters are applied to the Numerically
Controlled Oscillator (NCO) circuit to generate sine and cosine signals used by the digital mixer. The NCOs can
be synchronized with NCOs on other chips via the sync pin SI. This allows multiple down converter outputs to be
coherently combined, each with a unique phase and amplitude.
The tuning frequency is set by loading the FREQ register according to the formula FREQ = 232F/FCK, where F is
the desired tuning frequency and FCK is the chip’s clock rate. FREQ is a 2’s complement word. The range for F is
from
-FCK/2 to +FCK(1-2-31)/2.
If a sub-sampled signal is in an even Nyquist zone the sampling process causes the order of the I and Q
components to be reversed. Should this occur simply invert the polarity of the tuning frequency F.
Figure 62. Complex NCO Output Phase Dither Enabled
Example of NCO spurs due to phase truncation
(After Phase Dithering)
The 2’s complement format represents full-scale negative as 10000000 and full-scale positive as 01111111 for
an 8-bit example.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
31
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
The 16 bit phase offset is set by loading the PHASE register according to the formula PHASE = 216P/2π, where
P is the desired phase in radians ranging between 0 and 2π. PHASE is an unsigned 16-bit number. P ranges
from 0 to 2π(1-2-16). Phase dithering can be enabled to reduce the spurious signals created by the NCO due to
phase truncation. This truncation is unavoidable since the frequency resolution is much finer than the phase
resolution. With dither enabled, spurs due to phase truncation are below -100 dBc for all frequencies and phase
offsets. Each NCO has its own dither source and the initial state of one is maximally offset with respect to the
other so that they are effectively uncorrelated. The phase dither sources are on by default. They are
independently controlled by the DITH_A and DITH_B bits. The amplitude resolution of the ROM creates a worstcase spur amplitude of -101dBc rendering amplitude dither unnecessary.
The spectrum plots in Figure 61 and Figure 62show the effectiveness of phase dither in reducing NCO spurs due
to phase truncation for a worst-case example (just below FS/8). With dither off, the spur is at -86.4dBFS. With
dither on, the spur is below -125dBFS, disappearing into the noise floor. This spur is spread into the noise floor
which results in an SNR of -83.6dBFS. The channel filter’s processing gain will further improve the SNR.
Figure 63 shows the spur levels as the tuning frequency is scanned over a narrow portion of the frequency
range. The spurs are again a result of phase quantization but their locations move about as the frequency scan
progresses. As before, the peak spur level drops when dithering is enabled. When dither is enabled and the
fundamental frequency is exactly at FS/8, the worst-case spur due to amplitude quantization can be observed at 101dBc in Figure 64.
Four Stage CIC Filter
The mixer outputs are decimated by a factor of N in a four stage CIC filter. N is programmable to any integer
between 8 and 2048. Decimation is programmed in the DEC register where DEC = N - 1. The programmable
decimation allows the chip’s usable output bandwidth to range from about ±1.27kHz to ±650kHz when the input
data rate (which is equal to the chip’s clock rate, FCK) is 52 MHz. For the maximum sample rate of 65MHz, the
LM97593’s output bandwidth will range from about ±1.58kHz to ±812kHz. A block diagram of the CIC filter is
shown in Figure 65.
The CIC filter is primarily used to decimate the high-rate incoming data while providing a rough lowpass
characteristic. The lowpass filter will have a sin(x)/x response (similar to the AGC’s CIC shown in Figure 80)
where the first null is at FS/N.
Figure 63. Complex NCO Output Phase Dither Disabled
NCO Spurs due to Phase Quantization
32
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
Figure 64. Complex NCO Output Phase Dither Enabled
Worst Case Amplitude Spur, NCO at FS/8
The CIC filter has a gain equal to N4 (filter decimation^4) which must be compensated for in the “SHIFT UP”
circuit shown in Figure 65 as well as Figure 60. This circuit has a gain equal to 2(SCALE-44), where SCALE ranges
from 0 to 40. This circuit divides the input signal by 244 providing
Figure 65. Four-stage decimate by N CIC filter
maximum headroom through the CIC filter. For optimal noise performance the SCALE value is set to increase
this level until the CIC filter is just below the point of distortion. A value is normally calculated and loaded for
SCALE such that GAINSHIFTUP*GAINCIC ≤ 1. The actual gain of the CIC filter will only be unity for power-of-two
decimation values. In other cases the gain will be somewhat less than unity.
Channel Gain
The gain of each channel can be boosted up to 42 dB by shifting the output of the CIC filter left by 0 to 7 bits
prior to rounding it to 21 bits. For channel A, the gain of this stage is: GAIN = 2GAIN_A , where GAIN_A ranges
from 0 to 7. Overflow due to the GAIN circuit is saturated (clipped) at plus or minus full scale. Each channel can
be given its own GAIN setting.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
33
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
First Programmable FIR Filter (F1)
The CIC/GAIN outputs are followed by two stages of filtering. The first stage is a 21 tap decimate-by-2 symmetric
FIR filter with programmable coefficients. Typically, this filter compensates for a slight droop induced by the CIC
filter while removing undesired alias images above Nyquist. In addition, it often provides stopband assistance to
F2 when deep stop bands are required. The filter coefficients are 16-bit 2’s complement numbers. Unity gain will
be achieved through the filter if the sum of the 21 coefficients is equal to 216. If the sum is not 216, then F1 will
introduce a gain equal to (sum of coefficients)/ 216. The 21 coefficients are identified as coefficients h1(n), n =
0, ..., 20 where h1(10) is the center tap. The coefficients are symmetric, so only the first 11 are loaded into the
chip.
Two example sets of coefficients are provided here. The first set of coefficients, referred to as the standard set
(STD), compensates for the droop of the CIC filter providing a passband which is flat (0.01 dB ripple) over 95%
of the final output bandwidth with 70dB of out-of-band rejection (see Figure 66). The filter has a gain of 0.999
and is symmetric with the following 11 unique taps (1|21, 2|20, ..., 10|12, 11):
29, -85, -308, -56, 1068, 1405, -2056, -6009,
1303, 21121, 32703
Figure 66. F1 STD frequency response
The second set of coefficients (GSM set) are intended for applications that need deeper stop bands or need
oversampled outputs. These requirements are common in cellular systems where out of band rejection
requirements can exceed 100dB (see Figure 67). They are useful for wideband radio architectures where the
channelization is done after the ADC. These filter coefficients introduce a gain of 0.984 and are:
-49, -340, -1008, -1617, -1269, 425, 3027, 6030,
9115, 11620, 12606
Figure 67. F1 GSM frequency response
34
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
Second Programmable FIR Filter (F2)
The second stage decimate by two or four filter also uses externally downloaded filter coefficients. F2 determines
the final channel filter response. The filter coefficients are 16-bit 2’s complement numbers. Unity gain will be
achieved through the filter if the sum of the 63 coefficients is equal to 216. If the sum is not 216, then the F2 will
introduce a gain equal to (sum of coefficients)/216.
The 63 coefficients are identified as h2(n), n = 0, ..., 62 where h2(31) is the center tap. The coefficients are
symmetric, so only the first 32 are loaded into the chip. An example filter (STD F2 coefficients, see Figure 68)
with 80dB out-of-band rejection, gain of 1.00, and 0.03 dB peak to peak passband ripple is created by this set of
32 unique coefficients:
-14, -20, 19, 73, 43, -70, -82, 84, 171, -49, -269,
-34, 374, 192, -449,
-430, 460,751, -357, -1144, 81, 1581, 443, -2026,
-1337, 2437, 2886,
-2770, -6127, 2987, 20544, 29647
A second set of F2 coefficients (GSM set, see Figure 69) suitable for meeting the stringent wideband GSM
requirements with a gain of 0.999 are:
-536, -986, 42, 962, 869, 225, 141, 93, -280,
-708, -774, -579, -384,
-79, 536, 1056, 1152, 1067, 789, 32, -935, -1668,
-2104, -2137, -1444,
71, 2130, 4450, 6884, 9053, 10413, 10832
The filter coefficients of both filters can be used to tailor the spectral response to the user’s needs. For example,
the first can be loaded with the standard set to provide a flat response through to the second filter. The latter can
then be programmed as a Nyquist (typically a root-raised-cosine) filter for matched filtering of digital data.
Figure 68. F2 STD frequency response
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
35
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
Figure 69. F2 GSM frequency response
The complete channel filter response for standard coefficients is shown in Figure 70. Passband flatness is shown
in Figure 71. The complete filter response for GSM coefficients is shown in Figure 72. GSM Passband flatness is
shown in Figure 73.
The mask shown in Figure 72 is derived from the ETSI GSM 5.05 specifications for a normal Basestation
Transceiver (BTS). For interferers, 9dB was added to the carrier to interference (C/I) ratios. For blockers, 9dB
was added to the difference between the blocker level and 3dB above the reference sensitivity level.
Channel Bandwidth vs. Sample Rate
When the LM97593 is used for GSM systems, a bandwidth of about 200kHz is desired. With a sample rate of
52MHz, the total decimation of 192 provides the desired 270.833kHz output sample rate. This output sample rate
in combination with the FIR filter coefficients create the desired channel bandwidth. If the sample rate is
increased to 65MHz, the decimation must also be increased to 65MHz/270.833kHz or 240.
Figure 70. CIC, F1, & F2 STD frequency response
36
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
Figure 71. CIC, F1, & F2 STD Passband Flatness
Figure 72. CIC, F1, & F2 GSM frequency response
This new decimation rate will maintain the same output bandwidth. The output bandwidth may only be changed
in relation to the output sample rate by creating a new set of FIR filter coefficients. As the filter bandwidth
decreases relative to the output sample rate, the CIC droop compensation performed by F1 may no longer be
required.
Figure 73. CIC, F1, & F2 GSM Passband Flatness
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
37
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
Overall Channel Gain
The overall gain of the chip is a function of the amount of decimation (N), the settings of the “SHIFT UP” circuit
(SCALE), the GAIN setting, the sum of the F1 coefficients, and the sum of the F2 coefficients. The overall gain is
shown below in Equation 6.
(6)
Where:
(7)
and:
(8)
It is assumed that the DDC output words are treated as fractional 2’s complement words. The numerators of GF1
and GF2 equal the sums of the impulse response coefficients of F1 and F2, respectively. For the STD and GSM
sets, GF1 and GF2 are nearly equal to unity. Observe that the AGAIN term in Equation 6 is cancelled by the
DVGA operation so that the entire gain of the DRCS is independent of the DVGA setting when EXP_INH=0. The
1/2 appearing in Equation 6 is the result of the 6dB conversion loss in the mixer. For full-scale square wave
inputs the 1/2 should be set to 1 to prevent signal distortion.
Data Latency and Group Delay
The LM97593 latency calculation assumes that the FIR filter latency will be equal to the time required for data to
propagate through one half of the taps. The CIC filter provides 4N equivalent taps where N is the CIC decimation
ratio. F1 and F2 provide 21 and 63 taps respectively. When these filters are reflected back to the input rate, the
effective taps are increased by decimation. This results in a total of 151N taps.
The total latency is found by dividing the number of taps by 2 and adding pipeline delays. When the F2
decimation is 2 the latency is 80N. When the F2 decimation is 4 the latency is 82N. The LM97593 filters are
linear phase filters so the group delay remains constant.
OUTPUT MODES
After processing by the DDC, the data is then formatted for output.
All output data is two’s complement. The serial outputs power up in a tri-state condition and must be
enabled when the chip is configured. Parallel outputs are enabled by the POUT_EN pin.
Output formats include truncation to 8 or 32 bits, rounding to 16 or 24 bits, and a 12-bit floating point format (4-bit
exponent, 8-bit mantissa, 138dB numeric range). This function is performed in the OUTPUT CIRCUIT shown in
Figure 74.
38
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
Figure 74. LM97593 output circuit
The channel outputs are accessible through serial output pins and a 16-bit parallel output port. The RDY pin is
provided to notify the user that a new output sample period (OSP) has begun. OSP refers to the interval between
output samples at the decimated output rate. For example, if the input rate (and clock rate) is 52 MHz and the
overall decimation factor is 192 (N=48, F2 decimation=2) the OSP will be 3.69 microseconds which corresponds
to an output sample rate of 270.833kHz. An OSP starts when a sample is ready and stops when the next one is
ready.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
39
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
Refer to Figure 13 for detailed timing information.
Figure 75. Serial output formats
Serial Outputs
The LM97593 provides a serial clock (SCK), a frame strobe (SFS) and two data lines (AOUT and BOUT) to
output serial data. The MUX_MODE control register specifies whether the two channel outputs are transmitted
on two separate serial pins, or multiplexed onto one pin in a time division multiplexed (TDM) format. Separate
output pins are not provided for the I and Q halves of complex data. The I and Q outputs are always multiplexed
onto the same serial pin. The I-component is output first, followed by the Q-component. By setting the PACKED
mode bit to ‘1’ a complex pair may be treated as a single double-wide word. The RDY signal is used to identify
the first word of a complex pair of the TDM formatted output when the SFS_MODE bit is set to ‘0’. Setting
SFS_MODE to ‘1’ causes the LM97593 to output a single SFS pulse for each output period. This SFS pulse will
be coincident with RDY and only a single SCK period wide. The TDM modes are summarized in Table 2.
Table 2. TDM Modes
MUX_MODE
AOUT
BOUT
0
0
OUTA
OUTB
1
OUTA, OUTB
LOW
0
OUTA
OUTB
1
OUTA, OUTB
LOW
1
40
SERIAL OUTPUTS
SFS_MODE
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
The serial outputs use the format shown in Figure 75. Figure 75(a) shows the standard output mode (the
PACKED mode bit is low). The chip clocks the frame and data out of the chip on the rising edge of SCK (or
falling edge if the SCK_POL bit in the input control register is set high).
AOUT
BOUT
SFS
SCK
AOUT
SCK
SFS
RDY
To DSP
SCKMASTER = 2*SCKSLAVE
4
SDC_EN = 1
MUX_MODE = 0
12
12
12
4
SDC_EN = 0
MIX_MODE = 0
PACKED = 1
Master
CLC5903
12
Slave
CLC5903
POUT_SEL[0]
POUT_SEL[1]
POUT_SEL[2]
SCK_IN
ADCs and
DVGAs
ADCs and
DVGAs
Figure 76. Serial Daisy-Chain Mode
Data should be captured on the falling edge of SCK (rising if SCK_POL=1). The chip sends the I data first by
setting SFS high (or low if SFS_POL in the input control register is set high) for one clock cycle, and then
transmitting the data, MSB first, on as many SCK cycles as are necessary. Without a pause, the Q data is
transferred next as shown in Figure 75(a). If the PACKED control bit is high, then the I and Q components are
sent as a double length word with only one SFS strobe as shown in Figure 75(b). If both channels are
multiplexed out the same serial pin, then the subsequent I/Q channel words will be transmitted immediately
following the first I/Q pair as shown in Figure 75(c). Figure 75(c) also shows how SFS_MODE=1 allows the SFS
signal to be used to identify the A and B channels in the TDM serial transmission. The serial output rate is
programmed by the RATE register to CK divided by 1, 2, 4, 8, 16, or 32. The serial interface will not work
properly if the programmed rate of SCK is insufficient to clock out all the bits in one OSP.
Serial Port Daisy-Chain Mode
Two LM97593s can be connected in series so that a single DSP serial port can receive four DDC output
channels. This mode is enabled by setting the SDC_EN bit to ‘1’ on the serial daisy-chan (SDC) master. The
SDC master is the LM97593 which is connected to the DSP while the SDC slave’s serial output drives the
master. The SDC master’s RATE register must be set so that its SCK rate is twice that of the SDC slave, the
SDC master must have MUX_MODE=1, the SDC slave must have MUX_MODE=0 and PACKED=1, and both
chips must come out of a MR or SI event within four CK periods of each other. In this configuration, the master’s
serial output data is shifted out to the DSP and then the slave’s serial data is shifted out. All the serial output data
will be muxed onto the master’s AOUT pin as shown in Figure 76.
Serial Port Output Number Formats
Several numeric formats are selectable using the FORMAT control register. The I/Q samples can be rounded to
16 or 24 bits, or truncated to 8 bits. The packed mode works as described above for these fixed point formats. A
floating point format with 138dB of dynamic range in 12 bits is also provided. The mantissa (m) is 8 bits and the
exponent (e) is 4 bits. The MSB of each segment is transmitted first. When the packed mode is selected, the I/Q
samples are packed regardless of the state of MUX_MODE, and the data is sent as mI/eI/eQ/ mQ which allows
the two exponents to form an 8-bit word. This is shown in Figure 75(d). For all formats, once the defined length
of the word is complete, SCK stops toggling.
Parallel Outputs
Output data from the channels can also be taken from a 16-bit parallel port. A 3-bit word applied to the
POUT_SEL[2:0] pins determines which 16-bit segment is multiplexed to the parallel port. Table 3 defines this
mapping. To allow for bussing of multiple chips, the parallel port is tri-stated unless POUT_EN is low. The RDY
signal indicates the start of an OSP and that new data is ready at the parallel output. The user has one OSP to
cycle through whichever registers are needed. The RATE register must be set so that each OSP is at least 5
SCK periods.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
41
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
Parallel Port Output Numeric Formats
The I/Q samples can be rounded to 16 or 24 bits or the full 32 bit word can be read. By setting the word size to
32 bits it is possible to read out the top 16-bits and only observe the top 8 bits if desired. Additionally, the output
samples can be formatted as floating point numbers with an 8-bit mantissa and a 4 bit exponent. For the fixedpoint formats, the valid bits are justified into the MSBs of the registers of Table 3 and
Table 3. Register Selection for Parallel Output
POUT_SEL
Normal Register Contents
Floating Point Register Contents
0
IA upper 16-bits
0000/eIA/mIA
1
IA lower 16-bits
0x0000
2
QA upper 16-bits
0000/eQA/mQA
3
QA lower 16-bits
0x0000
4
IB upper 16-bits
0000/eIB/mIB
5
B lower 16-bits
0x0000
6
QB upper 16-bits
0000/eQB/mQB
7
QB lower 16-bits
0x0000
all other bits are set to zero. For the floating point format, the valid bits are placed in the upper 16-bits of the
appropriate channel register using the format 0000/eI/mI for the I samples.
AGC
The LM97593 AGC processor monitors the output level of the ADC and servos it to the desired setpoint. The
ADC input is controlled by the DVGA to maintain the proper setpoint level. DVGA operation results in a
compression of the signal through the ADC. The DVGA signal compression is reversed in the LM97593 to
provide > 120dB of linear dynamic range. This is illustrated in Figure 77.
Figure 77. Output Gain Scaling vs. Input Signal
In order to use the AGC, the DRCS Control Panel software may be used to calculate the programmable
parameters. To generate these parameters, only the desired setpoint, deadband+ hysteresis, and loop time
constant need to be supplied. All subsequent calculations are performed by the software. Complete details of the
AGC operation are provided in an appendix.
42
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
Figure 78. AGC Setup
AGC setpoint and deadband are illustrated in Figure 78. The loop time constant is a measure of how fast the
loop will track a changing signal. Values down to approximately 1.0 microsecond will be stable with the second
order LC noise filter. Since the DVGA operates with 6dB steps the deadband should always be greater than 6dB
to prevent oscillation. An increased deadband value will reduce the amount of AGC operation. A decreased
deadband value will increase the amount of AGC operation but will hold the ADC output closer to the setpoint.
The threshold should be set so that transients do not cause sustained overrange at the ADC inputs. The
threshold setting can also be used to set the ADC input near its optimal performance level.
The AGC will free run when AGC_HOLD_IC is set to ‘0’. It may be set to a fixed gain by setting AGC_HOLD_IC
to ‘1’ after programming the desired gain in the AGC_IC_A and AGC_IC_B registers. Allowing the AGC to free
run should be appropriate for most applications.
Programming the AGC_COMB_ORD register allows the AGC power detector bandwidth to be reduced if desired.
This will tend to improve the power detector’s ability to reject the signal carrier frequency and reduce overall AGC
activity. Figure 80 shows the power detector response.
The analog gain change from the DVGA must be compensated by the "Float To Fixed" converter after the
appropriate delay. This delay can be adjusted by the EXT_DELAY register value to make sure the analog gain
change is properly compensated in the digital domain.
Figure 79 shows the internal clock latency paths related to the DVGA and "Float To Fixed" timing conpensation.
In this diagram registers are represented by z-N where N is the sample delay in ADC clock periods. Following the
path from the output of the AGC integrator through the DVGA, bandpass filter, ADC and internal register delays
adds up to 6 clocks prior to the "Float To Fixed" converter excluding the bandpass filter and ADC. Following the
path from the AGC integrator to the "Float To Fixed" is also 6 clocks when EXT_DELAY = 0. The value
programmed in EXT_DELAY should be set to the pipeline latency of the ADC plus the latency of the bandpass
filter (typically one clock). If ASTROBE and BSTROBE are not used then subtract one from the resulting total
latency.
The LM97593 includes an integrated ADC with a pipeline latency of 7 clocks. Adding one additional clock period
for the bandpass filter requires EXT_DELAY = 7 when the DVGA ASTROBE and BSTROBE signals are used,
otherwise, program EXT_DELAY = 6. In most cases ASTROBE and BSTROBE signals are not used so
EXT_DELAY is typically set to 6.
More accurate time alignment may improve the equalizer / demodulator performance for EDGE modulated
signals and other signals with a large AM component.
Figure 79. Function controlled by EXT_DELAY register
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
43
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
POWER MANAGEMENT
The LM97593 can be placed in a low power (static) state by stopping the input clock and setting the PD pin high.
To prevent this from placing the LM97593 into unexpected states, the SI pin of the LM97593 should be asserted
prior to disabling the input clock and held asserted until the input clock has returned to a stable condition.
TESTABILITY
JTAG Boundary Scan
The LM97593 supports IEEE 1149.1 compliant JTAG Boundary Scan for the I/O's. The following pins are used:
TRST
(test reset)
TMS
(test mode select)
TDI
(test data in)
TDO
(test data out)
TCK
(test clock)
The following JTAG instructions are supported:
Instruction
Description
BYPASS
Connects TDI directly to TDO
EXTEST
Enables the test access port controller to drive the outputs
IDCODE
Connects the 32-bit ID register to TDO
SAMPLE/PRELOAD
Allows the test access port to sample the device inputs and preload
test output data
HIGHZ
Tri-states the outputs
The JTAG Boundary Scan can be used to verify printed circuit board continuity at the system level.
Test Register
The user is able to program a value into TEST_REG and substitute this for the normal channel inputs from the
AIN/ BIN pins by selecting it with the crossbar. With the NCO frequency set to zero this allows the DDCs and the
output interface of the chip to be verified. Also, the AGC loop can be opened by setting AGC_HOLD_IC high and
setting the gain of the DVGA by programming the appropriate value into the AGC_IC_A/B register.
Debug Access Port
Real-time access to the following signals is provided by configuring the control interface debug register:
• NCO sine and cosine outputs
• data after round following mixers
• data before F1 and F2
• data after CIC filter within the AGC
The access points are multiplexed to a 20-bit parallel output port which is created from signal pins POUT[15:0],
AOUT, BOUT, SFS, and RDY according to the table below:
44
Normal Mode Pin
Debug Mode Pin
POUT[15:0]
DEBUG[19:4]
RDY
DEBUG[3]
SFS
DEBUG[2]
AOUT
DEBUG[1]
BOUT
DEBUG[0]
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
SCK will be set to the proper strobe rate for each debug tap point. POUT_EN and PSEL[2:0] have no effect in
Debug Mode. The outputs are turned on when the Debug Mode bit is set. Normal serial outputs are also
disabled.
CONTROL REGISTERS
The chip is configured and controlled through the use of 8-bit control registers. These registers are accessed for
reading or writing using the control bus pins (CE, RD, WR, A[7:0], and D[7:0]) described in CONTROL
INTERFACE.
The two sets of FIR coefficients are overlaid at the same memory address. Use the PAGE_SEL registers
to access the second set of coefficients.
The register names and descriptions are listed below in Control Register Addresses and Defaults. A quick
reference table is provided in the Condensed LM97593 Address Map.
Control Register Addresses and Defaults
Register Name
Width
Type
Default (1)
Addr
Bit
Description
7:0
2:0
CIC decimation control. N=DEC+1. Valid range is from 7 to 2047.
Format is an unsigned integer. This affects both channels.
4
DEC
11b
R/W
7
0(LSBs)
1(MSBs)
DEC_BY_4
1b
R/W
0
1
Controls the decimation factor in F2. 0=Decimate by 2.
1=Decimate by 4. This affects both channels.
SCALE
6b
R/W
0
2
5:0
CIC SCALE parameter. Format is an unsigned integer
representing the number of left bit shifts to perform on the data
prior to the CIC filter. Valid range is from 0 to 40. This affects
both channels.
GAIN_A
3b
R/W
0
3
2:0
Value of left bit shift prior to F1 for channel A.
GAIN_B
3b
R/W
0
4
2:0
Value of left bit shift prior to F1 for channel B.
RATE
1B
R/W
1
5
7:0
Determines rate of serial output clock. The output rate is
FCK/(RATE+1). Unsigned integer values of 0, 1, 3, 7, 15, and 31
are allowed.
SOUT_EN
1b
R/W
0
6
0
Enables the serial output pins AOUT, BOUT, SCK, and SFS.
0=Tristate. 1=Enabled.
SCK_POL
1b
R/W
0
6
1
Determines polarity of the SCK output. 0=AOUT, BOUT, and
SFS change on the rising edge of SCK (capture on falling edge).
1=They change on the falling edge of SCK.
SFS_POL
1b
R/W
0
6
2
Determines polarity of the SFS output. 0=Active High. 1=Active
Low.
RDY_POL
1b
R/W
0
6
3
Determines polarity of the RDY output. 0=Active High. 1=Active
Low.
MUX_MODE
1b
R/W
0
6
4
Determines the mode of the serial outputs. 0=Each channel is
output on its respective pin, 1=Both channels are multiplexed and
output on AOUT.
PACKED
1b
R/W
0
6
5
Controls when SFS goes active. 0=SFS pulses prior to the start
of the I and the Q words. 1=SFS pulses only once prior to the
start of each I/Q sample pair (i.e. the pair is treated as a doublesized word) The I word precedes the Q word.
FORMAT
2b
R/W
0
6
7:6
Determines output number format. 0=Truncate serial output to 8
bits. Parallel output is truncated to 32 bits. 1=Round both serial
and parallel to 16-bits. All other bits are set to 0. 2=Round both
serial and parallel to 24-bits. All other bits are set to 0. 3=Output
floating point. 8-bit mantissa, 4-bit exponent. All other bits are set
to 0.
FREQ_A
4B
R/W
0
7-10
7:0
Frequency word for channel A. Format is a 32-bit, 2’s
complement number spread across 4 registers. The LSBs are in
the lower registers. The NCO frequency F is F/FCK=FREQ_A/
232.
(1)
These are the default values set by a master reset (MR). Sync in (SI) will not affect any of these values.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
45
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
Register Name
PHASE_A
FREQ_B
Width
2B
4B
Type
R/W
R/W
www.ti.com
Default (1)
0
0
Addr
11-12
13-16
Bit
Description
7:0
Phase word for channel A. Format is a 16-bit, unsigned
magnitude number spread across 2 registers. The LSBs are in
the lower registers. The NCO phase PHI is PHI=2*pi*PHASE_A/
216.
7:0
Frequency word for channel B. Format is a 32-bit, 2’s
complement number spread across 4 registers. The LSBs are in
the lower registers. The NCO frequency F is F/FCK=FREQ_B/
232.
2B
R/W
0
17-18
7:0
Phase word for channel B. Format is a 16-bit, unsigned
magnitude number spread across 2 registers. The LSBs are in
the lower registers. The NCO phase PHI is PHI=2*pi*PHASE_B/
216.
A_SOURCE
2
R/W
0
19
1:0
0=Select AIN as channel input source. 1=Select BIN. 2=3=Select
TEST_REG as channel input source.
B_SOURCE
2
R/W
1
19
3:2
0=Select AIN as channel input source. 1=Select BIN. 2=3=Select
TEST_REG as channel input source.
EXP_INH
1b
R/W
0
20
0
0=Allow exponent to pass into FLOAT TO FIXED converter.
1=Force exponent in DDC channel to a 7 (maximum digital gain).
This affects both channels.
Reserved
1b
R/W
1
20
1
AGC_FORCE on the CLC5902. Do not use.
Reserved
1b
R/W
0
20
2
AGC_RESET_EN on the CLC5902. Do not use.
AGC_HOLD_IC
1b
R/W
0
20
3
0=Normal closed-loop operation. 1=Hold integrator at initial
condition. This affects both channels.
AGC_LOOP_GAIN
2b
R/W
0
20
4:5
Bit shift value for AGC loop. Valid range is from 0 to 3. This
affects both channels.
Reserved
2B
R/W
0
21-22
7:0
AGC_COUNT on the CLC5902. Do not use.
AGC_IC_A
1B
R/W
0
23
7:0
AGC fixed gain for channel A. Format is an 8-bit, unsigned
magnitude number. The channel A DVGA gain will be set to the
inverted three MSBs.
AGC_IC_B
1B
R/W
0
24
7:0
AGC fixed gain for channel B. Format is an 8-bit, unsigned
magnitude number. The channel B DVGA gain will be set to the
inverted three MSBs.
AGC_RB_A
1B
R
0
25
7:0
AGC integrator readback value for channel A. Format is an 8-bit,
unsigned magnitude number. The user can read the magnitude
MSBs of the channel A integrator from this register.
AGC_RB_B
1B
R
0
26
7:0
AGC integrator readback value for channel B. Format is an 8-bit,
unsigned magnitude number. The user can read the magnitude
MSBs of the channel B integrator from this register.
7:0
5:0
Test input source. Instead of processing values from the
A|BIN pins, the value from this location is used
instead. Format is 14-bit 2s complement number spread across 2
registers.
PHASE_B
TEST_REG
14b
R/W
0
27(LSBs)
28(MSBs)
Reserved
1B
-
-
29
7:0
For future use.
Reserved
1B
-
-
30
7:0
For future use
DEBUG_EN
1b
R/W
0
31
0
46
0=Normal. 1=Enables access to the internal probe points.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
Register Name
SNWS019B – JULY 2007 – REVISED APRIL 2013
Width
Type
Default (1)
Addr
Bit
Description
Selects internal node tap for debug.
0 selects F1 output for BI, 20 bits
1 selects F1 output for BQ, 20 bits
2 selects F1 output for AQ, 20 bits
3 selects F1 output for AI, 20 bits
4 selects F1 input for BI, 20 bits
5 selects F1 input for BQ, 20 bits
6 selects F1 input for AI, 20 bits
7 selects F1 input for AQ, 20 bits
8 selects NCO A, cosine output. 17 bits, 3 LSBs are 0.
9 selects NCO A, sine output, 17 bits, 3 LSBs are 0.
10 selects NCO B, cosine output, 17 bits, 3 LSBs are 0.
11 selects NCO B, sine output, 17 bits, 3 LSBs are 0.
12 selects NCO AI, rounded output, 15 bits, 5 LSBs are 0.
13 selects NCO AQ, rounded output, 15 bits, 5 LSBs are 0.
14 selects NCO BI, rounded output, 15 bits, 5 LSBs are 0.
15 selects NCO BQ, rounded output, 15 bits, 5 LSBs are 0.
16 selects AGC CIC filter output. 9 MSBs from ch A, next 9 bits
from ch B, 2 LSBs are 0.
17-31 Reserved.
DEBUG_TAP
5b
R/W
0
31
5:1
DITH_A
1b
R/W
1
31
6
0=Disable NCO dither source for channel A. 1=Enable.
DITH_B
1b
R/W
1
31
7
0=Disable NCO dither source for channel B. 1=Enable.
32B
R/W
0
128-159
7:0
RAM space that defines key AGC loop parameters. Format is 32
separate 8-bit, 2’s complement numbers. This is common to both
channels.
7:0
Coefficients for F1. Format is 11 separate 16-bit, 2’s complement
numbers, each one spread across 2 registers. The LSBs are in
the lower registers. For example, coefficient h0[7:0] is in address
160, h0[15:8] is in address 161, h1[7:0] is in address 162,
h1[15:8] is in address 163. PAGE_SEL_F1=1 maps these
addresses to coefficient memory B.
AGC_TABLE
F1_COEFF
22B
R/W
0
160-181
64B
R/W
0
182-245
7:0
Coefficients for F2. Format is 32 separate 16-bit, 2’s complement
numbers, each one spread across 2 registers. The LSBs are in
the lower registers. For example, coefficient h0[7:0] is in address
182, h0[15:8] is in address 183, h1[7:0] is in address 184,
h1[15:8] is in address 185. PAGE_SEL_F2=1 maps these
addresses to coefficient memory B.
COEF_SEL_F1A
1b
R/W
0
246
0
Channel A F1 coefficient select register. 0=memory A, 1=memory
B.
COEF_SEL_F1B
1b
R/W
0
246
1
Channel B F1 coefficient select register. 0=memory A, 1=memory
B.
PAGE_SEL_F1
1b
R/W
0
246
2
F1 coefficient page select register. 0=memory A, 1=memory B.
COEF_SEL_F2A
1b
R/W
0
247
0
Channel A F2 coefficient select register. 0=memory A, 1=memory
B.
COEF_SEL_F2B
1b
R/W
0
247
1
Channel B F2 coefficient select register. 0=memory A, 1=memory
B.
PAGE_SEL_F2
1b
R/W
0
247
2
F2 coefficient page select register. 0=memory A, 1=memory B.
SFS_MODE
1b
R/W
0
248
0
0=SFS asserted at the start of each output word when
PACKED=1 or each I/Q pair when PACKED=0, 1=SFS asserted
at the start of each output sample period.
SDC_EN
1b
R/W
0
248
1
0=normal serial mode, 1=serial daisy-chain master mode.
AGC_COMB_ORD
2b
R/W
0
249
1:0
Enable reduced bandwidth AGC power detector. 0=2nd-order
decimate-by-eight CIC, 1=1-tap comb added to CIC, 2=4-tap
comb added to CIC.
EXT_DELAY
5b
R/W
0
249
6:2
Number of CK period delays needed to align the DVGA gain step
with the digital gain compensation step. Set this register to 7 if
ASTROBE and BSTROBE are not used. Otherwise set to 8.
F2_COEFF
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
47
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
Condensed LM97593 Address Map
Register Name Addr
Addr
Hex
Bit7
Bit6
Bit5
DEC
0
0x00
Dec7
Dec6
Dec5
DEC_BY_4
1
0x01
SCALE
2
0x02
GAIN_A
3
0x03
GAIN_B
4
0x04
RATE
5
0x05
Rate7
Rate6
Rate5
SERIAL_CTRL
6
0x06
FMT1
FMT0
Packed
FREQ_A
7
0x07
FA7
FA6
FA5
8
0x08
FA15
FA14
Bit4
Dec4
Bit3
Dec3
DecBy4
Sc5
Sc4
Sc3
Bit2
Bit1
Bit0
Dec2
Dec1
Dec0
Dec10
Dec9
Dec8
Sc2
Sc1
Sc0
GA2
GA1
GA0
GB2
GB1
GB0
Rate4
Rate3
Rate2
Rate1
Rate0
MuxMode
RDY_POL
SFS_POL
SCK_POL
SOUT_EN
FA4
FA3
FA2
FA1
FA0
FA13
FA12
FA11
FA10
FA9
FA8
9
0x09
FA23
FA22
FA21
FA20
FA19
FA18
FA17
FA16
10
0x0A
FA31
FA30
FA29
FA28
FA27
FA26
FA25
FA24
11
0x0B
PA7
PA6
PA5
PA4
PA3
PA2
PA1
PA0
12
0x0C
PA15
PA14
PA13
PA12
PA11
PA10
PA9
PA8
13
0x0D
FB7
FB6
FB5
FB4
FB3
FB2
FB1
FB0
14
0x0E
FB15
FB14
FB13
FB12
FB11
FB10
FB9
FB8
15
0x0F
FB23
FB22
FB21
FB20
FB19
FB18
FB17
FB16
16
0x10
FB31
FB30
FB29
FB28
FB27
FB26
FB25
FB24
17
0x11
PB7
PB6
PB5
PB4
PB3
PB2
PB1
PB0
18
0x12
PB15
PB14
PB13
PB12
PB11
PB10
PB9
PB8
Source
19
0x13
AGC_CTRL
20
0x14
AGC_COUNT
21
0x15
Reserved
Reserved
PHASE_A
FREQ_B
PHASE_B
BS1
BS0
AS1
AS0
AgcLG0
AgcHldlC
Reserved
Reserved
Explnh
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
AgcLG1
22
0x16
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
Reserved
AGC_IC_A
23
0x17
AgclcA7
AgclcA6
AgclcA5
AgclcA4
AgclcA3
AgclcA2
AgclcA1
AgclcA0
AGC_IC_B
24
0x18
AgclcB7
AgclcB6
AgclcB5
AgclcB4
AgclcB3
AgclcB2
AgclcB1
AgclcB0
AGC_RB_A
25
0x19
AgcRbA7
AgcRbA6
AgcRbA5
AgcRbA4
AgcRbA3
AgcRbA2
AgcRbA1
AgcRbA0
AGC_RB_B
26
0x1A
AGCRbB7 AGCRbB6
TEST_REG
27
0x1B
28
0x1C
DEBUG
31
0x1F
AGC_TABLE
128
0x80
159
0x9F
160
0xA0
181
0xB5
182
0xB6
245
0xF5
F1_CTRL
246
0xF6
PgSelF1
CfSelF1B
CfSelF1A
F2_CTRL
247
0xF7
PgSelF2
CfSelF2B
CfSelF2A
SERIAL_CTRL2
248
0xF8
SdcEn
SfsMode
AGC_CTRL2
249
0xF9
F1_COEFF
F2COEFF
48
Test7
DITH_B
Test6
DITH_A
AGCRbB5 AGCRbB4
AGCRbB3
AGCRbB2 AGCRbB1
AGCRbB0
Test5
Test4
Test3
Test2
Test1
Test0
Test13
Test12
Test11
Test10
Test9
Test8
TapSel3
TapSel2
TapSel1
TapSel0
DebugEnable
TapSel4
The AGC Table loads from the low address to the high address in this order:
"1st location, 2nd location..."
The FIR Coefficients load from the low address to the high address in this order:
"1st location low byte, 1st location high byte, 2nd location...""
The Page Select bits determine which set of coefficient memory is written.
ExtDelay4
ExtDelay3
ExtDelay2
Submit Documentation Feedback
ExtDelay1
ExtDelay0 CombOrd1
CombOrd0
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
AGC Theory of Operation
A block diagram of the AGC is shown in Figure 81. The DVGA interface comprises four pins for each of the
channels. The first three pins of this interface are a 3-bit binary word that controls the DVGA gain in 6dB steps
(AGAIN). The final pin is ASTROBE which allows the AGAIN bits to be latched into the DVGA’s register. A key
feature of the ASTROBE, illustrated Figure 82, is that it toggles only if the data on AGAIN has changed from the
previous cycle. Not shown is that ASTROBE and BSTROBE are independent. For example, ASTROBE only
toggles when AGAIN has changed. BSTROBE will not toggle because AGAIN has changed. This is done to
minimize unnecessary digital noise on the sensitive analog path through the DVGA. ASTROBE and BSTROBE
are asserted during MR and SI to properly initialize the DVGAs.
The absolute value circuit and the 2-stage, decimate-by-8 CIC filter comprise the power detection part of the
AGC. The power detector bandwidth is set by the CIC filter to FCK/8. The absolute value circuit doubles the
effective input frequency. This has the effect of reducing the power detector bandwidth from FCK/8 to FCK/16.
For a full-scale sinusoidal input, the absolute value circuit output is a dc value of 511*(2/π). Because the absolute
value circuit also generates undesired even harmonic terms, the CIC filter (response shown in Figure 80) is
required to remove these harmonics. The first response null occurs at FCK/8, where FCK is the clock frequency,
and the response magnitude is at least 25dB below the dc value from FCK/10 to 9FCK/10. Because the 2nd
harmonic from the absolute value circuit is about 10dB below the dc this means that the ripple in the detected
level is about 0.7dB or less for input frequencies between FCK/20 to 19FCK/20. Setting the AGC_COMB_ORD
register to either 1 or 2 will narrow the power detector’s bandwidth as shown in Figure 80.
Figure 80. Power detector filter response, 52 MHz
Figure 81. LM97593 AGC circuit, Channel A
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
49
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
Refer to Figure 7 for detailed timing information.
Figure 82. Timing diagram for AGC/DVGA interface, Channel A
The “FIXED TO FLOAT CONVERTER” takes the fixed point 9-bit output from the CIC filter and converts it to a
“floating point” number. This conversion is done so that the 32 values in the RAM can be uniformly assigned (dB
scale) to detected power levels (54 dB range). This provides a resolution of 1.7dB between detected power
levels. The truth table for this converter is given in Table 4. The upper three bits of the output represent the
exponent (e) and the lower 2 are the mantissa (m). The exponent is determined by the position of the leading ‘1’
out of the CIC filter. An output of ‘001XX’ corresponds to a leading ‘1’ in bit 2 (LSB is bit 0). The exponent
increases by one each time the leading ‘1’ advances in bit position. The mantissa bits are the two bits that follow
the leading ‘1’. If we define E as the decimal value of the exponent bits and M as the decimal value of the
mantissa bits, the output of the CIC filter, POUT, corresponding to a given “FIXED TO FLOAT CONVERTER”
output is:
(9)
The max() and min() operators account for row 1 of Table 4 which is a special case because M=POUT. Equation 9
associates each address of the RAM with a CIC filter output.
Table 4. Fixed to Float Converter Truth Table
INPUT
OUTPUT (eeemm)
0-3
000XX
4-7
001XX
8-15
010XX
16-31
011XX
32-63
100XX
64-127
101XX
128-255
110XX
256-511
111XX
As shown in Figure 81, the 32X8 RAM look-up table implements the functions of log converter, reference
subtraction, error amplifier, and deadband. The user must build each of these functions by constructing a set of
8-bit, 2’s complement numbers to be loaded into the RAM. Each of these functions and how to construct them
are discussed in the following paragraphs.
A log conversion is done in order to keep the loop gain independent of operating point. To see why this is
beneficial, the control gain of the DVGA computed without log conversion is:
where
•
•
50
G is the decimal equivalent of GAIN
Go accounts for the DVGA gain in excess of unity
Submit Documentation Feedback
(10)
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
This equation assumes that the DVGA gain control polarity is positive as is the case for the CLC5526. The gain
around the entire loop must be negative. Observe in Equation 10 that the control gain is dependent on operating
point G. If we instead compute the control gain with log conversion then:
(11)
which is no longer operating-point dependent. The log function is constructed by computing the CIC filter output
associated with each address (Equation 12) and converting these to dB. Full scale (dc signal) is
20log(511)=54dB.
The reference subtraction is constructed by subtracting the desired loop servo point (in dB) from the table values
computed in the previous paragraph. For example, if it is desired that the DVGA servo the ADC input level
(sinusoidal signal) to -6dBFS, the number to subtract from the data is:
(12)
The table data will then cross through zero at the address corresponding to this reference level. A deadband
wider than 6dB should then be constructed symmetrically about this point. This prevents the loop from hunting
due to the 6dB gain steps of the DVGA. Any deadband in excess of 6dB appears as hysteresis in the servo point
of the loop as illustrated in Figure 78. The deadband is constructed by loading zeros into those addresses on
either side of the one which corresponds to the reference level.
The last function of the RAM table is that of error amplification. All the operations preceding this one gave a table
slope SRAM = 1. This must now be adjusted in order to control the time constant of the loop given by:
(13)
The term GL in this equation is the loop gain:
(14)
The design equations are obtained by solving Equation 13 for GL and Equation 14 for SRAM. AGC_LOOP_GAIN
is a control register value that determines the number of bits to shift the output of the RAM down by. This allows
some of the loop gain to be moved out of the RAM so that the full output range of the table is utilized but not
exceeded. The valid range for AGC_LOOP_GAIN is from 0 to 3 which corresponds to a 0 to 3 bit shift to the left.
An example set of numbers to implement a loop having a reference of 6dB below full scale, a deadband of 8dB,
and a loop gain of 0.108 is:
-102 -102 -88 -80 -75 -70 -66
-63 -61 -56 -53 -50
-47 -42 -39 -36 -33 -29 -25
-22 -19 -15 -11 -0
0 0 0 0 0 13 17 20
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
51
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
20
AGC RAM CONTENTS
0
-20
-40
-60
-80
-100
-120
0
5
10
15
20
25
30
ADDRESS
Figure 83. Example of programmed RAM contents
20
AGC RAM CONTENTS
0
-20
-40
-60
-80
-100
-120
0
10
20
30
40
50
60
POUT (dB)
Figure 84. Example of programmed RAM contents
Table 5. 15-bit Mixer Output Alignment into the 22-bit SHIFT-UP Based On EXP
AGAINa
EXPb
Inputc
21d
20
19
18
17
16
15
14
...
8
7
6
5
4
3
2
1
0
000 = -12dB
111 = +0dB
-12dB
14
13
12
11
10
9
8
7
...
1
0
L
L
L
L
L
L
L
001 = -6dB
110 = -6dB
-12dB
14
14
13
12
11
10
9
8
...
2
1
0
L
L
L
L
L
L
010 = 0dB
101 = -12dB
-12dB
14
14
14
13
12
11
10
9
...
3
2
1
0
L
L
L
L
L
011 = +6dB
100 = -18dB
-12dB
14
14
14
14
13
12
11
10
...
4
3
2
1
0
L
L
L
L
100 = +12dB
011 = -24dB
-12dB
14
14
14
14
14
13
12
11
...
5
4
3
2
1
0
L
L
L
101 = +18dB
010 = -30dB
-12dB
14
14
14
14
14
14
13
12
...
6
5
4
3
2
1
0
L
L
110 = +24dB
001 = -36dB
-12dB
14
14
14
14
14
14
14
13
...
7
6
5
4
3
2
1
0
L
111 = +30dB
000 = -42dB
-12dB
14
14
14
14
14
14
14
14
...
8
7
6
5
4
3
2
1
0
These values are shown with respect to the table addresses in Figure 83, and the CIC filter output POUT in
Figure 84. For a 52MHz clock rate and AGC_LOOP_GAIN=2, these values result in a loop time constant of
1.5µs.
The error signal from the loop gain “SHIFT DOWN” circuit is gated into the loop integrator. A MUX within the
integrator feedback allows the integrator to be initialized to the value loaded into AGC_IC_A (channel B can be
set independently). The top eight bits of the integrator output can also be read back over the microprocessor
interface from the AGC_RB_A (or AGC_RB_B) register. The top 3 bits become AGAIN and are output along with
the ASTROBE signal on the DVGA interface pins. The valid range of AGAIN is from 0 to 7 which corresponds to
a valid range of 0 to 211-1 for the 11-bit, 2’s complement integrator output from which AGAIN is derived. This is
illustrated in Figure 85. The integrator saturates at these limits to prevent overshoots as the integrator attempts to
enter the valid range. The AGAIN value is inverted (EXP) and used to adjust the gain of the incoming signal to
provide a linear output dynamic range. The relationship between the DVGA analog gain (AGAIN) and the “FIXED
TO FLOAT CONVERTER” digital gain (EXP) is shown in Table 5. The DVGA’s compression of the incoming
signal in the analog domain vs. the subsequent expansion in the digital domain is shown in Figure 77.
52
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
The AGC may be forced to free run by setting AGC_HOLD_IC low. Writing an initial condition to AGC_IC_A|B
and then setting AGC_HOLD_IC high will force the AGC to a fixed gain. The three MSBs of the value written to
AGC_IC_A|B are inverted and output to drive the DVGA.
Allowing the AGC to free run should be appropriate for most applications. If the INH_EXP bit is not set, the
DVGA gain word (EXP) is routed to the “FLOAT TO FIXED CONVERTER” in the DDCs prior to the
programmable decimation filter. The EXP signals are delayed to account for the propagation delay of the DVGA
interface and the ADC12DL080 ADC.
For this range to be
the same size as all
others, the max
integrator output must
8
11
be limited to 8x2 - 1 = 2 - 1
7
6
A GAIN
5
4
3
2
The min integrator
output must be
limited to 0 so that
the sign of AGAIN
is positive.
1
0
0
8
1x2
2x2
8
3x2
8
4x2
8
5x2
8
6x2
8
7x2
8
8
8x2
INTEGRATOR OUTPUT
Figure 85. AGC integrator output limits
General Applications Information
OUTPUTS
Be very careful when driving a high capacitance bus. The more capacitance the output drivers must charge for
each conversion, the more instantaneous digital current flows through VDR and DRGND. These large charging
current spikes can cause on-chip ground noise and couple into the analog circuitry, degrading dynamic
performance. Adequate bypassing, limiting output capacitance and careful attention to the ground plane will
reduce this problem.
To minimize noise due to output switching, minimize the load currents at the digital outputs. Only one driven
input should be connected to each output pin. Additionally, inserting series resistors of about 100Ω at the digital
outputs, close to the ADC pins, will isolate the outputs from trace and other circuit capacitances and limit the
output currents, which could otherwise result in performance degradation. See Figure 86.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
53
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
FILTER COMPONENT VALUES
FREQUENCY
150 MHz
250 MHz
L1
43 nH
22 nH
18 pF
43 pF
C1
18 pF
43 pF
C2
0.5 pF
C3
-
24
0.01
PF
0.1
PF
23
10 PF 22
0.1
PF
15
0.01
PF
0.1
PF
0.1
PF
T3
M/A-COM PART NO.
ETK4-2T
VRNA
0.01
PF
1 nF
C1B
500Ö
27
26
500Ö
C2B
3
CLC5526
DVGA
REMOVE GND & POWER PLANES
FROM UNDER THESE COMPONENTS
& SIGNALS AS MUCH AS POSSIBLE
T3
M/A-COM PART NO.
ETK4-2T
VINBVINB+
LM97593
0.01
PF
REMOVE GND & POWER PLANES
FROM UNDER THESE COMPONENTS
& SIGNALS AS MUCH AS POSSIBLE
CLC5526
DVGA
1
1 nF
A/R
C3B
L1B
1 nF
2
4
BGAIN[2:0]
BANDPASS FILTER
2
5
40, 41, 42
1
A/R
VIN_A
VCOMA
0.1 PF
3
4
VRNB
16
VRPA
10 PF 17
Æ PCB parasitics may require component value adjustment.
Æ See LM97593 (% XVHU¶V JXLGH IRU FRPSOHWH GHWDLOV.
5
VRPB
0.1 PF
Note:
VIN_B
VCOMB
0.01
PF
500Ö
C1A
C3A
13
14
L1A
500Ö
C2A
3
1 nF
0.01
PF
VINAVINA+
BANDPASS FILTER
3
127, 126, 125
AGAIN[2:0]
Figure 86. Application Circuit
POWER SUPPLY CONSIDERATIONS
The power supply pins should be bypassed with a 10 µF capacitor and with a 0.1 µF ceramic chip capacitor near
each power pin. Leadless chip capacitors are preferred because they have low series inductance.
The LM97593 is sensitive to power supply noise. Accordingly, the noise on the analog supply pins should be
kept below 100 mVP-P.
No pin should ever have a voltage on it that is in excess of the supply voltages, not even on a transient basis. Be
especially careful of this during power turn on and turn off.
The VDR pin provides power for the output drivers and should be operated from a supply equivalent to VD.
LAYOUT AND GROUNDING
A proper grounding and proper routing of all signals are essential to ensure accurate conversion. Maintaining
separate analog and digital areas of the board, with the LM97593 between these areas, is required to achieve
specified performance.
The ground return for the data outputs (DRGND) carries the ground current for the output drivers. The output
current can exhibit high transients that could add noise to the conversion process. To prevent this from
happening, the DRGND pins should NOT be connected to system ground in close proximity to any of the
LM97593's other ground pins.
54
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
Capacitive coupling between the typically noisy digital circuitry and the sensitive analog circuitry can lead to poor
performance. The solution is to keep the analog circuitry separated from the digital circuitry, and to keep the
clock line as short as possible.
Digital circuits create substantial supply and ground current transients. The logic noise thus generated could
have significant impact upon system noise performance. The best logic family to use in systems with A/D
converters is one which employs non-saturating transistor designs, or has low noise characteristics, such as the
74LS, 74HC(T) and 74AC(T)Q families. The worst noise generators are logic families that create the largest
supply current transients during clock or signal edges, like the 74F and the 74AC(T) families.
The effects of the noise generated from the LM97593 output switching can be minimized through the use of
100Ω resistors in series with each data output line. Locate these resistors as close to the LM97593 output pins
as possible.
Since digital switching transients are composed largely of high frequency components, total ground plane copper
weight will have little effect upon the logic-generated noise. This is because of the skin effect. Total surface area
is more important than is total ground plane volume.
Generally, analog and digital lines should cross each other at 90° to avoid crosstalk. To maximize accuracy in
high speed, high resolution systems, however, avoid crossing analog and digital lines altogether. It is important to
keep clock lines as short as possible and isolated from ALL other lines, including other digital lines. Even the
generally accepted 90° crossing should be avoided with the clock line as even a little coupling can cause
problems at high frequencies. This is because other lines can introduce jitter into the clock line, which can lead to
degradation of SNR. Also, the high speed clock can introduce noise into the analog chain.
Best performance at high frequencies and at high resolution is obtained with a straight signal path. That is, the
signal path through all components should form a straight line wherever possible.
Be especially careful with the layout of inductors. Mutual inductance can change the characteristics of the circuit
in which they are used. Inductors should not be placed side by side, even with just a small part of their bodies
beside each other.
The analog input should be isolated from noisy signal traces to avoid coupling of spurious signals into the input.
Any external component (e.g., a filter capacitor) connected between the converter's input pins and ground or to
the reference input pin and ground should be connected to a very clean point in the ground plane.
All analog circuitry (input amplifiers, filters, reference components, etc.) should be placed in the analog area of
the board. All digital circuitry and I/O lines should be placed in the digital area of the board. The LM97593 should
be between these two areas. Furthermore, all components in the reference circuitry and the input signal chain
that are connected to ground should be connected together with short traces and enter the ground plane at a
single, quiet point. All ground connections should have a low inductance path to ground.
DYNAMIC PERFORMANCE
To achieve the best dynamic performance, the clock source driving the CLK input must be free of jitter. Isolate
the ADC clock from any digital circuitry with buffers, as with the clock tree shown in Figure 87. The gates used in
the clock tree must be capable of operating at frequencies much higher than those used if added jitter is to be
prevented.
Best performance will be obtained with a differential input drive, compared with a single-ended drive, as
discussed in Single-Ended Operation and Driving the Analog Inputs.
It is good practice to keep the clock line as short as possible and to keep it well away from any other signals.
Other signals can introduce jitter into the clock signal, which can lead to reduced SNR performance, and the
clock can introduce noise into other lines. Even lines with 90° crossings have capacitive coupling, so try to avoid
even these 90° crossings of the clock line.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
55
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
Figure 87. Isolating the ADC Clock from other Circuitry with a Clock Tree
Common Application Pitfalls
Driving the inputs (analog or digital) beyond the power supply rails. For proper operation, all inputs should
not go more than 100 mV beyond the supply rails (more than 100 mV below the ground pins or 100 mV above
the supply pins). Exceeding these limits on even a transient basis may cause faulty or erratic operation. It is not
uncommon for high speed digital components (e.g., 74F and 74AC devices) to exhibit overshoot or undershoot
that goes above the power supply or below ground. A resistor of about 47Ω to 100Ω in series with any offending
digital input, close to the signal source, will eliminate the problem.
Do not allow input voltages to exceed the supply voltage, even on a transient basis. Not even during power up or
power down.
Be careful not to overdrive the inputs of the LM97593 with a device that is powered from supplies outside the
range of the LM97593 supply. Such practice may lead to conversion inaccuracies and even to device damage.
Attempting to drive a high capacitance digital data bus. The more capacitance the output drivers must
charge for each conversion, the more instantaneous digital current flows through VDR and DRGND. These large
charging current spikes can couple into the analog circuitry, degrading dynamic performance. Adequate
bypassing and maintaining separate analog and digital areas on the pc board will reduce this problem.
The digital data outputs should be buffered (with 74ACQ541, for example) if they will drive a large capacitive
load. Dynamic performance can also be improved by adding series resistors at each digital output, close to the
LM97593, which reduces the energy coupled back into the part's output pins by limiting the output current. A
reasonable value for these resistors is 100Ω.
Using an inadequate amplifier to drive the analog input. As explained in Signal Inputs, the capacitance seen
at the input alternates between 8 pF and 7 pF, depending upon the phase of the clock. This dynamic load is
more difficult to drive than is a fixed capacitance. If the amplifier exhibits overshoot, ringing, or any evidence of
instability, even at a very low level, it will degrade performance. A small series resistor at each amplifier output
and a capacitor at the analog inputs will improve performance.
Also, it is important that the signals at the two inputs have exactly the same amplitude and be exactly 180º out of
phase with each other. Board layout, especially equality of the length of the two traces to the input pins, will
affect the effective phase between these two signals. Remember that an operational amplifier operated in the
non-inverting configuration will exhibit more time delay than will the same device operating in the inverting
configuration.
Operating with the reference pins outside of the specified range. As mentioned in Reference Pins, VREF
should be in the range of
0.8V ≤ VREF ≤ 1.2V
(15)
Operating outside of these limits could lead to performance degradation.
Inadequate network on Reference Bypass pins (VRPA, VRNA, VCOMA, VRPB, VRNB and VCOMB). As mentioned
in Reference Pins, these pins should be bypassed with 0.1 µF capacitors to ground at VRMA and VRMB and with
a 10 µF between pins VRPA and VRNA and between VRPB and VRNB for best performance.
Using a clock source with excessive jitter, using excessively long clock signal trace, or having other
signals coupled to the clock signal trace. This will cause the sampling interval to vary, causing excessive
output noise and a reduction in SNR and SINAD performance.
56
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
LM97593
www.ti.com
SNWS019B – JULY 2007 – REVISED APRIL 2013
Evaluation Hardware
Evaluation boards are available to facilitate designs based on the LM97593:
LM97593EB
The LM97593 evaluation board provides a complete narrowband receiver from IF to digital symbols.
SOFTWARE
Control panel software for the LM97593 supports complete device configuration on both evaluation boards.
Integrated capture software manages the capture of data and its storage in a file on a PC.
Matlab script files support data analysis: FFT, DNL, and INL plotting.
This software and additional application information is available on the Basestation CDROM.
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
57
LM97593
SNWS019B – JULY 2007 – REVISED APRIL 2013
www.ti.com
REVISION HISTORY
Changes from Revision A (April 2013) to Revision B
•
58
Page
Changed layout of National Data Sheet to TI format .......................................................................................................... 57
Submit Documentation Feedback
Copyright © 2007–2013, Texas Instruments Incorporated
Product Folder Links: LM97593
PACKAGE OPTION ADDENDUM
www.ti.com
10-Dec-2020
PACKAGING INFORMATION
Orderable Device
Status
(1)
Package Type Package Pins Package
Drawing
Qty
Eco Plan
(2)
Lead finish/
Ball material
MSL Peak Temp
Op Temp (°C)
Device Marking
(3)
(4/5)
(6)
LM97593VH/NOPB
ACTIVE
QFP
NND
128
66
RoHS & Green
SN
Level-3-245C-168 HR
-40 to 85
LM97593VH
(1)
The marketing status values are defined as follows:
ACTIVE: Product device recommended for new designs.
LIFEBUY: TI has announced that the device will be discontinued, and a lifetime-buy period is in effect.
NRND: Not recommended for new designs. Device is in production to support existing customers, but TI does not recommend using this part in a new design.
PREVIEW: Device has been announced but is not in production. Samples may or may not be available.
OBSOLETE: TI has discontinued the production of the device.
(2)
RoHS: TI defines "RoHS" to mean semiconductor products that are compliant with the current EU RoHS requirements for all 10 RoHS substances, including the requirement that RoHS substance
do not exceed 0.1% by weight in homogeneous materials. Where designed to be soldered at high temperatures, "RoHS" products are suitable for use in specified lead-free processes. TI may
reference these types of products as "Pb-Free".
RoHS Exempt: TI defines "RoHS Exempt" to mean products that contain lead but are compliant with EU RoHS pursuant to a specific EU RoHS exemption.
Green: TI defines "Green" to mean the content of Chlorine (Cl) and Bromine (Br) based flame retardants meet JS709B low halogen requirements of