Si3220/Si3225
P R E L I M I N A R Y D A TA S H E E T
D U A L P R O SLIC™ P R O G R A M M A B L E C M O S S L I C / C O D E C
Features
!
!
!
!
!
!
!
Performs all BORSCHT functions
!
Ideal for applications up to 18 kft
!
Internal balanced ringing to 65 Vrms
!
(Si3220)
!
External bulk ringer support (Si3225)
!
Low standby power consumption:
!
61 mA will disable threshold
detection.
Ground Key Detection
LONGLPF = [(2πf x 4096)/800]
Ground Key detection detects an alerting signal from
the terminal equipment during the Active linefeed state
(forward or reverse polarity). The functional blocks
required to implement a Ground Key detector are
shown in Figure 21, and the register set for detecting a
ground key event is provided in Table 22. The primary
input to the system is the Longitudinal Current Sense
value provided by the voltage/current/power monitoring
circuitry and reported in the ILONG RAM address. The
ILONG value is processed in the ISP provided the LFS
bits in the Linefeed register indicate the device is in an
Active state. The output of the ISP is the input to a
programmable digital low-pass filter, which removes
unwanted ac signal components before threshold
detection.
Where f = the desired cutoff frequency of the filter.
The low-pass filter coefficient is calculated using the
equation below and is entered into the LONGLPF RAM
location.
ILONG
Input
Signal
Processor
Digital
LPF
The programmable range of the filter is from 0 (blocks
all signals) to 4000h (unfiltered). A typical value of
10 Hz (0A10h) is sufficient to filter out any unwanted ac
artifacts while allowing the dc information to pass
through the filter.
The output of the low-pass filter is compared to the
programmable threshold, LONGHITH. Hysteresis is
enabled by programming a second threshold,
LONGLOTH, to detect when the ground key is released.
The threshold comparator output feeds a programmable
debounce filter. The output of the debounce filter
remains in its present state unless the input remains in
the opposite state for the entire period of time
programmed by the loop closure debounce interval,
LONGDBI. If the debounce interval is satisfied, the
LONGHI bit is set to indicate that a valid loop closure
has occurred.
+
Debounce
Filter
LONGHI
–
LONGLPF
LONGDBI
LFS
Interrupt
Logic
LONGS
Ground Key
Threshold
LONGE
LONGHITH LONGLOTH
Figure 21. Ground Key Detection Circuitry
Preliminary Rev. 0.91
37
Si3220/Si3225
Table 22. Register and RAM Locations Used for Ground Key Detection
Parameter
Register/
RAM
Mnemonics
Register/RAM
Bits
Programmable
Range
LSB
Size
Resolution
Ground Key Interrupt
Pending
IRQVEC2
LONGS
Yes/No
N/A
N/A
Ground Key Interrupt
Enable
IRQEN2
LONGE
Yes/No
N/A
N/A
Ground Key Linefeed
Shadow
LINEFEED
LFS[2:0]
Monitor only
N/A
N/A
Ground Key Detect Status
LCRRTP
LONGHI
Monitor only
N/A
N/A
Ground Key Detect
Debounce Interval
LONGDBI
LONGDBI[15:0]
0 to 40.96 s
1.25 ms
1.25 ms
ILONG
ILONG[15:0]
Monitor only
Ground Key Threshold
(enabled)
LONGHITH
LONGHITH[15:0]
0 to
101.09 mA*
3.097 µA
396.4 µA
Ground Key Threshold
(released)
LONGLOTH
LONGLOTH[15:0]
0 to
101.09 mA*
3.097 µA
396.4 µA
LONGLPF
LONGLPF[15:3]
0 to 4000h
N/A
N/A
Longitudinal Current Sense
Ground Key Filter Coefficient
See Table 19
*Note: The usable range for LONGHITH and LONGLOTH is limited to 16 mA. Setting a value > 16 mA will disable
threshold detection.
Automatic Dual Battery Switching
The Dual ProSLIC chipsets provide the ability to switch
between several user-provided battery supplies to aid
thermal management. Two specific scenarios where
this method may be required are as follows:
!
Ringing to off-hook state transition (Si3220):
During the on-hook operating state, the Dual
ProSLIC chipset must operate from the ringing
battery supply to provide the desired ringing signal
when required. Once an off-hook condition is
detected, the Dual ProSLIC chipset must transition
to the lower battery supply, typically –24 V, to reduce
power dissipation during the active state. The low
current consumed by the Dual ProSLIC chipset
during the on-hook state results in very little power
dissipation while being powered from the ringing
battery supply, which can have an amplitude as high
as –100 V depending on the desired ringing
amplitude.
! On-hook to off-hook state, short loop feed
(Si3225): When sourcing both long and short loop
lengths, the Dual ProSLIC chipset can automatically
switch from the typical –48 V off-hook battery supply
to a lower off-hook battery supply (e.g., –24 V) to
38
reduce the total off-hook power dissipation. The Dual
ProSLIC chipset continuously monitors the TIPRING voltage and selects the lowest battery voltage
required to power the loop when transitioning from
the on-hook to the off-hook state, thus assuring the
lowest power dissipation.
The BATSELa and BATSELb pins switch between the
two battery voltages based on the operating state and
the TIP-RING voltage. Figure 22 illustrates the chip
connections required to implement an automatic dual
battery switching scheme. When BATSEL is pulled
LOW, the desired channel is powered from the VBATL
supply. When BATSEL is pulled HIGH, the VBATH
source supplies power to the desired channel.
The BATSEL pins for both channels are controlled using
the BATSEL bit of the RLYCON register and can be
programmed to automatically switch to the lower battery
supply (VBATL) when the off-hook TIP-RING voltage is
low enough to allow proper operation from the lower
supply. When using the Si3220, this mode should
always be enabled to allow seamless switching
between the ringing and off-hook states. The same
switching scheme is used with the Si3225 to reduce
power by switching to a lower off-hook battery when
sourcing a short loop.
Preliminary Rev. 0.91
Si3220/Si3225
Two thresholds are provided to enable battery switching
with hysteresis. The BATHTH RAM location specifies
the threshold at which the Dual ProSLIC device
switches from the low battery (VBATL) to the high battery
(VBATH) due to an off-hook to on-hook transition. The
BATLTH RAM location specifies the threshold at which
the Si3220/Si3225 switches from VBATH to VBATL due to
a transition from the on-hook or ringing state to the offhook state or because the overhead during active OffHook mode is sufficient to feed the subscriber loop
using a lower battery voltage.
The low pass filter coefficient is calculated using the
equation below and is entered into the BATLPF RAM
location.
BATLPF = [(2πf x 4096)/800]
Where f = the desired cutoff frequency of the filter
The programmable range of the filter is from 0 (blocks
all signals) to 4000h (unfiltered). A typical value of
10 Hz (0A10h) is sufficient to filter out any unwanted ac
artifacts while allowing the dc information to pass
through the filter.
Table 23 provides the register and RAM locations used
for programming the battery switching functions.
Table 23. Register and RAM Locations Used for Battery Switching
Parameter
Register/RAM
Mnemonic
Register/RAM
Bits
Programmable
Range
Resolution
(LSB Size)
Battery Select Switch
RLYCON
BATSEL
Toggle
N/A
High Battery Detect Threshold
BATHTH
BATHTH[14:7]
0 to 160.173 V*
628 mV
(4.907 mV)
Low Battery Detect Threshold
BATLTH
BATLTH[14:7]
0 to 160.173 V*
628 mV
(4.907 mV)
Ringing Battery Switch (Si3220 only)
RLYCON
GPO
Toggle
N/A
Battery Select Indicator
RLYCON
BSEL
Toggle
N/A
Battery Switching LPF
BATLPF
BATLPF[15:3]
0 to 4000h
N/A
*Note: Usable range for BATHTH and BATLTH is limited to VBATH.
Si3220
Si3225
SVBAT
Battery
Control
Logic
Battery
Sense
Circuit
BATSEL
40.2 kΩ
806 kΩ
Si3200
Linefeed
Circuitry
VBATL
BATSEL
Battery
Select
Control
VBATL
VBAT
VBATH
VBATH
Figure 22. External Battery Switching Using the Si3220/Si3225
Preliminary Rev. 0.91
39
Si3220/Si3225
When generating a high-voltage ringing amplitude using
the Si3220, the power dissipated during the OHT state
typically increases due to operating from the ringing
battery supply in this mode. To reduce power, the
Si3220/Si3200 chipset provides the ability to
accommodate up to three separate battery supplies by
implementing a secondary battery switch using a few
low-cost external components as illustrated in Figure
22. The Si3220’s BATSEL pin is used to switch between
the VBATH (typically –48 V) and VBATL (typically
–24 V) rails using the switch internal to the Si3200. The
Si3220’s GPO pin is used along with the external
transistor circuit to switch the VRING rail (the ringing
voltage battery rail) onto the Si3200’s VBAT pin when
ringing is enabled. The GPO signal is driven
automatically by the ringing cadence provided that the
RRAIL bit of the RLYCON register is set to 1 (signifying
that a third battery rail is present).
Table 24. 3-Battery Switching Components
Component
Value
Comments
R102
10 kΩ,1/10 W, ± 5%
R103
402 kΩ,1/10 W,± 1%
Ringing Generation
The Si3220-based Dual ProSLIC chipset provides a
balanced ringing waveform, with or without dc offset.
The ringing frequency, cadence, waveshape, and dc
offset are register programmable.
Using a balanced ringing scheme, the ringing signal is
applied to both the TIP and the RING lines using ringing
waveforms that are 180° out of phase with each other.
The resulting ringing signal seen across TIP-RING is
twice the amplitude of the ringing waveform on either
the TIP or the RING line, which allows the ringing
circuitry to withstand half the total ringing amplitude
seen across TIP-RING.
Si3220
806 kΩ
VRING
GPO
BATSEL
SVBAT
RING
0.1 µF
Si3200
R101
VBAT
VBAT
R9
40.2 kΩ
VBATH
VBATH
R6
40.2 kΩ
CXT5401
VBATL
BATSEL
SLIC
Q1
0.1 µF
VBATL
VOFF
R102
TIP
10 kΩ
402 kΩ
Q2
IN4003
VTIP
R103
D1
CXT5551
Figure 23. 3-Battery Switching with Si3220/
Si3200
GND
VCM
VTIP
V PK
Table 24. 3-Battery Switching Components
VOFF
Component
Value
Comments
D1
200 V, 200 mA
1N4003 or similar
Q1
100 V PNP
CXT5401 or
similar
Q2
100 V NPN
CXT5551 or
similar
Figure 24. Balanced Ringing
R101
1/10 W, ± 5%
2.4 kΩ for
VDD=3.3 V
3.9 kΩ for
VDD=5 V
An internal ringing scheme provides >40 Vrms into a
5REN load at the terminal equipment using a userprovided ringing battery supply. The specific ringing
supply voltage required depends on the ringing voltage
desired. The ringing amplitude at the terminal
equipment depends on the loop impedance as well and
40
VRING
VOV
VBATH
Preliminary Rev. 0.91
Si3220/Si3225
the load impedance in REN. The following equation can
be used to determine the TIP-RING ringing amplitude
required for a specific load and loop condition.
R LOOP = ( 0.09Ω per foot for 26AWG wire )
R OUT = 320Ω
RLOOP
+
ROUT
7000Ω
R LOAD = -----------------#REN
RLOAD
VRING
VTERM
–
Figure 25. Simplified Loop Circuit During
Ringing
R LOAD
V TERM = V RING × --------------------------------------------------------------------( R LOAD + R LOOP + R OUT )
When ringing longer loop lengths, adding a dc offset
voltage is necessary to reliably detect a ring trip
condition (off-hook phone). Adding dc offset to the
ringing signal decreases the maximum possible ringing
amplitude. Adding significant dc offset also increases
the power dissipation in the Si3200 and may require
additional airflow or modified PCB layout to maintain
acceptable operating temperatures in the line feed
circuitry. The Dual ProSLIC chipset automatically
applies and removes the ringing signal during VOCcrossing periods to reduce noise and crosstalk to
adjacent lines. Table 25 provides a list of registers
required for internal ringing generation
where
Table 25. Register and RAM Locations Used for Ringing Generation
Parameter
Register/
RAM
Mnemonic
Register/RAM
Bits
Programmable
Range
Resolution
(LSB Size)
Ringing Waveform
RINGCON
TRAP
Sinusoid/Trapezoid
N/A
Ringing Active Timer Enable
RINGCON
TAEN
Enabled/Disabled
N/A
Ringing Inactive Timer Enable
RINGCON
TIEN
Enabled/Disabled
N/A
Ringing Oscillator Enable
Monitor
RINGCON
RINGEN
Enabled/Disabled
N/A
Ringing Oscillator Active Timer
RINGTALO/
RINGTAHI
RINGTA[15:0]
0 to 8.19 s
125 µs
Ringing Oscillator Inactive
Timer
RINGTILO/
RINGTIHI
RINGTI[15:0]
0 to 8.19 s
125 µs
Linefeed Control
(Initiates Ringing State)
LINEFEED
LF[2:0]
000 to 111
N/A
On-Hook Line Voltage
VOC
VOC[15:0]
0 to 63.3 V
1.005 V
(4.907 mV)
Ringing Voltage Offset
RINGOF
RINGOF[15:0]
0 to 63.3 V
1.005 V
(4.907 mV)
Ringing Frequency
RINGFRHI/
RINGFRLO
RINGFRHI[14:3]/
RINGFRLO[14:3]
4 to 100 Hz
Ringing Amplitude
RINGAMP
RINGAMP[15:0]
0 to 160.173 V
Preliminary Rev. 0.91
628 mV
(4.907 mV)
41
Si3220/Si3225
Table 25. Register and RAM Locations Used for Ringing Generation (Continued)
Parameter
Register/
RAM
Mnemonic
Register/RAM
Bits
Ringing Initial Phase
Sinusoidal
Trapezoid
External Ringing
RINGPHAS
RINGPHAS[15:0]
Ringing Relay Driver Enable
(Si3225 only)
RELAYCON
VOVRING
Ringing Overhead Voltage
Programmable
Range
Resolution
(LSB Size)
N/A
0 to 1.024 s
0 to 662.83 mA
N/A
31.25 µs
2.6 mA (20.3 µA)
RDOE
Enabled/Disabled
N/A
VOVRING[15:0]
0 to 63.3 V
1.005 V
(4.907 mV)
Internal Sinusoidal Ringing
A sinusoidal ringing waveform is generated by the onchip digital tone generator. The tone generator used to
generate ringing tones is a two-pole resonator with a
programmable frequency and amplitude. Since ringing
frequencies are low compared to the audio band
signaling frequencies, the sinusoid is generated at a
1 kHz rate. The ringing generator is programmed via the
RINGFREQ, RINGAMP, and RINGPHAS registers. The
equations are as follows:
2πf
coeff = cos ---------------------
1000Hz
DesiredV PK
1 1 – coeff
15
RINGAMP = --- ----------------------- × ( 2 ) × --------------------------------4 1 + coeff
160.173V
allow on/off cadence settings up to 8 s on/8 s off. In
addition to controlling ringing cadence, these timers
control the transition into and out of the ringing state.
To initiate ringing, the user must program the
RINGFREQ, RINGAMP, and RINGPHAS RAM
addresses as well as the RINGTA and RINGTI registers,
and select the ringing waveshape and dc offset. After
this is done, TAEN and TIEN bits are set as desired.
Ringing state is invoked by a write to the linefeed
register. At the expiration of RINGTA, the Dual ProSLIC
turns off the ringing waveform and goes to the on-hook
transmission state. At the expiration of RINGTI, ringing
is initiated again. This process continues as long as the
two timers are enabled and the linefeed register
remains in the ringing state.
Internal Trapezoidal Ringing
RINGPHAS = 0
For example, to generate a 60 Vrms (87 VPK), 20 Hz
ringing signal, the equations are as follows:
23
RINGFREQ = coeff ( 2 )
2π20
coeff = cos --------------------- = 99211
1000Hz
RINGPHAS = 4 x Period x 8000
RINGAMP = (Desired V/160.8 V) x (215)
23
RINGFREQ = 99211 × ( 2 ) = 8322461 =
0x7EFD9D
1 00789
85
15
RINGAMP = --- -------------------- × ( 2 ) × --------------------- = 273 = 0x111
4 1.99211
160.173
In addition to the variable frequency and amplitude, a
selectable dc offset (VOFF), which can be added to the
waveform is included. The dc offset is defined in the
RINGOF RAM location.
As with the tone generators, the ringing generator has
two timers which function as described above. They
42
In addition to the traditional sinusoidal ringing
waveform, the Dual ProSLIC can generate a trapezoidal
ringing waveform similar to the one illustrated in
Figure 26.
The
RINGFREQ,
RINGAMP,
and
RINGPHAS RAM addresses are used for programming
the ringing wave shape as follows:
RINGFREQ = (2 x RINGAMP)/(tRISE x 8000)
RINGFREQ is a value that is added or subtracted from
the waveform to ramp the signal up or down in a linear
fashion. This value is a function of rise time, period, and
amplitude, where rise time and period are related
through the following equation for the crest factor of a
trapezoidal waveform.
3
1
t RISE = --- T 1 – ----------2-
4
CF
where
Preliminary Rev. 0.91
Si3220/Si3225
Ringing DC Offset Voltage
1
T = Period = -------------- CF = desired crest factor
f RING
So for a 90 VPK, 20 Hz trapezoidal waveform with a
crest factor of 1.3, the period is 0.05 s and the rise time
requirement is 0.015 s.
A dc offset voltage can be added to the Si3220’s ac
ringing waveform by programming the RINGOF RAM
location to the appropriate setting. The value of
RINGOF is calculated as follows:
RINGPHAS = 4 x 0.05 x 8000 = 1600 (0x0640)
V OFF
15
RINGOF = --------------- × 2
64.32
RINGAMP = 90/160.8 x (215) = 18340 (0x47A5)
RINGFREQ = (2 x RINGAMP)/(0.0153 x 8000) = 300
(0x012C)
The time registers and interrupts described in the
sinusoidal ring description also apply to the trapezoidal
ring waveform.
Ringing Coefficients
The ringing coefficients are calculated in decimal for
sinusoidal and trapezoidal waveforms. The RINGPHAS
and RINGAMP hex values are decimal to hex
conversions in 16-bit, 2’s complement representations
for their respective RAM locations.
To obtain sinusoidal RINGFREQ RAM values, the
RINGFREQ decimal number is converted to a 24-bit 2’s
complement value. The lower 12 bits are placed in
RINGFRLO bits 14:3. RINGFRLO bits 15 and 2:0 are
cleared to 0. The upper 12 bits are set in a similar
manner in RINGFRHI, bits 13:3. RINGFRHI bit 14 is the
sign bit and RINGFRHI bits 2:0 are cleared to 0.
For
example,
the
register
RINGFREQ=0x7EFD9D are as follows:
values
for
RINGFRHI = 0x3F78
RINGFRLO = 0x6CE8
To obtain trapezoidal RINGFREQ RAM values, the
RINGFREQ decimal number is converted to an 8-bit, 2’s
complement value. This value is loaded into RINGFRHI.
RINGFRLO is not used.
External Unbalanced Ringing
The Si3225 supports centralized, battery-backed
unbalanced ringing schemes by providing a ringing
relay driver as well as inputs from an external ring trip
circuit. Using this scheme, line-card designers can use
the Dual ProSLIC chipset in existing system
architectures with minimal system changes.
Linefeed Overhead Voltage Considerations
During Ringing
The ringing mode output impedance allows ringing
operation without overhead voltage modification
(VOVR = 0). If an offset of the ringing signal from the
ring lead is desired, VOVR can be used for this
purpose.
Ringing Power Considerations
The total power consumption of the Si3220/Si3200
chipset using internal ringing generation is dependent
on the VDD supply voltage, the desired ringing
amplitude, the total loop impedance, and the AC load
impedance (number of REN). The following equations
can be used to approximate the total current required
for each channel during ringing mode.
VDD = 3.3 V:
I DD,AVE = 22mA + ( 6mA × REN )
VDD = 5 V:
VTIP-RING
I DD,AVE = 26mA + ( 6mA × REN )
And
v RING,RMS
2.04
I BAT,RMS = ---------------------------------------------------------------- × ----------R LOAD + R LOOP + R OUT
π
VOFF
Where:
T = 1/freq
REN = number of REN
RLOAD = 7000/REN for North America
tRISE
time
RLOOP = loop impedance
ROUT = ProSLIC output impedance = 320 Ω
Figure 26. Trapezoidal Ringing Waveform
Preliminary Rev. 0.91
43
Si3220/Si3225
Ring Trip Detection
A ring trip event signals that the terminal equipment has
transitioned to an off-hook state after ringing has
commenced, ensuring that the ringing signal is removed
before normal speech begins. The Dual ProSLIC is
designed to implement either an ac- or dc-based
internal ring trip detection scheme or a combination of
both schemes. The system design is flexible to address
varying loop lengths of different applications. An ac ring
trip detection scheme cannot reliably detect an off-hook
condition when sourcing longer loop lengths, as the
20 Hz ac impedance of an off-hook long loop is
indistinguishable from a heavily loaded (5 REN) short
loop in the on-hook state. Therefore, a dc ring trip
detection scheme is required when sourcing longer loop
lengths.
The Si3220 can implement either an ac- or dc-based
ring trip detection scheme, depending on the application.
The Si3225 allows external dc ring trip detection when
using a battery-backed external ringing generator by
monitoring the ringing feed path through two sensing
inputs on each channel. By monitoring this path, the
Dual ProSLIC detects a dc current flowing in the loop
once the end equipment has gone off-hook. Table 26
provides recommended register and RAM settings for
various applications, and Table 27 lists the register and
RAM addresses that must be written or monitored to
correctly detect a ring trip condition.
Figure 27 illustrates the internal functional blocks that
correctly detect and process a ring trip event. The
primary input to the system is the loop current sense
(ILOOP) value provided by the loop monitoring circuitry
and reported in the ILOOP RAM location register. The
ILOOP RAM location value is processed by the ISP
block when the LFS bits in the Linefeed register indicate
the device is in the ringing state. The output of the ISP
then feeds into a pair of programmable digital low-pass
filters; one for the ac ring trip detection path and one for
44
the dc path. The ac path also includes a full wave
rectifier block prior to the LPF block. The outputs of
each low pass filter block are then passed on to a
programmable ring trip threshold (RTACTH for ac
detection and RTDCTH for dc detection). Each
threshold block output is then fed to a programmable
debounce filter to ensure a valid ring trip event. The
output of each debounce filter remains constant unless
the input remains in the opposite state for the entire
period of time set using the ac and dc ring trip debounce
interval registers, RTACDB and RTDCDB, respectively.
The outputs of both debounce filter blocks are then
ORed together. If either the ac or the dc ring trip circuits
indicate a valid ring trip event has occurred, the RTP bit
is set. Either the ac or dc ring trip detection circuits are
disabled by setting the respective ring trip threshold
sufficiently high so that it does not trip under any
condition. A ring trip interrupt also generates if the
RTRIPE bit is enabled.
Ringtrip Timeout Counter
The Dual ProSLIC incorporates a ringtrip timeout
counter (RTCOUNT) that will monitor the status of the
ringing control. When exiting ringing, the Dual ProSLIC
will allow the ringtrip timeout counter amount of time
(RTCOUNT x 1.25 ms/LSB) for the mode to switch to
On-hook Transmission or Active. The mode that is
being exited to is governed by whether the command to
exit ringing is a ringing active timer expiration (on-hook
transmission) or ringtrip/manual mode change (Active
mode). The ringtrip timeout counter will assure ringing is
exited within its time setting (RTCOUNT x 1.25 ms/LSB,
typically 200 ms).
Ringtrip Debounce Interval
The ac and dc ring trip debounce intervals can be
calculated based on the following equations:
RTACDB = tdebounce (1600/RTPER)
RTDCDB = tdebounce (1600/RTPER)
Preliminary Rev. 0.91
Si3220/Si3225
RTACTH
AC Ring Trip
Threshold
LFS
_
ILOOP
Input
Signal
Processor
Full Wave
Rectifier
Digital
LPF
+
Debounce
Filter_AC
RTP
Interrupt
Logic
RTACDB
RTPER
Digital
LPF
+
_
DC Ring Trip
Threshold
RTRIPS
RTRIPE
Debounce
Filter_DC
RTDCDB
RTDCTH
Figure 27. Ring Trip Detect Processing Circuitry
Preliminary Rev. 0.91
45
Si3220/Si3225
Table 26. Recommended Values for Ring Trip Registers and RAM Addresses1
Ringing
Method
Internal
(Si3220)
External
(Si3225)
Ringing
Frequency
16–32 Hz
33–60 Hz
16–32 Hz
33–60 Hz
DC
Offset
Added?
Yes
No
Yes
No
Yes
Yes
RTPER
RTACTH
RTDCTH
RTACDB/
RTDCDB
800/fRING
800/fRING
2(800/fRING)
2(800/fRING)
800/fRING
2(800/fRING)
221 x RTPER
1.59 x VRING,PK x RTPER
221 x RTPER
1.59 x VRING,PK x RTPER
32767
32767
0.577(RTPER x VOFF)
32767
0.577(RTPER x VOFF)
32767
0.067 x RTPER x VOFF
0.067 x RTPER x VOFF
See Note 2
Notes:
1. All calculated values should be rounded to the nearest integer.
2. Refer to Ring Trip Debounce Interval for RTACDB and RTDCDB equations.
Table 27. Register and RAM Locations Used for Ring Trip Detection
Parameter
Register/RAM
Mnemonic
Register/RAM
Bits
Programmable
Range
Resolution
Ring Trip Interrupt Pending
IRQVEC2
RTRIPS
Yes/No
N/A
Ring Trip Interrupt Enable
IRQEN2
RTRIPE
Enabled/Disabled
N/A
AC Ring Trip Threshold
RTACTH
RTACTH[15:0]
See Table 26
DC Ring Trip Threshold
RTDCTA
RTDCTH[15:0]
See Table 26
Ring Trip Sample Period
RTPER
RTPER[15:0]
See Table 26
LINEFEED
LFS[2:0]
N/A
N/A
Ring Trip Detect Status
(monitor only)
LCRRTP
RTP
N/A
N/A
AC Ring Trip Detect Debounce
Interval
RTACDB
RTACDB[15:0]
0 to 40.96 s
1.25 ms
DC Ring Trip Detect Debounce
Interval
RTDCDB
RTDCDB[15:0]
0 to 40.96 s
1.25 ms
ILOOP
ILOOP[15:0]
0 to 101.09 mA
See
Table 19
Linefeed Shadow (monitor only)
Loop Current Sense
(monitor only)
Loop Closure Mask
The Dual ProSLIC implements a loop closure mask to
ensure mode change between Ringing and Active or
On-hook Transmission without causing an erroneous
loop closure detection. The loop closure mask register,
LCRMASK, should be set such that loop closure
detection is ignored for LCRMASK 1.25 ms/LSB
amount of time. The programmed time is set to mask
detection of transitional currents that occur when exiting
the ringing mode while driving a reactive load (i.e., 5
REN). A typical setting is 80 ms (LCRMASK = 0x40).
Si3220 Ring Trip Detection
The Si3220 provides the ability to process a ring trip
event using an ac-based detection scheme. Using this
scheme eliminates the need to add dc offset to the
46
ringing signal, which reduces the total power dissipation
during the ringing state and maximizes the available
ringing amplitude. This scheme is valid for shorter loop
lengths only since it cannot reliably detect a ring trip
event if the off-hook line impedance overlaps the onhook impedance at 20 Hz.
The Si3220 also can add a dc offset component to the
ringing signal and detect a ring trip event by monitoring
the dc loop current flowing once the terminal equipment
transitions to the off-hook state. Although adding dc
offset reduces the maximum available ringing amplitude
(using the same ringing supply), this method is required
to reliably detect a valid ring trip event when sourcing
longer loop lengths. The dc offset can be programmed
from 0 to 64.32 V in the RINGOF RAM address as
Preliminary Rev. 0.91
Si3220/Si3225
required to produce adequate dc loop current in the offhook state. Depending on the loop length and the ring
trip method, the ac or dc ring trip detection circuits are
disabled by setting their respective ring trip thresholds
(RTACTH or RTDCTH) sufficiently high so it does not
trip under any condition.
VDD
VCC
Si3220/
Si3225
3 V/5 V Relay
(polarized or
non-polarized)
Si3225 Ring Trip Detection
The Si3225 implements an external ring trip detection
scheme when using a standard battery-backed, external
ringing generator. In this application, the centralized
ringing generator produces an unbalanced ringing
signal that is distributed to individual TIP/RING pairs. A
per-channel ringing relay is required to disconnect the
Si3225 from the TIP/RING pair and apply the ringing
signal. By monitoring the ringing feed path across a ring
feed sense resistor (RRING in Figure 31) in series with
the ringing source, the Si3225 can detect the dc current
path created when the hook switch inside the terminal
equipment closes. The internal ring trip detection
circuitry is identical to that illustrated in Figure 27.
Figure 31 illustrates the typical external ring trip circuitry
required for the Si3225. Because of the long loop nature
of these applications, a dc ring trip detection scheme is
used typically. The user can disable the ac ring trip
detection circuitry by setting the RTACTH threshold
sufficiently high so it does not trip under any condition.
Relay Driver Considerations
The Dual ProSLIC devices include up to three
dedicated relay drivers to drive external ringing and/or
test relays. Test relay drivers TRD1a, TRD1b, TRD2a,
and TRD2b are provided in all product versions, and
ringing relay drivers RRDa and RRDb are included for
the Si3225 only. In most applications, the relay can be
driven directly from the Dual ProSLIC with no external
relay drive circuitry required. Figure 28 illustrates the
internal relay driver circuitry using a 3 V or 5 V relay.
Relay
Driver
Logic
RRDa/b
TRD1a/b
TRD2a/b
GDD
Figure 28. Dual ProSLIC Internal Relay Drive
Circuitry
The internal driver logic and drive circuitry is powered
by the same VDD supply as the chip’s main VDD supply
(VDD1–VDD4 pins). When operating external relays
from a VCC supply equal to the chip’s VDD supply, an
internal diode network provides protection against
overvoltage conditions from flyback spikes when the
relay is opened. Both 3 V or 5 V relays can be used in
the configuration shown in Figure 28 and either
polarized or non-polarized relays are acceptable if the
VCC and VDD supplies are identical. The input
impedance (RIN) of the relay driver pins is a constant
11 Ω while sinking less than the maximum rated 85 mA
into the pin.
If the operating voltage of the relay (VCC) is higher than
the Dual ProSLIC’s VDD supply voltage, an external
drive circuit is required to eliminate leakage from VCC to
VDD through the internal protection diode. In this
configuration, a polarized relay will provide optimal
overvoltage
protection
and
minimal
external
components. Figure 29 illustrates the required external
drive circuit and Table 28 provides recommended
values for RDRV for typical relay characteristics and VCC
supplies. The output impedance (ROUT) of the relay
driver pins is a constant 63 Ω while sourcing less than
the maximum rated 28 mA out of the pin.
Preliminary Rev. 0.91
47
Si3220/Si3225
VCC
VDD
Si3220/
Si3225
Polarized
relay
RRDa/b
TRD1a/b
TRD2a/b
IDRV
Q1
RDRV
Figure 29. Driving Relays with VCC > VDD
The maximum allowable RDRV value can be calculated with the following equation:
( V DD,MIN – 0.6 V ) ( R RELAY ) ( β Q1,MIN )
MaxR DRV = ------------------------------------------------------------------------------------------------- – R SOURCE
V CC,MAX – 0.3 V
Where βQ1,MIN ~ 30 for a 2N2222.
Table 28. Recommended RDRV Values
48
ProSLIC VDD
Relay VCC
Relay RCOIL
Maximum RDRV
Recommended 5% Value
3.3 V ±5%
3.3 V ±5%
64 Ω
Not Required
—
5 V ±5%
5 V ±5%
178 Ω
Not Required
—
3.3 V ±5%
5 V ±5%
178 Ω
2718 Ω
2.7 kΩ
3.3 V ±5%
12 V ±10%
1028 Ω
6037 Ω
5.6 kΩ
3.3 V ±5%
24 V ±10%
2880 Ω
8364 Ω
8.2 kΩ
3.3 V ±5%
48 V ±10%
7680 Ω
11092 Ω
11 kΩ
5 V ±5%
12 V ±10%
1028 Ω
9910 Ω
9.1 kΩ
5 V ±5%
24 V ±10%
2880 Ω
13727 Ω
13 kΩ
5 V ±5%
48 V ±10%
7680 Ω
18202 Ω
18 kΩ
Preliminary Rev. 0.91
IRINGXSCAL
D
ZERDELAY
COUNTER0
0
1
2
3
4
COUNTER1
5
6
7
8
9
10
11
0
1
2
3
4
5
6
7
8
9
10
11
0
1
2
3
4
5
6
7
8
9
10
11
0
1
2
3
4
5
6
0
1
2
3
4
5
6
7
8
9
10
11
0
1
2
3
4
5
6
7
8
9
10
11
0
1
2
3
4
5
6
7
8
9
10
11
0
Preliminary Rev. 0.91
RINGEN
RRD
On
Off
On
LF
Active
Ringing
LFSDELAY
LFS
Off
Ringing
LFSDELAY
OHT
Ringing
49
Si3220/Si3225
Figure 30. Timing Characteristics for Ringing Relay Control
Active
Si3220/Si3225
VOFF VRING
+
510 Ω
_
806 kΩ
RTRP
BLkRING
806 kΩ
Relay
Phone
RING
Si3200
Si3225
Hook
Switch
Protection
RRD
TIP
VDD
Figure 31. Si3225 External Ring Trip Circuitry
Ringing Relay Activation During Zero Crossings
The Si3225 is for applications that use a centralized
ringing generator and a per-channel ringing relay to
connect the ringing signal to the TIP/RING pair. The
Si3225 has one relay driver output per channel (RRDa
and RRDb) that can drive a mechanical or solid-state
DPDT relay. To reduce impulse noise that can couple
into adjacent lines, the relay should be closed when
there is zero voltage across the relay contacts and
opened during periods when there is zero current
through the contacts.
Closing the Relay at Zero Voltage
Internal voltage monitoring circuitry closes the relay at
zero voltage with respect to the line voltage. By
observing the phase of the ringing signal and constantly
monitoring the open-circuit T-R voltage, VOC, the
Si3225 can detect the next time when there is zero
voltage across the relay contacts.
Opening the Relay at Zero Current
Opening the ringing relay at zero current also is
accomplished using the internal monitoring circuitry and
prevents arcing from excess current flow when the relay
contacts are opened. The current flowing through the
ringing relay is continuously monitored in the IRNGNG
RAM address, and two internal counters (COUNTER0
and COUNTER1) detect time elapsed since the last two
zero current crossings based on the ringing period and
predict when the next zero crossing occurs. The ringing
relay current and internal counters are both updated at
an 8 kHz rate. To account for the mechanical delay of
the relay, a programmable advance firing timer allows
the user to initiate relay opening up to 10 ms prior to the
zero current crossing event. Figure 30 illustrates the
50
timing sequence for a typical ringing relay control
application.
During a typical ringing sequence, the Si3225 monitors
both the ringing relay current (IRNGNG) and the
RINGEN bit of the RINGCON register. The RINGEN bit
toggles because of pre-programmed ringing cadence or
a change in operating state. COUNTER0 and
COUNTER1 are re-started at each alternating zero
current crossing event, and the delay period
ZERDELAY equal to the ringing frequency period less
the desired advance firing time (D) is entered by the
user. If either counter reaches the same value as
ZERDELAY, the relay control signal is enabled when the
RINGEN bit transition has already occurred. During
typical ringing bursts, the LFS bits of the Linefeed
register toggle between the RINGING and OHT states
based on the pre-programmed ringing cadence. The
transition from OHT to RINGING is synchronized with
the RRD state transitions so the ringing burst starts
immediately. The transition from RINGING to OHT is
gated by a user-programmed delay period LFSDELAY
that ensures the ringing burst has ceased before going
to the OHT state or to the ACTIVE state in response to a
Linefeed state change.
Preliminary Rev. 0.91
Si3220/Si3225
Polarity Reversal
The Dual ProSLIC devices support polarity reversal for
message waiting functionality and various signaling
modes. The ramp rate can be programmed for a smooth
transition or an abrupt transition to accommodate
different application requirements. A wink function is
provided for special equipment that responds to a
smooth ramp to VOC = 0 V. Table 29 illustrates the
register bits required to program the polarity reversal
modes.
Setting the Linefeed register to the opposite polarity
immediately reverses (hard reversal) the line polarity.
For example, to transition from Forward Active mode to
Reverse Active mode changes LF[2:0] from 001 to 101.
Polarity reversal is accommodated in the OHT and
ground start modes. The POLREV bit is a read-only bit
that reflects if the device is in Polarity Reversal mode.
For smooth polarity reversal, set the PREN bit to 1 and
the RAMP bit to 0 or 1 depending on the desired ramp
rate (see Table 29). Polarity reversal is then
accomplished by toggling the linefeed register from
forward to reverse modes as desired.
A wink function slowly ramps down the TIP-RING
voltage (VOC) to 1 followed by a return to the original
VOC value (set in the VOC RAM location). This scheme
lights a message-waiting lamp in certain handsets. No
change to the linefeed register is necessary to enable
this function. Instead, the user sets the VOCZERO bit to
1 so that the TIP-RING voltage collapses to 0 V at the
rate programmed by the RAMP bit. Setting the
VOCZERO bit back to 0 returns the TIP-RING voltage
to its normal setting. With a software timer, the user can
automate the cadence of the wink function. Figure 32
illustrates the wink function.
Table 29. Register and RAM Locations used for Polarity Reversal
Parameter
Programmable Range
Linefeed
See table 15
Polarity Reversal Status
Read only
Wink Function
(Smooth transition to Voc=0V)
1 = Ramp to 0 V
0 = Return to previous VOC
Smooth Polarity Reversal Enable
Smooth Polarity Reversal Ramp
Rate
Register/RAM Register/RAM
Bits
Mnemonic
LF[2:0]
LINEFEED
POLREV
POLREV
VOCZERO
POLREV
0 = Disabled
1 = Enabled
PREN
POLREV
0 = 1 V/1.25 ms
1 = 2 V/1.25 ms
RAMP
POLREV
Preliminary Rev. 0.91
51
Si3220/Si3225
VTIP/RING (V)
VOC = 48 V
50
40
2 V/1.25 ms slope
set by RAMP bit
30
20
10
Time (ms)
0
0
10
20
30
40
Set VOCZERO bit to 1
50
60
70
80
Set VOCZERO bit to 0
Figure 32. Wink Function with Programmable Ramp Rate
Two-Wire Impedance Synthesis
Two-wire impedance synthesis is performed on-chip to
optimally match the output impedance of the Dual
ProSLIC to the impedance of the subscriber loop to
minimize the receive path signal reflected back onto the
transmit path. The Dual ProSLIC chipset provides onchip digitally programmable, two-wire impedance
synthesis to meet return loss requirements against
virtually any global two-wire impedance requirement.
Real and complex two-wire impedances are realized by
a programmable digital filter block. (See Z block in
Figure 11 on page 22.)
RP
RS
CP
Figure 33. Two-Wire Impedance Synthesis
Configuration
Table 30. Two-Wire Impedance
Synthesis Limitations
Desired
Configuration
Programmable Limits
RS only
100–1000 Ω
RS + CP
RS x CP > 0.5 ms
RS + RP||CP
RS/(RS + RP) > 0.1
The two-wire impedance is programmed by loading the
desired real or complex impedance value into the
Si322X Coefficient Generator software in the format RS
+ RP||CP, as shown in Figure 33. The software
calculates the appropriate hex coefficients and loads
them
into
the
appropriate
control
registers
(registers 33–52). The two-wire impedance can be set
to any real or complex value within the boundaries set in
Table 30. The actual impedance presented to the
subscriber loop varies with series impedance from
protection devices placed between the Dual ProSLIC
chipset outputs and the TIP/RING pair according to the
following equation:
Z T = 2R PROT + ( R S + R P || C P )
Where:
52
Preliminary Rev. 0.91
ZT is the termination impedance presented
to the TIP/RING pair
Si3220/Si3225
RPROT is the series resistance caused by
protection devices
RS is the series portion of the synthesized
impedance
RP||CP is the parallel portion of the
synthesized impedance
The user must enter the value of RPROT into the
software so the equalizer block can compensate for
additional series impedance. (See Figure 11 on page
22.) Figure 34 illustrates the simplified two-wire
impedance circuit including external protection
resistors, where ZL is the actual line impedance for the
specific geographical region. The Dual ProSLIC devices
can accomodate up to 50 Ω of series protection
impedance per leg. The Dual ProSLIC devices load a
600 Ω default setting into the RS register if the user
does not define the impedance setting, which assumes
there is no additional series protection resistance.
The ac impedance generation scheme is comprised of
analog and DSP-based coefficients. To turn off the
analog coefficients (RS, ZP, and ZZ bits in the ZRS and
ZZ registers), the user can simply set the ZSDIS bit of
the ZZ register to 0. To turn off the DSP coefficients
(ZA1H1 through ZB3LO registers), each register must
be loaded with 0x00.
TIP
ZT
RING
Si3200
ZL
RPROT
Dual
ProSLIC
Transhybrid Balance Filter
The Dual ProSLIC devices provide a transhybrid
balance function via a digitally programmable balance
filter block. (See “H” block in Figure 11 on page 22.) The
Dual ProSLIC devices implement a 8-tap FIR filter and a
second order IIR filter, both running at a 16 kHz sample
rate. These two filters combine to form a digital replica
of the reflected signal (echo) from the transmit path
inputs. The user can filter settings on a per-line basis by
loading the desired impedance cancellation coefficients
into the appropriate registers. The Si322X Coefficient
Generator software interface is provided for calculating
the appropriate coefficients for the FIR and IIR filter
blocks.
The transhybrid balance filters can be disabled to
implement loopback diagnostic modes. To disable the
transhybrid balance filter (zero cancellation), set the
HYBDIS bit in the DIGCON register to 1. With the hybrid
balance cancellation scheme disabled, the user can
accurately measure the full transmit path signal to
measure the two-wire return loss.
Note: The user must enter values into each register location
to ensure correct operation when the hybrid balance
block is enabled.
Tone Generators
Dual ProSLIC devices have two digital tone generators
that allow a wide variety of single or dual tone frequency
and amplitude combinations that spare the user the
effort of generating the required POTS signaling tones
on the PCM highway. DTMF, FSK (caller ID), call
progress, and other tones can all be generated on-chip.
The tones are sent to the receive or transmit paths.
(See Figure 11 on page 22.)
Tone Generator Architecture
RPROT
Figure 34. Two-Wire Impedance Simplified
Circuit
A simplified diagram of the tone generator architecture
is shown in Figure 35. The oscillator, active/inactive
timers, interrupt block, and signal routing block are
connected for flexibility in creating audio signals.
Control and status register bits are placed in the figure
to indicate their association with the tone generator
architecture. The register set for tone generation is
summarized in Table 31 on page 55.
Preliminary Rev. 0.91
53
Si3220/Si3225
8 kHz
Clock
8 kHz
Clock
ZEROENn
Zero Cross
ENSYNCn
OSCnEN
16-Bit
Modulo
Counter
OSCnTA
Expire
Zero
Cross
Logic
Two-Pole
Resonant
Register Oscillator
Load
Logic
OSCnTI
Expire
to TX Path
Enable
Signal
Routing
Load
to RX Path
OSCnTA
OSCnFREQ
REL*
INT
Logic
OSCnTAEN
OSnTIS
OSCnTI
ROUTn
OSCnAMP
OSnTIE
OSnTAS
INT
Logic
OSCnTIEN
OSCnPHAS
OSnTAE
*Tone Generator 1 Only
n = "1" or "2" for Tone Generator 1 and 2, respectively
Figure 35. Tone Generator Diagram
Oscillator Frequency and Amplitude
Each of the two tone generators contains a two-pole
resonant oscillator circuit with a programmable
frequency and amplitude, which are programmed via
RAM addresses OSC1FREQ, OSC1AMP, OSC1PHAS,
OSC2FREQ, OSC2AMP, and OSC2PHAS. The sample
rate for the two oscillators is 8000 Hz. The equations
are as follows:
1
15
OSC1AMP = --- 0.21556
--------------------- × ( 2 – 1 ) × 0.5 = 1424
4 1.78434
= 0x590
OSC1PHAS = 0
coeff2 = cos (2π 1336 / 8000) = 0.49819
OSC2FREQ = 0.49819 (214) = 8162 = 0x1FE2
coeffn = cos(2π fn/8000 Hz),
where fn is the frequency to be generated;
1
15
OSC2AMP = --- 0.50181
--------------------- × ( 2 – 1 ) × 0.5 = 2370
4 1.49819
14
OSCnFREQ = coeffn x (2 );
= 0x942
1 1 – coeff
DesiredVrms
15
OSCnAMP = --- ----------------------- × ( 2 – 1 ) × -------------------------------------4 1 + coeff
1.11Vrms
where desired Vrms is the amplitude to be generated;
OSCnPHAS = 0,
n = 1 or 2 for oscillator 1 or oscillator 2, respectively.
For example, to generate a DTMF digit of 8, the two
required tones are 852 Hz and 1336 Hz. Assuming we
want to generate half-scale values (ignoring twist), the
following values are calculated:
2π852
coeff 1 = cos ----------------- = 0.78434
8000
14
OSC1FREQ = 0.78434 ( 2 ) = 12851 = 0x3233
54
OSC2PHAS = 0
The computed values above are written to the
corresponding registers to initialize the oscillators. Once
the oscillators are initialized, the oscillator control
registers can be accessed to enable the oscillators and
direct their outputs.
Tone Generator Cadence Programming
Each of the two tone generators contains two timers,
one for setting the active period and one for setting the
inactive period. The oscillator signal is generated during
the active period and suspended during the inactive
period. Both the active and inactive periods can be
programmed from 0 to 8 seconds in 125 µs steps. The
active period time interval is set using OSC1TA for tone
generator 1 and OSC2TA for tone generator 2.
Preliminary Rev. 0.91
Si3220/Si3225
To enable automatic cadence for tone generator 1,
define the OSC1TA and OSC1TI registers and then set
the OSC1TAEN and OSC1TIEN bits. This enables each
of the timers to control the state of the Oscillator Enable
bit, OSC1EN. The 16-bit counter counts until the active
timer expires, when the 16-bit counter resets to zero
and begins counting until the inactive timer expires. The
cadence continues until the user clears the OSC1TA
and OSC1TIEN control bits. Setting the ZEROEN1 bit
implements the zero crossing detect feature. This
ensures that each oscillator pulse ends without a dc
component. The timing diagram in Figure 36 is an
example of an output cadence that uses the zero
crossing feature.
One-shot oscillation is possible with OSC1EN and
OSC1TAEN. Direct control over the cadence is
achieved by setting the OSC1EN bit directly if
OSC1TAEN and OSC1TIEN are disabled.
The operation of tone generator 2 is identical to that of
tone generator 1 using its respective control registers.
Note: Tone Generator 2 should not be enabled simultaneously with the ringing oscillator because of resource
sharing within the hardware.
Table 31. Register and RAM Locations Used for Tone Generation
Tone Generator 1
Parameter
Register/RAM
Mnemonics
Register/RAM Bits
Description/Range
(LSB Size)
Oscillator 1 Frequency
Coefficient
OSC1FREQ
OSC1FREQ[15:3]
Sets oscillator frequency
Oscillator 1 Amplitude Coefficient
OSC1AMP
OSC1AMP[15:0]
Sets oscillator amplitude
Oscillator 1 Initial Phase
Coefficient
OSC1PHAS
OSC1PHAS[15:0]
Sets initial phase
(default = 0)
Oscillator 1 Active Timer
O1TALO/O1TAHI
OSC1TA[15:0]
0 to 8.19 s (125 µs)
Oscillator 1 Inactive Timer
O1TILO/O1TIHI
OSC1TI[15:0]
0 to 8.19 s (125 µs)
Oscillator 1 Control
OMODE, OCON
FSKSSEN, OSC1FSK,
ZEROEN1, ROUT1,
ENSYNC1, OSC1TAEN,
OSC1TIEN, OSC1EN
Enables all Oscillator 1 parameters
Oscillator 1 Interrupts
IRQVEC1, IRQEN1
OS1TAS, OS1TIS, OS1TAE,
OS1TIE
Interrupt enable/status
Tone Generator 2
Parameter
Location
Register/RAM Address
Description/Range
Oscillator 2 Frequency
Coefficient
OSC2FREQ
OSC2FREQ[15:3]
Sets oscillator frequency
Oscillator 2 Amplitude Coefficient
OSC2AMP
OSC2AMP[15:0]
Sets oscillator amplitude
Oscillator 2 Initial Phase
Coefficient
OSC2PHAS
OSC2PHAS[15:0]
Sets initial phase
(default = 0)
Oscillator 2 Active Timer
O2TALO/O2TAHI
OSC2TA[15:0]
0 to 8.19 s (125 µs)
Oscillator 2 Inactive Timer
O2TILO/O2TIHI
OSC2TI[15:0]
0 to 8.19 s (125 µs)
Oscillator 2 Control
OMODE, OCON
ZEROEN2, ROUT2,
ENSYNC2, OSC2TAEN,
OSC2TIEN, OSC2EN
Enables all Oscillator 2 parameters
Oscillator 2 Interrupts
IRQVEC1, IRQEN1
OS2TAS, OS2TIS, OS2TAE,
OS2TIE
Interrupt enable/status
Preliminary Rev. 0.91
55
Si3220/Si3225
OSC1EN
0,1 ...
..., OSC1TA 0,1 ...
...
..., OSC1TI 0,1 ... ..., OSC1TA 0,1 ...
...
ENSYNC1
Tone
Gen. 1
Signal
Output
Figure 36. Tone Generator Timing Diagram
First
Ring Burst
Message
Type
Channel
Seizure
Message
Length
Parameter 1
Message Header
Parameter
Type
Mark
Data
Packet
Parameter 2
Second
Ring Burst
Parameter n
Message Body
Data
Length
Data
Content
Figure 37. On-Hook Caller ID Transmission Sequence
56
Preliminary Rev. 0.91
Checksum
Si3220/Si3225
Tone Generator Interrupts
Both the active and inactive timers can generate an
interrupt to signal “on/off” transitions to the software.
The timer interrupts for tone generator 1 can be
individually enabled by setting the OS1TAE and OS1TIE
bits. Timer interrupts for tone generator 2 are OS2TAE
and OS2TIE. A pending interrupt for each of the timers
is determined by reading the OS1TAS, OS1TIS,
OS2TAS, and OS2TIS bits in the IRQVEC1 register.
Caller ID Generation
The Dual ProSLIC devices generate caller ID signals in
compliance with various Bellcore and ITU specifications
as described in Table 32 by providing continuous phase
binary frequency shift key (FSK) modulation.
Oscillator 1 is required because it preserves phase
continuity during frequency shifts whereas Oscillator 2
does not. Figure 37 illustrates a typical caller ID
transmission sequence in accordance with Bellcore
requirements.
Table 32. FSK Modulation Requirements
Parameter
ITU-T V.23
Bellcore GR30-CORE
Mark Frequency (logic 1)
1300 Hz
1200 Hz
Space Frequency (logic 0)
2100 Hz
2200 Hz
Transmission Rate
1200 baud
Table 33. Register and RAM Locations used for Caller ID Generation
Parameter
Register/RAM
Mnemonic
Register/RAM Bits
Description/Range
OMODE
O1FSK8
Enable/disable
O1TALO/O1TAHI
OSC1TA[15:0]
0 to 8.19 s/125 µs
FSKDAT
FSKDAT[7:0]
Caller ID data
FSK Frequency for Space
FSKFREQ0
FSKFREQ0[15:3]
Audio range
FSK Frequency for Mark
FSKFREQ1
FSKFREQ1[15:3]
Audio range
FSK Amplitude for Space
FSKAMP0
FSKAMP0[15:3]
FSK Amplitude for Mark
FSKAMP1
FSKAMP1[15:3]
FSK 0-1 Transition Freq, High
FSK01HI
FSK01HI[15:3]
FSK 0-1 Transition Freq, Low
FSK01LO
FSK01LO[15:3]
FSK 1-0 Transition Freq, High
FSK10HI
FSK10HI[15:3]
FSK 1-0 Transition Freq, Low
FSK10LO
FSK10LO[15:3]
FSK Start & Stop Bit Enable
Oscillator 1 Active Timer
FSK Data Byte
The register and RAM locations for caller ID generation
are listed in Table 33. Caller ID data is entered into the
8-bit FSKDAT register. The data byte is double buffered
so that the Dual ProSLIC can generate an interrupt
indicating the next data byte can be written when
processing begins on the current data byte. The caller
ID data can be transmitted in one of two modes
controlled by the O1FSK8 register bit. When
O1FSK8 = 0 (default case), the 8-bit caller ID data is
transmitted with a start bit and stop bit to create a 10-bit
data sequence. If O1FSK8 = 1, the caller ID data is
transmitted as a raw 8-bit sequence with no start or stop
bits. The value programmed into the OSC1TA register
determines the bit rate, and the interrupt rate is equal to
the bit rate divided by the data sequence length (8 or 10
bits).
Pulse Metering Generation
The Si3220 offers an additional tone generator to
generate tones above the audio frequency band. This
oscillator generates billing tones which are typically
12 kHz or 16 kHz. The generator follows the same
algorithm as described in "Tone Generator Architecture"
on page 53 with the exception that the sample rate for
computation is 64 kHz instead of 8 kHz. The equation is
as follows:
Preliminary Rev. 0.91
Coeff = cos (2πf / 64000 Hz)
PMFREQ = coeff x (214 – 1)
57
Si3220/Si3225
DesiredV PK
15
1 1 – coeff
PMAMPL = --- ----------------------- × ( 2 – 1 ) × -------------------------------------FullScaleV PK
4 1 + coeff
where Full Scale VPK = 0.5 V.
The pulse metering oscillator has a volume envelope
(linear ramp) on the on/off transitions of the oscillator.
The ramp is controlled by the value in the PMRAMP
RAM address, and the sinusoidal generator output is
multiplied by this volume before it is sent to the Pulse
Metering DAC. The volume value is incremented by the
value in PMRAMP at an 8 kHz rate. The volume will
ramp from 0 to 7FFF in increments of PMRAMP to allow
the value of PMRAMP to set the slope of the ramp. The
clip detector stops the ramp once the signal seen at the
transmit path exceeds the amplitude threshold set by
PMAMPTH, which provides an automatic gain control
(AGC) function to prevent the audio signal from clipping.
When the pulse metering signal is turned off, the
volume ramps down to 0 by decrementing according to
the value of PMRAMP. Figure 38 illustrates the
functional blocks involved in pulse metering generation,
and Table 34 presents the register and RAM locations
required that must be set to generate pulse metering
signals.
Table 34. Register and RAM Locations Used for Pulse Metering Generation
Parameter
Register/RAM
Mnemonic
Register/RAM
Bits
Description/Range
(LSB Size)
Pulse Metering Frequency
Coefficient
PMFREQ
PMFREQ[15:3]
Sets oscillator frequency
Pulse Metering Amplitude
Coefficient
PMAMPL
PMAMPL[15:0]
Sets oscillator amplitude
Pulse Metering Attack/Decay
Ramp Rate
PMRAMP
PMRAMP[15:0]
0 to PMAMPL
(full amplitude)
Pulse Metering Active Timer
PMTALO/PMTAHI
PULSETA[15:0]
0 to 8.19 s (125 µs)
Pulse Metering Inactive Timer
PMTILO/PMTIHI
PULSETI[15:0]
0 to 8.19 s (125 µs)
Pulse Metering, Control
Interrupt
IRQVEC1, IRQEN1
PULSTAE,
PULSTIE,
PULSTAS,
PULSTIS
Interrupt Status and control
registers
Pulse Metering AGC
Amplitude Threshold
PMAMPTH
PMAMPTH[15:0]
0 to 500 mV
PM Waveform Present
PMCON
ENSYNC
Indicates signal present
PM Active Timer Enable
PMCON
TAEN
Enable/disable
PM Inactive Timer Enable
PMCON
TIEN
Enable/disable
Pulse Metering Enable
PMCON
PULSE1
Enable/disable
58
Preliminary Rev. 0.91
Si3220/Si3225
Decimation
Filter
ADC
12/16 kHz
Bandpass
Peak Detector
PMAMPTH
–
IBUF
ZA
++
+
DAC
PMRAMP
Pulse
Metering
DAC
±+
x+
Pulse
Metering
Oscillator
Volume
Clip
Logic
7FFF
or 0
8 kHz
Figure 38. Pulse Metering Generation Block Diagram
DTMF Detection
On-chip DTMF detection, also known as Touch Tone, is
available in the Si3220 and Si3225.
It is an in-band signaling system that replaces the pulsedial signaling standard. In DTMF, two tones generate a
DTMF digit. One tone is chosen from the four possible
row tones and one tone is chosen from the four possible
column tones. The sum of these tones constitute one of
16 possible DTMF digits. The row and column tones
and corresponding digits are shown in Table 35.
DTMF detection is performed using a modified Goertzel
algorithm to compute the DFT for each of the eight
DTMF frequencies and their second harmonics. At the
end of the DFT computation, the squared magnitudes of
the DFT results for the 8 DTMF fundamental tones are
computed. The row results are sorted to determine the
strongest row frequency, and the column frequencies
are sorted as well. At the completion of this process,
checks are made to determine if the strongest row and
column tones constitute a DTMF digit.
overwritten by a new one. There is no buffering of the
digit information.
Table 35. DTMF Row/Column Tones
697 Hz
1
2
3
A
770 Hz
4
5
6
B
852 Hz
7
8
9
C
941 Hz
*
0
#
D
1209 Hz
1336 Hz
1477 Hz
1633 Hz
Table 36 outlines the hex code corresponding to the
detected DTMF digits.
The detection process occurs twice within the 45 ms
minimum tone time. A digit must be detected on two
consecutive tests after a pause to be recognized as a
new digit. If all tests pass, an interrupt is generated and
the DTMF digit value is loaded into the DTMF register
according to the following table. If tones occur at the
maximum rate of 100 ms per digit, the interrupt must be
serviced within 85 ms so that the current digit is not
Preliminary Rev. 0.91
59
Si3220/Si3225
attenuation for signals above 3.4 kHz are part of the
combined decimation filter characteristic of the A/D
converter. One more digital filter is available in the
transmit path, THPF. THPF implements the high-pass
attenuation requirements for signals below 65 Hz. An
equalizer block then equalizes the transmit signal path
to compensate for series protection resistance (RPROT)
outside of the ac-sensing inputs. The linear PCM data
stream output from the equalizer block is amplified by
the transmit-path programmable gain amplifier, TPGA,
which can be programmed from –∞ to 6 dB. The DTMF
decoder receives the linear PCM data stream and
performs the digit extraction if enabled by the user. The
final step in the transmit path signal processing is the
A-law or µ-law compression which can reduce the data
stream word width to 8 bits. Depending on the PCM
Mode Select register selection, every 8-bit compressed
serial data word occupies one time slot on the PCM
highway, or every 16-bit uncompressed serial data
word occupies two time slots on the PCM highway.
Table 36. DTMF Hex Codes
Digit
1
2
3
4
5
6
7
8
9
0
*
#
A
B
C
D
Hex code
0x1
0x2
0x3
0x4
0x5
0x6
0x7
0x8
0x9
0xA
0xB
0xC
0xD
0xE
0xF
0x0
Receive Path
Modem Tone Detection
The Dual ProSLIC devices are capable of detecting a
2100 Hz modem tone as described in ITU-T
Recommendation V.8. The detection scheme can be
implemented in both transmit and receive paths, and is
enabled by programming the appropriate register bit.
The detection scheme should be disabled for power
conservation after the modem tone window has passed.
Once a valid modem tone is detected, a register bit will
be set accordingly and the user can check the results by
reading the register value. A programmable debounce
interval is provided to eliminate false detection and can
be programmed in increments of 67 ms by writing to the
appropriate register.
Audio Path Processing
Unlike traditional SLICs, the Dual ProSLIC devices
integrate the codec function into the same IC. The onchip 16-bit codec offers programmable gain/attenuation
blocks and multiple loopback modes for self testing. The
signal path block diagram is shown in Figure 11 on page
22.
Transmit Path
In the transmit path, the analog signal fed by the
external ac coupling capacitors is passed through an
anti-aliasing filter before being processed by the A/D
converter. An analog mute function is provided directly
prior to the A/D converter input. The output of the A/D
converter is an 8 kHz, 16-bit wide, linear PCM data
stream. The standard requirements for transmit path
60
In the receive path, the optionally compressed 8-bit data
is first expanded to 16-bit words. The PCMF register bit
can bypass the expansion process, so that two 8-bit
words are assembled into one 16-bit word. RPGA is the
receive path programmable gain amplifier which can be
programmed from –∞ dB to 6 dB. An 8 kHz, 16-bit
signal is then provided to a D/A converter. An analog
mute function is provided directly after the D/A
converter. When not muted, the resulting analog signal
is applied at the input of the transconductance amplifier,
Gm, which drives the off-chip current buffer, IBUF.
TPGA/RPGA Gain/Attenuation Blocks
The TPGA and RPGA blocks are essentially linear
multipliers with the structure illustrated in Figure 39.
Both blocks can be independently programmed from –∞
to +6 dB (0 to 2 linear scale). The TXGAIN and RXGAIN
RAM locations are used to program each block. A
setting of 0000h will mute all audio signals; a setting of
4000h will pass the audio signal with no gain or
attenuation (0 dB), and a setting of 7FFFh will provide
the maximum 6 dB of gain to the incoming audio signal.
The DTXMUTE and DRXMUTE bits in the DIGCON
register are also available in order to allow muting of the
transmit and receive paths without requiring
modifications to the TXGAIN or RXGAIN settings.
Preliminary Rev. 0.91
Si3220/Si3225
TPGA or RPGA
PCM
In
X
PCM
Out
The group delay distortion in either path is limited to no
more than the levels indicated in Figure 9. The
reference in Figure 9 is the smallest group delay for a
sine wave in the range of 500 Hz to 2500 Hz at 0 dBm0.
M
where M = {0, 1/16384, 2/16384,...32767/16384}
Figure 39. TPGA and RPGA structure
Audio Characteristics
The dominant source of distortion and noise in both the
transmit and receive paths is the quantization noise
introduced by the µ-law or the A-law compression
process. Figure 5 on page 18 specifies the minimum
Signal-to-Noise-and-Distortion Ratio for either path for a
sine wave input of 200 Hz to 3400 Hz.
Both the µ-law and the A-law speech encoding allow the
audio codec to transfer and process audio signals larger
than 0 dBm0 without clipping. The maximum PCM code
is generated for a µ-law encoded sine wave of
3.17 dBm0 or an A-law encoded sine wave of
3.14 dBm0. The device overload clipping limits are
driven by the PCM encoding process. Figure 6 on page
19 shows the acceptable limits for the analog-to-analog
fundamental power transfer-function, which bounds the
behavior of the device.
The transmit path gain distortion versus frequency is
shown in Figure 7 on page 19. The same figure also
presents the minimum required attenuation for out-ofband analog signals applied on the line. The presence
of a high-pass filter transfer-function ensures at least
30 dB of attenuation for signals below 65 Hz. The lowpass filter transfer function attenuates signals above
3.4 kHz. It is implemented as part of the A-to-D
converter.
The receive path transfer function requirement, shown
in Figure 8 on page 20, is very similar to the transmit
path transfer function. The PCM data rate is 8 kHz so no
frequencies greater than 4 kHz are digitally encoded in
PCLK
the data stream. At frequencies greater than 4 kHz, the
plot in Figure 8 is interpreted as the maximum allowable
magnitude of spurious signals that are generated when
a PCM data stream representing a sine wave signal in
the range of 300 Hz to 3.4 kHz at a level of 0 dBm0 is
applied at the digital input.
The block diagram for the voice-band signal processing
paths are shown in Figure 11 on page 22. Both the
receive and the transmit paths employ the optimal
combination of analog and digital signal processing for
maximum performance while maintaining sufficient
flexibility for users to optimize their particular application
of the device. The two-wire (TIP/RING) voice-band
interface to the device is implemented with a small
number of external components. The receive path
interface consists of a unity-gain current buffer, IBUF,
while the transmit path interface is an ac coupling
capacitor. Signal paths, although implemented
differentially, are shown as single-ended for simplicity.
System Clock Generation
The Dual ProSLIC devices generate the internal clock
frequencies from the PCLK input. PCLK must be
synchronous to the 8 kHz FSYNC clock and run at one
of the following rates: 256 kHz, 512 kHz, 786 kHz,
1.024 MHz, 1.536 MHz, 1.544 MHz, 2.048 MHz,
4.096 MHz or 8.192 MHz. The ratio of the PCLK rate to
the FSYNC rate is determined by a counter clocked by
PCLK. The three-bit ratio information is transferred into
an internal register, PLL_MULT, after a device reset.
The PLL_MULT controls the internal PLL which
multiplies PCLK to generate the rate required to run the
internal filters and other circuitry.
The PLL clock synthesizer settles quickly after powerup
or update of the PLL_MULT register. However, the
settling time depends on the PCLK frequency and it is
approximately predicted by the following equation:
Tsettle = 64/FPCLK
VCO
PFD
÷2
÷2
28.672 MHz
DIV M
RESET
PLL_MULT
Figure 40. PLL Frequency Synthesizer
Preliminary Rev. 0.91
61
Si3220/Si3225
Interrupt Logic
The Dual ProSLIC devices are capable of generating
interrupts for the following events:
!
Loop current/ring ground detected.
Ring trip detected.
! Ground Key detected.
! Power alarm.
! DTMF digit detected.
! Active timer 1 expired.
! Inactive timer 1 expired.
! Active timer 2 expired.
! Inactive timer 2 expired.
! Ringing active timer expired.
! Ringing inactive timer expired.
! Pulse metering active timer expired.
! Pulse metering inactive timer expired.
! RAM address access complete.
! Receive path modem tone detected.
! Transmit path modem tone detected.
The interface to the interrupt logic consists of six
registers. Three interrupt status registers (IRQ0–IRQ3)
contain 1 bit for each of the above interrupt functions.
These bits are set when an interrupt is pending for the
associated resource. Three interrupt mask registers
(IRQEN1–IRQEN3) also contain 1 bit for each interrupt
function. For interrupt mask registers, the bits are active
high. Refer to the appropriate functional description text
for operational details of the interrupt functions.
!
When a resource reaches an interrupt condition, it
signals an interrupt to the interrupt control block. The
interrupt control block sets the associated bit in the
interrupt status register if the mask bit for that interrupt
is set. The INT pin is a NOR of the bits of the interrupt
status registers. Therefore, if a bit in the interrupt status
registers is asserted, IRQ asserts low. Upon receiving
the interrupt, the interrupt handler should read interrupt
status registers to determine which resource requests
service. All interrupt bits in the interrupt status registers
IRQ0–IRQ3 are cleared following a register read
operation. If the interrupt status registers are non-0, the
INT pin remains asserted.
eight devices (up to sixteen channels). The device
operates with both 8-bit and 16-bit SPI controllers. Each
SPI operation consists of a control byte, an address
byte (of which only the seven LSBs are used internally),
and either one or two data bytes depending on the width
of the controller and whether the access is to an 8-bit
register or 16-bit RAM address.
Bytes are always
transmitted MSB first.
There are variations of usage on this four-wire interface
as follows:
!
Continuous clocking. During continuous clocking,
the data transfers are controlled by the assertion of
the CS pin. CS must be asserted before the falling
edge of SCLK on which the first bit of data is
expected during a read cycle, and must remain low
for the duration of the 8-bit transfer (command/
address or data), going high after the last rising of
SCLK after the transfer.
! Clock during transfer only. In this mode, only the
clock is cycling during the actual byte transfers. Each
byte transfer will consist of eight clock cycles in a
return to “1” format.
! SDI/SDO wired operation. Independent of the
clocking options described, SDI and SDO can be
treated as two separate lines or wired together if the
master is capable of tri-stating its output during the
data byte transfer of a read operation.
! Soft reset. The SPI state machine resets whenever
CS asserts during an operation on an SCLK cycle
that is not a multiple of eight. This is a mechanism
for the controller to force the state machine to a
known state when the controller and the device are
out of synchronization.
SPI Control Interface
The control interface to the Dual ProSLIC devices is a
4-wire interface modeled after micro-controller and
serial peripheral devices. The interface consists of a
clock (SCLK), chip select (CS), serial data input (SDI),
and serial data output (SDO). In addition, the Dual
ProSLIC devices include a serial data through output
(SDI_THRU) to support a daisy-chain operation of up to
62
Preliminary Rev. 0.91
Si3220/Si3225
The control byte has the following structure and is presented on the SDI pin MSB first.
7
6
BRDCST
R/W
5
4
REG/RAM Reserved
3
2
1
0
CID[0]
CID[1]
CID[2]
CID[3]
The bits are defined as follows:
Table 37. SPI Control Interface
7
6
BRDCST Indicates a broadcast operation that is intended for all devices in the daisy chain. This is
only valid for write operations, since it would cause contention on the SDO pin during a
read.
R/W
Read/Write Bit.
0 = Write operation.
1 = Read operation.
5
REG/RAM Register/RAM Access Bit.
0 = RAM access.
1 = Register access.
4
Reserved
3:0
CID[3:0]
Indicates the channel that is targeted by the operation. The 4-bit channel value is provided
LSB first. The devices reside on the daisy chain such that device 0 is nearest to the controller and device 15 is furthest down the SDI/SDU_THRU chain. (See Figure 41.)
As the CID information propagates down the daisy chain, each channel decrements the
CID by 1. The SDI nodes between devices reflects a decrement of 2 per device since each
device contains two channels. The device receiving a value of 0 in the CID field responds
to the SPI transaction. (See Figure 42.) If a broadcast to all devices connected to the chain
is requested, the CID does not decrement. In this case, the same 8-bit or 16-bit data is presented to all channels regardless of the CID values.
Preliminary Rev. 0.91
63
Si3220/Si3225
SDI0
SDI
S DO
C PU
CS
CS
S DI
SDO
Channel 0
SDI1
Dual P roS LIC #1
Channel 1
SDITHRU
SDI2
SDI
CS
Channel 2
SDI3
SDO
Dual P roS LIC #2
Channel 3
SDITHRU
SDI4
SDI14
SDI
CS
Channel 14
SDI15
SDO
Channel 15
SDITHRU
Figure 41. SPI Daisy-Chain Mode
64
Preliminary Rev. 0.91
Dual P roS LIC #8
Si3220/Si3225
In Figure 42 the CID field is 0. As this field is decremented (in LSB to MSB order) the value decrements for each
SDI down the line. The BRDCST, R/W, and REG/RAM bits remain unchanged as the control word passes through
the entire chain. The odd SDIs are internal to the device and represent the SDI to SDI_THRU connection between
channels of the same device. A unique CID is presented to each channel, and the channel receiving a CID value of
zero is the target of the operation (channel 0 in this case). The last line of Figure 42 illustrates that in Broadcast
mode, all bits pass through the chain without permutation.
SPI Control Word
BRDCST
R/W
REG/RAM
Reserved
CID[0]
CID[1]
CID[2]
CID[3]
SDI0
0
A
B
C
0
0
0
0
SDI1 (Internal)
0
A
B
C
1
1
1
1
SDI2
0
A
B
C
0
1
1
1
SDI3 (Internal)
0
A
B
C
1
0
1
1
SDI 14
0
A
B
C
0
1
0
0
SDI15 (Internal)
0
A
B
C
1
0
0
0
SDI0-15
1
A
B
C
D
E
F
G
Figure 42. Sample SPI Control Word to Address Channel 0
Preliminary Rev. 0.91
65
Si3220/Si3225
Figures 43 and 44 illustrate WRITE and READ operations to register addresses via an 8-bit SPI controller. These
operations are performed as a 3-byte transfer. CS is asserted between each byte which is required for CS to be
asserted before the first falling edge of SCLK after the DATA byte to indicate to the state machine that one byte only
should be transferred. The state of SDI is a “don’t care” during the DATA byte of a read operation.
CS
SCLK
SDI
CONTROL
ADDRESS
DATA [7:0]
Hi-Z
SDO
Figure 43. Register Write Operation via an 8-Bit SPI Port
CS
SCLK
SDI
CONTROL
ADDRESS
XXXXXXXX
Data [7:0]
SDO
Figure 44. Register Read Operation via an 8-Bit SPI Port
CS
SCLK
SDI
CONTROL
ADDRESS
Data [7:0]
XXXXXXXX
Hi - Z
SDO
Figure 45. Register Write Operation via a 16-Bit SPI Port
CS
SCLK
SDI
CONTROL
ADDRESS
XXXXXXXX
Data [7:0]
SDO
XXXXXXXX
Data [7:0]
Same byte repeated twice.
Figure 46. Register Read Operation via a 16-Bit SPI Port
66
Preliminary Rev. 0.91
Si3220/Si3225
Figures 45 and 46 illustrate WRITE and READ
operations to register addresses via a 16-bit SPI
controller. These operations require a 4-byte transfer
arranged as two 16-bit words. The absence of CS going
high after the eighth bit of data indicates to the SPI state
machine that eight more SCLK pulses follow to
complete the operation. For a WRITE operation, the last
eight bits are ignored. For a read operation, the 8-bit
data value repeats so that the data is captured during
the last half of a data transfer if required by the
controller.
During register accesses, the CONTROL, ADDRESS,
and DATA are captured in the SPI module. At the
completion of the ADDRESS byte of a READ access,
the contents of the addressed register move into the
data register of the SPI data register. At the completion
of the DATA byte of a WRITE access, the data is
transferred from the SPI to the addressed register.
Figures 47–50 illustrate the various cycles for accessing
RAM addresses. RAM addresses are 16-bit entities;
therefore, the accesses always require four bytes.
During RAM address accesses, the CONTROL,
ADDRESS, and DATA are captured in the SPI module.
At the completion of the ADDRESS byte of a READ
access, the contents of the channel-based data buffer
move into the data register in the SPI for shifting out
during the DATA portion of the SPI transfer. This is the
data loaded into the data buffer in response to the
previous RAM address read request. Therefore, there is
a one-deep pipeline nature to RAM address READ
operations. At the completion of the DATA portion of the
READ cycle, the ADDRESS is transferred to the
channel-based address buffer register and a RAM
address is logged for that channel. The RAMSTAT bit in
each channel is polled to monitor the status of RAM
address accesses that are serviced twice per sample
period at dedicated windows in the DSP algorithm.
A RAM access interrupt in each channel indicates that
the pending RAM access request is serviced. For a
RAM access, the ADDRESS and DATA is transferred
from the SPI registers to the address and data buffers in
the appropriate channel. The RAM WRITE request will
be then logged. As for READ operations, the status of
the pending request is monitored by either polling the
RAMSTAT bit for the channel or enabling the RAM
access interrupt for the channel. By keeping the
address, data buffers, and RAMSTAT register on a per
channel basis, RAM address accesses can be
scheduled for both channels without interface.
CS
SCLK
SDI
CONTROL
ADDRESS
DATA [15:8]
DATA [7:0]
SDO
Hi-Z
Figure 47. RAM Write Operation via an 8-Bit SPI Port
CS
SCLK
SDI
SDO
CONTROL
ADDRESS
xxxxxxxx
xxxxxxxx
DATA [15:8]
DATA [7:0]
Figure 48. RAM Read Operation via an 8-Bit SPI Port
Preliminary Rev. 0.91
67
Si3220/Si3225
CS
SCLK
SDI
CONTROL
ADDRESS
Data [15:8]
Data [7:0]
Hi - Z
SDO
Figure 49. RAM Write Operation via a 16-Bit SPI Port
CS
SCLK
SDI
CONTROL
ADDRESS
Data [15:8]
SDO
Figure 50. RAM Read Operation via a 16-Bit SPI Port
68
Preliminary Rev. 0.91
Data [7:0]
Si3220/Si3225
PCM Interface
The Dual ProSLIC devices contain a flexible
programmable interface for the transmission and
reception of digital PCM samples. PCM data transfer is
controlled by the PCLK and FSYNC inputs, PCM Mode
Select, PCM Transmit Start Count (PCMTXHI/
PCMTXLO), and PCM Receive Start Count (PCMRXHI/
PCMRXLO) registers. The interface can be configured
to support from 4 to 128 8-bit timeslots in each frame.
This corresponds to PCLK frequencies of 256 kHz to
8.192 MHz in power of 2 increments. (768 kHz,
1.536 MHz and 1.544 MHz also are available.)
Timeslots for data transmission and reception are
independently configured with the PCMTXHI,
PCMTXLO, PCMRXHI, and PCMRXLO. Setting the
correct starting point of the data configures the part to
support long FSYNC and short FSYNC variants, IDL2 8bit, 10-bit, B1 and B2 channel time slots. DTX data is
high-impedance except for the duration of the 8-bit PCM
transmit. DTX returns to high-impedance on the
negative edge of PCLK during the LSB or on the
positive edge of PCLK following the LSB. This is based
on the setting of the PCMTRI bit of the PCM Mode
Select register. Tristating on the negative edge allows
the transmission of data by multiple sources in adjacent
timeslots without the risk of driver contention. In addition
to 8-bit data modes, there is a 16-bit mode provided for
testing. This mode can be activated via the PCMF bits
of the PCM Mode Select register. Setting the PCMTXHI/
PCMTXLO or PCMRXHI/PCMRXLO register greater
than the number of PCLK cycles in a sample period
stops data transmission because PCMTXHI/PCMTXLO
or PCMRXHI/PCMRXLO do not equal the PCLK count.
Figures 51–53 illustrate the usage of the PCM highway
interface to adapt to common PCM standards.
PCLK
FSYNC
0
PCLK_CNT
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
DRX
MSB
LSB
MSB
LSB
DTX
HI-Z
HI-Z
Figure 51. Example, Timeslot 1, Short FSYNC (TXS/RXS = 1)
PCLK
FSYNC
0
PCLK_CNT
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
DRX
MSB
LSB
MSB
LSB
DTX
HI-Z
HI-Z
Figure 52. Example, Timeslot 1, Long FSYNC (TXS/RXS = 0)
Preliminary Rev. 0.91
69
Si3220/Si3225
PCM Companding
The Dual ProSLIC devices support both µ-255 Law (µLaw) and A-Law companding formats in addition to
Linear Data mode. The data format is selected via the
PCMF bits of the PCM Mode Select register. µ-Law
mode is more commonly used in North America and
Japan, and A-Law is primarily used in Europe and other
countries. These 8-bit companding schemes follow a
segmented curve formatted as a sign bit (MSB) followed
by three chord bits and four step bits. A-Law typically
uses a scheme of inverting all even bits while µ-Law
does not. Dual ProSLIC devices also support A-Law
with inversion of even bits, inversion of all bits, or no bit
inversion by programming the ALAW bits of the PCM
Mode Select register to the appropriate setting. Tables
38 and 39 define the µ-Law and A-Law encoding
formats.
The Dual ProSLIC devices also support a 16-bit linear
data format with no companding. This Linear mode is
typically used in systems that convert to another
companding format such as adaptive delta PCM
(ADPCM) or systems that perform all companding in an
external DSP. The data format is 2’s complement with
MSB first (sign bit). Transmitting and receiving data via
Linear mode requires two continuous time slots. An 8-bit
Linear mode enables 8-bit transmission without
companding.
PCLK
FSYNC
0
PCLK_CNT
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
DRX
MSB
LSB
MSB
LSB
DTX
HI-Z
HI-Z
Figure 53. Example, IDL2 Long FSYNC, B2, 10-Bit Mode (TXS/RXS = 10)
PCLK
FSYNC
0
PCLK_CNT
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
DRX
DTX
MSB
LSB
MSB
LSB
HI-Z
HI-Z
Figure 54. 16-Bit Linear Mode Example, Timeslots 1 and 2, Long FSYNC
70
Preliminary Rev. 0.91
Si3220/Si3225
Table 38. µ-Law Encode-Decode Characteristics*
Segment
Number
#Intervals X Interval Size
Value at Segment Endpoints
Digital Code
Decode Level
8
16 X 256
8159
.
.
.
4319
4063
10000000b
8031
10001111b
4191
.
.
.
2143
2015
10011111b
2079
.
.
.
1055
991
10101111b
1023
.
.
.
511
479
10111111b
495
.
.
.
239
223
11001111b
231
.
.
.
103
95
11011111b
99
.
.
.
35
31
11101111b
33
.
.
.
3
1
0
11111110b
11111111b
2
0
7
6
5
4
3
2
1
16 X 128
16 X 64
16 X 32
16 X 16
16 X 8
16 X 4
15 X 2
__________________
1 X 1
*Note: Characteristics are symmetrical about analog zero with sign bit = 0 for negative analog values.
Preliminary Rev. 0.91
71
Si3220/Si3225
Table 39. A-Law Encode-Decode Characteristics1,2
Segment
Number
#intervals X interval size
Value at segment endpoints
7
16 X 128
4096
3968
.
.
2176
2048
6
5
4
3
2
1
16 X 64
16 X 32
16 X 16
16 X 8
16 X 4
32 X 2
Digital Code
Decode Level
10101010b
4032
10100101b
2112
.
.
.
1088
1024
10110101b
1056
.
.
.
544
512
10000101b
528
.
.
.
272
256
10010101b
264
.
.
.
136
128
11100101b
132
.
.
.
68
64
11110101b
66
.
.
.
2
0
11010101b
1
Notes:
1. Characteristics are symmetrical about analog zero with sign bit = 0 for negative values.
2. Digital code includes inversion of even numbered bits. Other available formats include inversion of odd bits, inversion of
all bits, or no bit inversion. See "PCM Companding" on page 70 for more details.
72
Preliminary Rev. 0.91
Si3220/Si3225
General Circuit Interface
The Dual ProSLIC devices also contain an alternate
communication interface to the SPI and PCM control
and data interface. The general circuit interface (GCI) is
used for the transmission and reception of both control
and data information onto a GCI bus. The PCM and GCI
interfaces are both four-wire interfaces and share the
same pins. The SPI control interface is not used as a
communication interface in the GCI mode, but rather as
hard-wired channel selector pins. The selection
between PCM and GCI modes is performed out of reset
using the SDITHRU pin. Tables 40 and 41 illustrate how
to select the communication mode and how the pins are
used in each mode.
If GCI mode is selected, the following pins must be tied
to the correct state to select one of eight sub-frame
timeslots in the GCI frame (described below). These
pins must remain in this state while the Dual ProSLIC is
operating. Selecting a particular subframe causes that
individual Dual ProSLIC device to transmit and receive
on the appropriate sub-frame in the GCI frame, which is
initiated by an FSYNC pulse. No further register settings
are needed to select which sub-frame a device uses,
and the sub-frame for a particular device cannot be
changed while in operation.
Table 42. GCI Mode Sub-Frame Selection
SDI
SDO
CS
GCI Subframe 0 Selected
(Voice channels 1–2)
1
1
1
GCI Subframe 1 Selected
(Voice channels 3–4)
1
1
0
GCI Subframe 2 Selected
(Voice channels 5–6)
1
0
1
Note: Values shown are the states of the pins at the rising
edge of RESET.
GCI Subframe 3 Selected
(Voice channels 7–8)
1
0
0
Table 41. Pin Functionality in PCM or GCI Mode
GCI Subframe 4 Selected
(Voice channels 9–10)
0
1
1
Pin Name
PCM Mode
1
0
SPI Chip Select
GCI Subframe 5 Selected
(Voice channels 11–12)
0
CS
GCI Subframe 6 Selected
(Voice channels 13–14)
0
0
1
GCI Subframe 7 Selected
(Voice channels 15–16)
0
0
0
Table 40. PCM or GCI Mode Selection
SDITHRU SCLK
0
0
1
SCLK
SDI
SDO
SDITHRU
FSYNC
PCLK
DTX
DRX
0
1
x
Mode Selected
GCI Mode—1x PCLK (2.048 MHz)
GCI Mode—2x PCLK (4.096 MHz)
PCM Mode
GCI Mode
Channel Selector,
bit 0
SPI Clock Input
PCLK Rate
Selector
SPI Serial Data Input Channel Selector,
bit 2
SPI Serial Data
Channel Selector,
Output
bit 1
SPI Data ThroughPCM/GCI Mode
put pin for Daisy
Selector
Chaining Operation
(Connects to the SDI
pin of the subsequent device in the
daisy chain)
PCM Frame Sync
GCI Frame Sync
Input
Input
PCM Input Clock
GCI Input Clock
PCM Data Transmit GCI Data Transmit
PCM Data Receive GCI Data Receive
Note: This table denotes pin functionality after the rising
edge of RESET and mode selection.
In GCI mode, the PCLK input requires either a
2.048 MHz or a 4.096 MHz clock signal, and the
FSYNC input requires an 8 kHz frame sync signal. The
overall unit of data used to communicate on the GCI
highway is a frame 125 µs in length. Each frame is
initiated by a pulse on the FSYNC pin, whose rising
edge signifies the beginning of the next frame. In 2x
PCLK mode, the user sees twice as many PCLK cycles
during each 125 µs frame versus 1x PCLK mode. Each
frame consists of eight fixed timeslot sub-frames, which
are assigned by the Sub-Frame Select pins as
described above (SDI, SDO, and CS). Within each subframe are four channels (bytes) of data, including two
voice data channels B1 and B2, one Monitor channel M
used for initialization and setup of the device, and one
Signaling and Control channel (SC) used for
communicating status of the device and for initiating
commands. Within the SC channel are six Command/
Preliminary Rev. 0.91
73
Si3220/Si3225
Indicate (C/I) bits and two handshaking bits, MR and
MX. The C/I bits indicate status and command
communication, while the handshaking bits Monitor
Receive (MR) and Monitor Transmit (MX), exchange
data in the Monitor channel. Figure 55 illustrates the
contents of a GCI highway frame.
to sixteen voice channels in 8-bit GCI mode). Table 43
describes the GCI mode sub-frame selection for 16-bit
GCI mode.
Table 43. Sub-Frame Selection 16-Bit GCI Mode
SDI
SDO
GCI Subframe 0 Selected
(Voice channels 0–1)
1
1
GCI Subframe 1 Selected
(Voice channels 2–3)
1
0
GCI Subframe 2 Selected
(Voice channels 4–5)
0
1
GCI Subframe 3 Selected
(Voice channels 6–7)
0
0
16-Bit GCI Mode
In addition to the standard 8-bit GCI mode, the Dual
ProSLIC devices also offer a 16-bit GCI mode for
passing 16-bit voice data to the upstream host
processor. This mode can be used for testing purposes
or for passing non-companded voice data to an
upstream DSP for further processing.
In 16-bit GCI mode, both of the 8-bit voice data
channels (B1 and B2, Figure 56) of each sub-frame are
required to pass the 16-bit voice data to the host. Each
125 µs frame can therefore accommodate up to eight
voice channels (the Dual ProSLIC can accommodate up
125 µs = 1 Frame
FS
SF0
SF1
SF2
SF3
SF4
SF5
SF6
Sub-Frame
8
8
8
B1
B2
M
SC
0
1
2
3
1
Channel
C/I
1
MR MX
Figure 55. Time-Multiplexed GCI Highway Frame Structure
74
Preliminary Rev. 0.91
SF7
Si3220/Si3225
125 µs = 1 Frame
FS
CH0
CH1
CH2
CH3
Sub-Frame
16
8
6
1
1
16
16
B1
M
C/I
MR
MX
B2
Unused
Figure 56. GCI Highway Frame Structure for 16-Bit GCI Mode
Monitor Channel
as the data sent by the Dual ProSLIC devices to a host.
The Monitor channel is used for initialization and setup
of the Dual ProSLIC devices. It is also for general
communication with the Dual ProSLIC by allowing read
and write access to the Dual ProSLIC devices registers.
Use of the monitor channel requires manipulation of the
MR and MX handshaking bits, located in bits 1 and 0 of
the SC channel described. For purposes of this
specification, “downstream” is identified as the data sent
by a host to the Dual ProSLIC. “Upstream” is identified
The following diagram illustrates the Monitor channel
communication protocol. For successful communication
with the Dual ProSLIC, the transmitter should anticipate
the falling edge of the receiver’s acknowledgement.
This also maximizes communication speed. Because of
the handshaking protocol required for successful
communication, the data transfer rate using the Monitor
channel is less than 8 kbps.
1st Byte
2nd Byte
3rd Byte
MX
Transm itter
MX
MR
Receiver
MR
ACK
1st Byte
ACK
2nd Byte
ACK
3rd Byte
125 µ s
Figure 57. Monitor Handshake Timing
Preliminary Rev. 0.91
75
Si3220/Si3225
The Idle state is achieved by the MX and MR bits being
held inactive for two or more frames. When a
transmission is initiated by a host device, an active state
is seen on the downstream MX bit. This signals the Dual
ProSLIC that a transmission has begun on the Monitor
channel and it should begin accepting data from it. The
Dual ProSLIC, after reading the data on the Monitor
channel, acknowledges the initial transmission by
placing the upstream MR bit in an active state. The data
is received and the upstream MR becomes active in the
frame immediately following the downstream MX
becoming active. The upstream MR then remains active
until either the next byte is received or an end of
message is detected (signaled by the downstream MX
being held inactive for two or more consecutive frames).
Upon receiving acknowledgement from the Dual
ProSLIC that the initial data was received (signaled by
the upstream MR bit transitioning from an inactive to an
active state), the host device places the downstream
MX bit in the inactive state for one frame and then either
transmit another byte by placing the downstream MX bit
in an active state again, or signal an end of message by
leaving the downstream MX bit inactive for a second
frame.
When the host is performing a write command, the host
only manipulates the downstream MX bit, and the Dual
ProSLIC only manipulates the upstream MR bit. If a
read command is performed, the host initially
manipulates the downstream MX bit to communicate
the command, but then manipulates the downstream
MR bit in response to the Dual ProSLIC responding with
the requested data. Similarly, the Dual ProSLIC initially
manipulates its upstream MR bit to receive the read
command, and will then manipulate its upstream MX bit
to respond with the requested data. If the host is
transmitting data, the Dual ProSLIC always transmits a
$FF value on its Monitor data byte. While the Dual
ProSLIC is transmitting data, the host should always
transmit a $FF value on its Monitor byte. If the Dual
ProSLIC is transmitting data and detects a value other
than a $FF on the downstream Monitor byte, the Dual
ProSLIC signals an Abort.
For read and write commands, an initial address must
be specified. The Dual ProSLIC responds to a read or a
write command at this address, and then subsequently
increment this address after every register access. In
this manner, multiple consecutive registers can be read
or written in one transmission sequence. By correctly
manipulating the MX and MR bits, a transmission
sequence can continue from the beginning specified
address until an invalid memory location is reached. To
end a transmission sequence, the host processor must
signal an End-of-Message (EOM) by placing the
76
downstream MX and MR bits inactive for two
consecutive frames. The transmission can also be
stopped by the Dual ProSLIC by signaling an Abort.
This is signaled by placing the upstream MR bit inactive
for at least two consecutive cycles in response to the
downstream MX bit going active. An abort is signaled by
the Dual ProSLIC for the following reasons:
!
A read or write to an invalid memory address is
attempted.
! An invalid command sequence is received.
! A data byte was not received for at least two
consecutive frames.
! A collision occurs on the Monitor data bytes while
the Dual ProSLIC is transmitting data.
! Downstream monitor byte not $FF while upstream
monitor byte is transmitting.
! MR/MX protocol violation
Whenever the Dual ProSLIC aborts due to an invalid
command sequence, the state of the Dual ProSLIC
does not change. If a read or write to an invalid memory
address is attempted, all previous reads or writes in that
transmission sequence are valid up to the read or write
to the invalid memory address. If an end-of-message is
detected before a valid command sequence is
communicated, the Dual ProSLIC returns to the idle
state and remains unchanged.
The data presented to the Dual ProSLIC in the
downstream Monitor bits must be present for two
consecutive frames to be considered valid data. The
Dual ProSLIC is designed to ensure it has received the
same data in two consecutive frames. If it does not, it
does not acknowledge receipt of the data byte and waits
until it does receive two consecutive identical data bytes
before acknowledging to the transmitter it has received
the data. If the transmitter attempts to signal
transmission of a subsequent data byte by placing the
downstream MX bit in an inactive state while the Dual
ProSLIC is still waiting to receive a valid data byte
transmission of two consecutive identical data bytes,
the Dual ProSLIC signals an abort and ends the
transmission. Figure 58 shows a state diagram for the
Receiver Monitor channel for the Dual ProSLIC.
Figure 59 shows a state diagram for the Transmitter
Monitor channel for the Dual ProSLIC.
Preliminary Rev. 0.91
Si3220/Si3225
Idle
MR = 1
M X * LL
Initial
S tate
MX
1s t By te
Rec eiv ed
MR = 0
MX
A bort
MR = 1
MX
ABT
MX
A ny
S tate
MX
MX
Wait
f or LL
MR = 0
M X * LL
M X * LL
By te
V alid
MR = 0
M X * LL
MX
M X * LL
MX
New By te
MR = 1
MX
nth by te
rec eiv ed
MR = 1
M X * LL
Wait
f or LL
MR = 0
MX
MR : MR bit calcu lated and tran s m itted on d ata ups tream (D TX) line.
MX: MX b it received d ata dow n s trea m (D R X) lin e.
LL: Las t look of m onitor b yte received on D R X line.
ABT: Abort ind ication to in terna l s ource.
Figure 58. Dual ProSLIC Monitor Receiver State Diagram
Preliminary Rev. 0.91
77
Si3220/Si3225
M R * M XR
M XR
Idle
MR = 1
M R * M XR
Wait
MX = 1
M R * M XR
A bort
MX = 1
Initial
S tate
M R * RQT
MR
1s t By te
MX = 0
M R * RQT
EOM
MX = 1
MR
M R * RQT
nth By te
ac k
MX = 1
MR
M R * RQT
MR
Wait f or
ac k
MX = 0
M R * RQT
CLS /A B T
A ny S tate
MR : MR bit received on D R X line.
MX: MX bit calculated and expected on D TX line.
MXR : MX bit s am pled on D TX line.
C LS: C ollis ion w ithin the m onitor data byte on D TX
line.
R QT: R eques t for trans m is s ion from internal s ource.
ABT: Abort reques t/indication.
Figure 59. Dual ProSLIC Monitor Transmitter State Diagram
Figures 60 and 61 are example timing diagrams of a
register read and a register write to the Dual ProSLIC
using the GCI. As noted in Figure 59, the transmitter
should always anticipate the acknowledgement of the
78
receiver for correct communication with the Dual
ProSLIC. Devices that do not accept this “best
case” timing scenario will not be able to
communicate with the Dual ProSLIC.
Preliminary Rev. 0.91
M onitor Da ta Dow nstre a m
$FF
$FF
$91
$91
$81
$81
$10
$10
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$FF
125 µ s
1 Frame
M X Dow nstre a m Bit
M R Dow nstre a m Bit
Preliminary Rev. 0.91
EOM Signalled
M onitor Da ta Upstre a m
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$91
$91
C ontents of C ontents of C ontents of C ontents of C ontents of
R eg ister $10 R eg ister $10 R eg ister $11 R eg ister $11 R eg ister $12
(ig nored by
host)
$FF
M X Upstre a m Bit
= Acknow ledgem ent of data reception
s ends
addres s before
data
Figure 60. Example Read of Registers $10 and $11 in Channel 0 of the Dual ProSLIC
EOM
Acknow ledge
79
Si3220/Si3225
M R Upstre a m Bit
$FF
$FF
$91
$91
$01
$01
$10
$10
D ata to be
D ata to be
D ata to be
D ata to be
written to $10 written to $10 written to $11 written to $11
$FF
$FF
125 µ s
1 Frame
M X Dow nstre a m Bit
M R Dow nstre a m Bit
Preliminary Rev. 0.91
EOM Signalled
M onitor Da ta Upstre a m
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$FF
$FF
M X Upstre a m Bit
M R Upstre a m Bit
= Acknow ledgem ent of data reception
EOM
Acknow ledge
Figure 61. Example Write to Registers $10 and $11 in Channel 0 of the Dual ProSLIC
Si3220/Si3225
80
M onitor Da ta Dow nstre a m
Si3220/Si3225
Programming the Dual ProSLIC Using the
Monitor Channel
The Dual ProSLIC devices use the monitor channel to
Transfer Status or Operating mode information to and
from the host processor. Communication with the Dual
ProSLIC should be in the following format:
Byte 1: Device Address Byte
Byte 2: Command Byte
MSB
LSB
Bit
7
6 5 4 3 2 1
0
Address Byte
1
0 0 A 0 0 0
0
Command Byte
0
0 0 0 0 0 0
0
Immediately after the last bit of the CID command is
received, the Dual ProSLIC responds with a fixed twobyte identification code as follows:
Byte 3: Register Address Byte
Bytes 4-n: Data Bytes
Bytes n+1, N+2: EOM
MSB
Device Address Byte
The Device Address Byte identifies which device
receives the particular message. This address must be
the first byte sent to the Dual ProSLIC at the beginning
of each transmission sequence. The Device Address
Byte has the following structure:
MSB
LSB
7
6
5
4
3
2
1
0
1
0
0
A
B
0
0
C
LSB
Bit
7
6 5 4 3 2 1
0
Address Byte
1
0 0 A 0 0 0
0
Command Byte
1
0 1 1 1 1 1
0
A = 1: Channel A is the source
A = 0: Channel B is the source
A = 1:
Channel A receives the command
A = 0:
Channel A does not receive the command
Upon sending the two-byte CID command, the Dual
ProSLIC sends an EOM signal (MR = MX = 1) for two
consecutive frames. When C = 0, B must be 0 or the
Dual ProSLIC signals an abort due to an invalid
command. In this mode, only bit C is programmable.
B = 1:
Channel B receives the command
Command Byte
B = 0:
Channel B does not receive the command
The Command Byte has the following structure:
C = 1:
Normal command follows
C = 0:
Channel identification command
MSB
When C = 1, bits A and B are channel enable bits.
When these bits are set to 1, the corresponding
channels receives the command in the next command
byte. The channels with corresponding bits set to 0
ignore the subsequent command byte.
Channel Identification (CID) Command
The lowest programmable bit of the Device Address
Byte, C, enables a special Channel Identification
Command to identify themselves by software. The
structure of this command is as follows:
A = 1: Channel A is the destination
A = 0: Channel B is the destination
LSB
RW
CMD[6:0]
RW = 1: A Read operation is performed from the Dual
ProSLIC
RW = 0: A Write operation is performed to the Dual
ProSLIC
CMD[6:0] = 0000001: Read or Write from the Dual
ProSLIC
CMD[6:0] = 0000010-1111111: Reserved
Register Address Byte
The Register Address Byte has the following structure:
MSB
LSB
ADDRESS[7:0]
This byte contains the actual 8-bit address of the
register to be read or written.
Preliminary Rev. 0.91
81
Si3220/Si3225
SC Channel
The downstream and upstream SC channels are
continuously carrying I/O information to and from the
Dual ProSLIC during every frame. The upstream
processor has immediate access to the receive
(downstream) and transmit (upstream) data present on
the Dual ProSLIC’s digital I/O port when used in GCI
mode. The SC channel consists of six C/I bits and two
handshaking bits as described in the tables below. The
functionality of the handshaking bits is defined in the
Monitor Channel section. This section defines the
functionality of the six C/I bits whether they are being
transmitted to the GCI bus via the DTX pin (upstream)
or received from the GCI bus via the DRX pin
(downstream). The structure of the SC channel is
shown in Figure 62.
MSB
LSB
7
6
5
4
3
2
1
0
CI2A
CI1A
CI0A
CI2B
CI1B
CI0B
MR
MX
Figure 62. SC Channel Structure
Downstream (Receive) SC Channel Byte
The first six bits in the downstream SC channel control
both channels of the Dual ProSLIC where the C/I bits
are defined as follows:
CI2A, CI1A, CI0A
Used to select operating mode for
channel A
CI2B, CI1B, CI0B
Used to select operating mode for
channel B
MR, MX
Monitor channel handshake bits
Table 44. Programming Operating Modes Using
Downstream SC Channel C/I Bits
Channel Specific C/I bits
Dual ProSLIC Operating
Mode
CI2x
CI1x
CI0x
0
0
0
Open (high impedence,
no line monitoring)
0
0
1
Forward Active
0
1
0
Forward On-Hook Transmission
0
1
1
Ground Start (Tip Open)
82
Table 44. Programming Operating Modes Using
Downstream SC Channel C/I Bits (Continued)
1
0
0
Ringing
1
0
1
Reverse Active
1
1
0
Reverse On-Hook Transmission
1
1
1
Ground Start (Ring Open)
Note: x = A or B, corresponding to channel A or channel B.
Figure 63 illustrates the transmission protocol for the C/I
bits within the downstream SC channel. New data
received by either channel must be present and match
for two consecutive frames to be considered valid.
When a new command is communicated via the
downstream C/I bits, this data must be sent for at least
two consecutive frames to be recognized by the Dual
ProSLIC.
The current state of the C/I bits is stored in a primary
register P. If the received C/I bits are identical to the
current state, no action is taken. If the received C/I bits
differ from those in register P, the new set of C/I bits is
loaded into secondary register S and a latch is set.
When the next set of C/I bits is received during the
frame that immediately follows, the following rules
apply:
!
If the received C/I bits are identical to the contents of
register S, the stored C/I bits are loaded into register
P and a valid C/I bit transition is recognized. The
latch is reset and the Dual ProSLIC responds
accordingly to the command represented by the new
C/I bits.
! If the received C/I bits differ from both the contents of
register S and the contents of register P, the newly
received C/I bits are loaded into register S and the
latch remains set. This cycle continues as long as
any new set of C/I bits differs from the contents of
registers S and P.
! If the newly received C/I bits are identical to the
contents of register P, the contents of register P
remain unchanged and the latch is reset.
Preliminary Rev. 0.91
Si3220/Si3225
Receive New
C/I Code
= P?
Yes
No
P: C/I Primary Register Contents
Store in S
S: C/I Secondary Register Contents
Receive New
C/I Code
= S?
Load C/I Register
With New C/I Bits
Yes
No
= P?
Yes
No
Figure 63. Protocol for Receiving C/I Bits in the Dual ProSLIC
When the Dual ProSLIC is set to GCI mode at
initialization, the default setting ignores the downstream
SC channel byte and allows linefeed state commands to
be directed through the monitor channel. This default
configuration is enabled by initializing the GCILINE bit
of the PCMMODE register to 0, which prevents the Dual
ProSLIC from transitioning between linefeed operating
states due to invalid data that may exist within the
downstream SC channel byte. To transfer direct linefeed
control to the downstream SC channel, the user must
set the GCILINE bit to 1. Once the GCILINE bit has
been set, the Dual ProSLIC follows the commands that
are contained in the downstream SC channel byte as
described in Figure 62.
The Dual ProSLIC architecture also enables automatic
transitions between linefeed operating states to reduce
the amount of interaction required between the host
processor and the Dual ProSLIC. When a GCI bus is
implemented, the user must ensure that these
automatic linefeed state transitions are consistent with
the linefeed commands contained within the
downstream SC channel byte.
In normal operation these automatic linefeed state
transitions are accompanied by the setting of a
threshold detection flag and an interrupt bit, if enabled.
To allow the Dual ProSLIC to automatically detect the
appropriate thresholds and control the linefeed
transitions, the downstream SC channel byte should be
updated accordingly once the interrupt bit is read from
the upstream SC channel byte. To disable the automatic
transitions, the user must set the GCILINE bit. Enabling
this Manual mode requires the host processor to read
the upstream SC channel information and provide the
appropriate downstream SC channel byte command to
program the correct linefeed state.
Table 45 presents the automatic linefeed state
transitions and their associated registers that cause the
transition.
The transition to the OPEN state stemming from power
alarm detection is intended to protect the Dual ProSLIC
circuit in the event that too much power is dissipated in
the Si3200 LFIC. This alarm is typically due to a fault in
the application circuit or on the subscriber loop, but can
be caused by intermittent power spikes depending on
the threshold to which the alarm is set. The user can reinitialize the linefeed operating state that was in effect
just prior to the power alarm by toggling the downstream
SC channel byte to the OPEN state for two consecutive
cycles and then resetting the downstream SC channel
byte to the intended linefeed state for two consecutive
Preliminary Rev. 0.91
83
Si3220/Si3225
cycles. If the Dual ProSLIC continues to automatically
transition to the OPEN state, the power alarm threshold
might be set incorrectly. If this problem persists after the
power alarm settings are verified, a system fault is
probable and the user should take measures to
diagnose the problem.
Table 45. Automatic Linefeed State Transitions
Initiating Action
Automatic Linefeed State
Transition
Loop closure detected On-hook active → off-hook active,
Off-hook active → on-hook active
Ring trip detected
Ringing → off-hook active
Ringing burst
cadence
Ringing → on-hook transmission
On-hook transmission → ringing
Power alarm detected Any state → open
Detection/Control Bits
Interrupt Enable/Status
Bits
LCR (Register 9)
LOOPE, LOOPS
(Register 16/19)
RTP (Register 9)
RTRIPE, RTRIPS
(Register 16/19)
T1EN, T2EN
(Register 23)
RINGT1E, RINGT2E,
RINGT1S, RINGT2S
(Register 15/18)
PQ1DL (RAM 50)
PQ1E, PQ1S
(Registers 17/20)
Upstream (Transmit) SC Channel Byte
The upstream SC channel byte looks similar to the
downstream SC channel byte, except that the
information quickly transfers the most time-critical
information from the Dual ProSLIC to the GCI bus. Each
upstream SC channel byte transfer from the Dual
ProSLIC lasts for at least two consecutive frames to
represent a valid transfer. The upstream C/I bits are
defined as follows:
CI2A, CI1A, CI0A
Monitors status data for channel A
CI2B, CI1B, CI0B
Monitors status data for channel B
MR, MX
Monitor channel handshake bits
(see Monitor Channel section)
Table 46. Monitored Data via Upstream SC Channel C/I Bits
C/I Bit
Information Provided
Context
CI2A
Interrupt information on channel A
CI1A
Hook status information on channel A
CI1A = 0: Channel A is on-hook
CI1A = 1: Channel A is off-hook
CI0A
Ground key information on channel A
CI0A = 0: No longitudinal current detected
CI0A = 1: Longitudinal current detected in ch A
CI2B
Interrupt information on channel B
CI1B
Hook status information on channel B
CI1A = 0: Channel B is on-hook
CI1A = 1: Channel B is off-hook
CI0B
Ground key information on channel B
CI0A = 0: No longitudinal current detected
CI0A = 1: Longitudinal current detected in ch B
The interrupt information for channels A and B is a
single bit that indicates that one or more interrupts might
exist on the respective channel. Each of the individual
interrupt flags (see registers 18–20) can be individually
masked by writing the appropriate bit in registers 21–23
84
CI2A = 0: No interrupt on channel A
CI2A = 1: Interrupt present on channel A
CI2A = 0: No interrupt on channel B
CI2A = 1: Interrupt present on channel B
to ignore specific interrupts. When using the GCI mode,
the user should verify that each of the desired interrupt
bits are set so the upstream SC channel byte includes
the required interrupt functions.
Preliminary Rev. 0.91
Si3220/Si3225
System Testing
Line Test and Diagnostics
The Dual ProSLIC devices include a complete suite of
test tools to test the functionality of the line card and
detect fault conditions present on the TIP/RING pair.
Using one of the loopback test modes with the signal
generation and measurement tools eliminates the need
for per-line test relays and centralized test equipment.
The Dual ProSLIC devices provide a variety of signal
generation and measurement tools that facilitate fault
detection and parametric diagnostics on the TIP/RING
pair and line card functionality verification. The Dual
ProSLIC generates test signals, measures the
appropriate voltage/current/signal levels, and processes
the results to provide a meaningful result to the user.
Interaction is required from the host microprocessor to
load the test parameters into the appropriate registers,
initiate the test(s), and read the results from the
registers. In some cases, the host processor might also
be required to perform some simple mathematics to
achieve the results. Software modules are available to
simplify integration of the diagnostics functions into the
system. The need for test relays and a separate test
head is eliminated in most applications. To address
legacy applications, all versions of the Dual proSLIC
include test-in and test-out relay drivers to switch in a
centralized test head.
Loopback Modes
Three loopback test options are available for the Dual
ProSLIC devices:
!
The codec loopback path encompasses almost
entirely the electronics of both the transmit and
receive paths. The analog signal at the output of the
receive path is fed back to the input of the transmit
path through a feedback path on the analog side of
the audio codec. Both the impedance synthesis and
transhybrid balance functions are disabled in this
mode. (See DLM3 path in Figure 11 on page 22.)
The signal path starts with 8-bit PCM data input to
the receive path and ends with 8-bit PCM data at the
output of the transmit path. The user can bypass the
companding process and interface directly to the 16bit data.
! A second digital loopback takes the receive path
digital stream and routes it back to the transmit path
via the transhybrid feedback path. (See DLM2 path
through block H in Figure 11.) This mode
characterizes the transhybrid filter response. The
transhybrid block also can be disabled (set to unity
gain) in this mode for diagnosing the digital gain
blocks and filter stages in both transmit and receive
paths. The signal path starts with 8-bit PCM data
input to the receive path and ends with 8-bit PCM
data at the output of the transmit path. The user can
bypass the companding process and interface
directly to the 16-bit data.
! A third digital loopback takes the digital stream at the
output of the µ-Law/A-Law expander and feeds it
back to the input of the µ-Law/A-Law compressor.
(See DLM1 path in Figure 11.) This path verifies that
the host is connected correctly with the Dual
ProSLIC through the PCM interface and that the
PCLK and FSYNC signals are correctly set. This
mode also can test the µ-Law/A-Law companding
process. The signal path starts with 8-bit PCM data
input to the receive path and ends with 8-bit PCM
data at the output of the transmit path. The user can
also connect directly to the 16-bit data to eliminate
the µ-Law/A-Law companding process when testing
the PCM interface.
The Dual ProSLIC’s line test and diagnostics
capabilities are categorized into three sections: signal
generation tools, measurement tools, and diagnostics
capabilities. Using these signal generation and
measurement tools, a variety of other diagnostics
functions can also be performed to meet the unique
requirements of specific applications. Table 47
summarizes the ranges and capabilities of the signal
generation and measurement tools.
Preliminary Rev. 0.91
85
Si3220/Si3225
Table 47. Summary of Signal Generation and Measurement Tools
Function
Range
Accuracy/Resolution
Comments
Signal Generation Tools
DC Current Generation
18 to 45 mA
0.875 mA
DC Voltage Generation
0 to 63.3 V
1.005 V
Audio Tone Generation
200 to 3400 Hz
Ringing Signal Generation
4 to 15 Hz
16 to 100 Hz
±5%
±1%
Measurement Tools
8-Bit DC/Low Frequency
Monitor A/D Converter
High Range:
0 to 160.173 V
0 to 101.09 mA
628 mV
396.4 µA
Low Range: 0 to 64.07 V
0 to 50.54 mA
251 mV
198.2 µA
Programmable Timer
0 to 8.19 s
125 µs
AC Low Pass Filter
3 to 400 Hz
16-Bit Audio A/D Converter
0 to 2.5 V
Transmit Path Notch Filter
300 to 3400 Hz
Transmit Path Bandpass Filter
300 to 3400 Hz
Signal Generation Tools
!
TIP/RING DC signal generation. The Dual ProSLIC
line feed D/A converter can program a constant
current linefeed from 18–45 mA in 0.87 mA steps
with a ±10% total accuracy. In addition, the opencircuit TIP/RING voltage can be programmed from 0
to 63 V in 1 V steps. The linefeed circuitry also can
generate a controlled polarity reversal.
! Tone generation. The Dual ProSLIC devices can
generate single or dual tones over the entire audio
band, and can direct them into either the transmit or
the receive path depending on the diagnostic
requirements. Ringing signals from 4–100 Hz also
can be generated.
! Diagnostics mode ringing generation. The Dual
ProSLIC devices can generate an internal low-level
ringing signal to test for the presence of REN without
causing the terminal equipment to ring audibly. This
ringing signal can be either balance or unbalanced
86
800 Hz update rate
acrms, acPK, and dc
post-processing blocks
38 µV
Single or dual notch,
≥90 dB attenuation
depending on the state of the RINGUNB bit of the
RINGCON register. This feature is also available
with the Si3225 provided that a sufficient battery
voltage is present.
Preliminary Rev. 0.91
Si3220/Si3225
PEAK
DETECT
DIAGPK
VTIP
VRING
DIAGDCCO
VLOOP
VLONG
DIAGDC
LPF
ILOOP
DIAGACCO
ILONG
VRING,EXT
FULL WAVE
RECTIFY
IRING,EXT
LPF
DIAGAC
Figure 64. SLIC Diagnostic Filter Structure
Measurement Tools
"
VLOOP = VTIP-VRING = metallic (loop) voltage
!
"
VLONG = (VTIP+VRING)/2 = longitudinal voltage
"
ILOOP = ITIP-IRING = metallic (loop) current
"
ILOOP = ITIP-IRING = metallic (loop) current
"
ILONG = (ITIP+IRING)/2 = longitudinal current
"
VRING, EXT = ringing voltage when using an external
ringing source (Si3225 only)
IRING,EXT = ringing current when using an external
ringing source (Si3225 only)
8-Bit monitor A/D converter. This 8-bit A/D
converter monitors all dc and low frequency voltage
and current data from TIP to ground and RING to
ground. Two additional values, TIP – RING and
TIP + RING, are calculated and stored in on-chip
registers to analyze metallic and longitudinal effects.
The A/D operates at an 800 Hz update rate to allow
measurement bandwidth from dc to 400 Hz. A dualrange capability allows high-voltage/high-current
measurement in the high range but also can
measure lower voltages and currents with a tighter
resolution.
! Programmable bandpass filter. A bandpass filter
discriminates certain frequency ranges such as
ringing frequencies and 50 Hz/60 Hz induction from
nearby or crossed power leads.
! SLIC diagnostics filter. Several post-processing
filter blocks monitor peak dc and ac characteristics of
the Monitor A/D converter outputs and values
derived from these outputs. Setting the SDIAG bit in
the DIAG register enables the filters. There are
separate filters for each channel, and their control is
independent. These filters require DSP processing
which is available only when voice band processing
is not being performed. If an off-hook or a ring trip
condition is detected while the SDIAG bit is set, the
bit is cleared and the diagnostic information is not
processed.
The following parameters can be selected as inputs
to the diagnostic block by setting the SDIAG bits in
the DIAG register to values 0–7 corresponding to the
order below:
"
VTIP = voltage on the TIP lead
"
VRI NG = voltage on the RING lead
"
The SLIC diagnostic capability consists of a peak detect
block and two filter blocks, one for dc and one for ac.
The topology is illustrated in Figure 64.
The peak detect filter block reports the magnitude of the
largest positive or negative value without sign. The dc
filter block consists of a single pole IIR low pass filter
with a coefficient held in the DIAGDCCO RAM location.
The filter output is read from the DIAGDC RAM location.
The ac filter block consists of a full wave rectifier,
followed by a single pole IIR low pass filter with a
coefficient held in the DIAGACCO RAM location. The
peak value is read from the DIAGPK RAM location. The
peak value is cleared and the filters are flushed on the
0-1 transition of the SDIAG bit and when the input
source changes. The user can write 0 to the DIAGPK
RAM location to get peak information for a specific time
interval.
!
16-bit audio A/D converter. The A/D converter
portion of the audio codec is made available for
processing test data received back through the
transmit audio path. The audio path offers a 2.5 V
peak voltage measurement capability and a coarse
attenuation stage for scenarios where the incoming
signal amplitude must be attenuated by as much as
3 dB to bring it into the allowable input range without
clipping.
Preliminary Rev. 0.91
87
Si3220/Si3225
!
Programmable timer. The Dual ProSLIC devices
incorporate several digital oscillator circuits to
program the on- and off-times of the ringing and
pulse metering signals. The tone generation
oscillator can be used to program a time period for
averaging specific measured test parameters.
! Transmit audio path diagnostics filter. Transmit
path audio diagnostics are facilitated by
implementing a sixth-order IIR filter followed by peak
detection and power estimation blocks. This filter
can be programmed to eliminate or amplify specific
signals for the purpose of measuring the peak
amplitude and power content of individual
components in the audio spectrum. Figure 11 on
page 22 illustrates the location of the diagnostics
filter block.
The sixth order IIR filter operates at an 8 kHz sample
rate and is implemented as three second-order filter
stages in cascade. Each second-order filter offers
five fully programmable coefficients (a1, a2, b0, b1,
and b2) with 25-bit precision by providing several
user-accessible registers. Each filter stage is
implemented with the following format:
–1
The power averaging filter time constant is absolute
value programmable, and the average power result is
read from the TESTAVO RAM location.
Diagnostics Capabilities
!
!
!
–2
( b0 + b1z + b2z )
H ( z ) = ------------------------------------------------------–1
–2
( 1 – a1z – a2z )
If any of the second-order filter stages are not
required, they can be programmed to H(z)=1 by
setting a1=0, a2=0, b0=1, b1=0, and b2=0. This
flexible filter block can be programmed any of the
following characteristics:
"
"
"
"
Single notch. Used for measuring noise/distortion in
the presence of a single tone. 90 dB attenuation is
provided at the notched frequency. Implemented by
placing two 0s on the unit circle at the notch frequency
and two poles inside the unit circle at the notch
frequency.
Dual notch. Used for measuring noise/distortion in the
presence of dual tones.
Single notch/single peak. Used for measuring
particular harmonic in the presence of a single tone.
Dual notch/single peak. Used to measure particular
intermodulation product in the presence of dual tones.
Because each second-order filter stage is fully
programmable, there are many other possible filter
implementations.
The IIR filter output is measured for power and peak
post-processing. The peak measurement window
duration is programmable by entering a value into the
TESTWLN RAM address. The peak value (TESTPKO)
is updated at the end of each window period. Power
measurement is performed by using a single pole IIR
filter to average the output of the sixth-order IIR filter.
88
!
!
Foreign voltages test. The Dual ProSLIC devices
can detect the presence of foreign voltages
according to GR-909 requirements of ac voltages >
10 V and dc voltages > 6 V from T-G or R-G. This
test is performed when it has been determined that a
hazardous voltage is not present on the line.
Resistive faults (leakage current) test. Resistive
fault conditions are measured from T-G, R-G, or T-R
for dc resistance per GR-909 specifications. If the dc
resistance is < 150 kΩ, it is considered a resistive
fault. To perform this test, program the Dual ProSLIC
chipset to generate a constant open-circuit voltage
and measure the resulting current. The resistance is
then calculated.
Receiver off-hook test. Uses a similar procedure
as described in the Resistive Faults test above, but
is measured across T-R only. In addition, two
measurements must be performed at different opencircuit voltages to verify the resistive linearity. If the
calculated resistance has more than 15%
nonlinearity between the two calculated points and
the voltage/current origin, it is determined to be a
resistive fault.
Ringers (REN) test. Verifies the presence of REN at
the end of the TIP/RING pair per TA-909
specifications. It can be implemented by generating
a 20 Hz ringing signal between 7 Vrms and 17 Vrms
and measuring the 20 Hz ac current using the 8-bit
monitor ADC. The resistance (REN) can then be
calculated using the software module. The
acceptable REN range is > 0.175 REN (< 40 kΩ) or
< 5 REN (> 1400 Ω). A returned value of < 1400 Ω is
determined to be a resistive fault from T-R, and a
returned value > 40 kΩ is determined to be a loop
with no handset attached.
ac line impedance measurement. Determines the
ac loop impedence across T-R. It can be
implemented by sending out multiple discrete tones,
one at a time, and measuring the returned amplitude
with the hybrid balance filter disabled. By calculating
the voltage difference between the initial amplitude
and the received amplitude and dividing the result by
the audio current, the line impedance can then be
calculated.
Preliminary Rev. 0.91
Si3220/Si3225
!
!
!
!
!
!
Line capacitance measurement. Implemented like
the ac line impedance measurement test above, but
the frequency band of interest is between 1 kHz and
3.4 kHz. Knowing the synthesized 2-wire impedance
of the Dual ProSLIC, the roll-off effect can be used to
calculate the ac line capacitance.
Ringing voltage verification. Verifies that the
desired ringing signal is correctly applied to the TIP/
RING pair and can be measured in the 8-bit monitor
ADC, which senses low frequency signals directly
across T-R.
Idle channel noise measurement. Given any
transmission mode with no tone generated and the
hybrid balance filter turned off, the voice band
energy can be measured through the normal audio
path and read through the appropriate register.
Echo path gain measurement.
Harmonic distortion measurement. Detects the
power content of a particular harmonic. It can be
implemented by programming two of the IIR
diagnostics filter stages to provide a notch at the
fundamental frequency and a peak at the harmonic
of interest. Performing this procedure on all relevant
harmonics individually and summing the results
provide the total harmonic distortion.
Intermodulation distortion measurement (twotone method). Measures the intermodulation
distortion product in the presence of two tones. It can
be implemented by programming the three IIR
diagnostics filter stages to provide two notches at the
two tone frequencies and a peak at the frequency of
interest.
Preliminary Rev. 0.91
89
Si3220/Si3225
8-Bit Control Register Summary
Any register not listed here is reserved and must not be written. Shaded registers are read only. All registers are
assigned a default value during initialization and following a system reset. Only registers 0, 2, 3, and 14 are
available until a PLL lock is established or during a clock failure. Refer to AN58 “Dual ProSLIC Programmer Guide”
for detailed register descriptions and recommended settings.
(Ordered alphabetically by mnemonic except in cases of high, medium and low bytes which are ordered high to low.)
Reg
Addr3
Mnemonic
Description
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Type
R/W
Def.
Hex
Init
R/W
0x00
Audio
21
AUDGAIN
Audio Gain Control
ATXMUTE
ARXMUTE
Calibration
11
CALR1
Calibration Register 1
12
CALR2
Calibration Register 2
CAL
CALOFFR
CALOFFT
CALOFFRN
CALOFFTN
CALDIFG
CALCMG
Init
R/W
0x3F
CALLKGR
CALLKGT
CALMADC
CALDACO
CALADCO
CALCMBAL
Init
R/W
0x3F
Diag
R/W
Diag
R/W
0x00
Oper
R/W
0x00
Init
R
0x—
Init
R/W
0x05
Diagnostic Tools
13
DIAG
Diagnostics Tool Enable
IQ2HR
IQ1HR
TSTRING
TXFILT
SDIAG
SDIAGIN[2:0]
Digital Control and Loopback
22
DIGCON
Digital Control and
Loopback Enable
CODECLB
PCMLB
HYBLB
HYBDIS
THPFDIS
RHPFDIS
DTXMUTE
DRXMUTE
FSK Data Byte
68
FSKDAT
FSK Data Byte
FSKBYTE[7:0]
Chip ID
0
ID
PARTNUM[2:0]4
Chip ID
REV[3:0]4
Loop Current Limit
10
ILIM
Loop Current Limit
ILIM[4:0]
Interrupts
14
IRQ0
Interrupt Status 0
CLKIRQ4,6
IRQ3B4,6
IRQ2B4,6
IRQ1B4,6
IRQ3A4,6
IRQ2A4,6
IRQ1A4,6
Oper
R
0x00
15
IRQ1
Interrupt Status 1
PULSTAS
PULSTIS
RINGTAS
RINGTIS
OS2TAS
OS2TIS
OS1TAS
OS1TIS
Oper
R/W
0x00
16
IRQ2
Interrupt Status 2
RXMDMS
TXMDMS
RAMIRS
DTMFS
VOCTRKS
LONGS
LOOPS
RTRIPS
Oper
R/W
0x00
17
IRQ3
Interrupt Status 3
CMBALS
PQ6S
PQ5S
PQ4S
PQ3S
PQ2S
PQ1S
Oper
R/W
0x00
18
IRQEN1
Interrupt Enable 1
PULSTAE
PULSTIE
RINGTAE
RINGTIE
OS2TAE
OS2TIE
OS1TAE
OS1TIE
Init
R/W
0x00
19
IRQEN2
Interrupt Enable 2
RXMDME
TXMDME
RAMIRE
DTMFE
VOCTRKE
LONGE
LOOPE
RTRIPE
Init
R/W
0x00
20
IRQEN3
Interrupt Enable 3
CMBALE
PQ6E
PQ5E
PQ4E
PQ3E
PQ2E
PQ1E
Init
R/W
0x00
VOCTST4
LONGHI4
RTP4
LCR4
Oper
R
0x40
Oper
R/W
0x00
Linefeed Control
9
LCRRTP
Loop Closure/Ring Trip/
Ground Key
Detection
CMH4
6
LINEFEED
Linefeed
LFS[2:0]4
Notes:
1.
2.
3.
4.
5.
6.
7.
90
SPEED4
LF[2:0]
Any register not listed is reserved and must not be written. Default hex value is loaded to register following any RESET. Only registers ID, MSTREN, MSTRSTAT, and IRQ0 are valid while the
PLL is not locked (MSTRSTAT[PLOCK]).
Reserved bit values are indeterminate.
Register address is in decimal.
Read only.
Protected bits.
Per channel bit(s).
Si3220 only.
Preliminary Rev. 0.91
Si3220/Si3225
Reg
Addr3
Mnemonic
Description
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Type
R/W
Def.
Hex
Init
R/W
0x00
Init
R/W
0x00
SPI
2
MSTREN
Master Initialization
Enable
PLLFLT
FSFLT
PCFLT
3
MSTRSTAT
Master Initialization
Status
PLLFAULT
FSFAULT
PCFAULT
SRCLR4
PLOCK4
FSDET4
FSVAL4
PCVAL4
Oscillators
61
O1TAHI
Oscillator 1 Active Timer—
High Byte
OSC1TA[15:8]
Init
R/W
0x00
60
O1TALO
Oscillator 1 Active Timer—
Low Byte
OSC1TA[7:0]
Init
R/W
0x00
63
O1TIHI
Oscillator 1 Inactive Timer—
High Byte
OSC1TI[15:8]
Init
R/W
0x00
62
O1TILO
Oscillator 1 Inactive Timer—
Low Byte
OSC1TI[7:0]
Init
R/W
0x00
65
O2TAHI
Oscillator 2 Active Timer—
High Byte
OSC2TA[15:8]
Init
R/W
0x00
64
O2TALO
Oscillator 2 Active Timer—
Low Byte
OSC2TA[7:0]
Init
R/W
0x00
67
O2TIHI
Oscillator 2 Inactive Timer—
High Byte
OSC2TI[15:8]
Init
R/W
0x00
66
O2TILO
Oscillator 2 Inactive Timer—
Low Byte
OSC2TI[7:0]
Init
R/W
0x00
59
OCON
Oscillator Control
ENSYNC24
OSC2TAEN
Oper
R/W
0x00
58
OMODE
Oscillator Mode Select
FSKSSEN
ZEROEN2
ROUT1[1:0]
Init
R/W
0x00
6
6
PCMF[1:0]6
Init
R/W
0x05
PCMRX[9:8]
Init
R/W
0x00
Init
R/W
0x00
Init
R/W
0x00
Init
R/W
0x00
OSC2TIEN
OSC2EN
ROUT2[1:0]
ENSYNC14
OSC1TAEN
OSC1FSK
ZEROEN1
OSC1TIEN
OSC1EN
PCM Control
53
PCMMODE
PCM Mode Select
57
PCMRXHI
PCM RX Clock Slot—
High Byte
56
PCMRXL0
PCM RX Clock Slot—
Low Byte
55
PCMTXHI
PCM TX Clock Slot—
High Byte
54
PCMTXLO
PCM TX Clock Slot—
Low Byte
GCILINE
PCLK2X
PCMTRI6
ALAW[1:0]6
PCMEN
PCMRX[7:0]
PCMTX[9:8]
PCMTX[7:0]
Pulse Metering
28
PMCON
Pulse Metering Control
ENSYNC4,7
TAEN17
TIEN17
PULSE17
Oper
R/W
0x00
Init
R/W
0x00
30
PMTAHI
Pulse Metering Oscillator
Active Timer—
High Byte
PULSETA[15:8]7
29
PMTALO
Pulse Metering Oscillator
Active Timer—
Low Byte
PULSETA[7:0]7
Init
R/W
0x00
32
PMTIHI
Pulse Metering Oscillator
Inactive Timer—
High Byte
PULSETI[15:8]7
Init
R/W
0x00
Notes:
1.
2.
3.
4.
5.
6.
7.
Any register not listed is reserved and must not be written. Default hex value is loaded to register following any RESET. Only registers ID, MSTREN, MSTRSTAT, and IRQ0 are valid while the
PLL is not locked (MSTRSTAT[PLOCK]).
Reserved bit values are indeterminate.
Register address is in decimal.
Read only.
Protected bits.
Per channel bit(s).
Si3220 only.
Preliminary Rev. 0.91
91
Si3220/Si3225
Reg
Addr3
Mnemonic
Description
31
PMTILO
Pulse Metering Oscillator
Inactive Timer—
Low Byte
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
PULSETI[7:0]7
Type
R/W
Def.
Hex
Init
R/W
0x00
Polarity Reversal
7
POLREV
POLREV4
Polarity Reversal Settings
VOCZERO
PREN
RAMP
Init
R/W
RAM Access
103
RAMADDR
RAM Address
RAMADDR[7:0]
Oper
R/W
0x00
102
RAMDATHI
RAM Data—
High Byte
RAMDAT[15:8]
Oper
R/W
0x00
101
RAMDATLO
RAM Data—
Low Byte
RAMDAT[7:0]
Oper
R/W
0x00
4
RAMSTAT
RAM Address Status
RAMSTAT4
Init
R
0x00
RESETB
RESETA
Init
R/W
0x00
UNBPOLR
TRAP
Init
R/W
0x00
Soft Reset
1
RESET
Soft Reset
Ringing
ENSYNC4
RDACEN4
RINGUNB
TAEN
TIEN
RINGEN4
23
RINGCON
Ringing Configuration
25
RINGTAHI
Ringing Oscillator
Active Timer—High Byte
RINGTA[15:8]
Init
R/W
0x00
24
RINGTALO
Ringing Oscillator
Active Timer—Low Byte
RINGTA[7:0]
Init
R/W
0x00
27
RINGTIHI
Ringing Oscillator
Inactive Timer—High Byte
RINGTI[15:8]
Init
R/W
0x00
26
RINGTILO
Ringing Oscillator
Inactive Timer—Low Byte
RINGTI[7:0]
Init
R/W
0x00
Diag
R/W
0x00
Init
R/W
0xE0
Oper
R/W
0x45
PASSCNT[3:0]
Oper
R/W
0x00
DTMFDIGIT[3:0]4
Oper
R
0x00
Init
R/W
0x00
Init
R/W
0x00
Init
R/W
0x00
Relay Configuration
5
RLYCON
BSEL5
Relay Driver and
Battery Switching
Configuration
RRAIL
RDOE
RRD/GPO
TRD2
TRD1
SLIC Bias Control
8
SBIAS
OBIAS[1:0]5
SLIC Bias Control
ABIAS[1:0]5
Si3200 Thermometer
72
THERM
Si3200 Thermometer
4
STAT
Tone Detection
70
TONDET
Modem Tone Detection
69
TONDTMF
DTMF Detection
71
TONDEN
Tone Detection Enable
FAILCNT[3:0]
VALID4
VALTONE4
DTMF
RXMDM
TXMDM
Impedance Synthesis Coefficients
49
ZA1HI
Impedance Synthesis
Coeff A1— High Byte
48
ZA1MID
Impedance Synthesis
Coeff A1—Middle Byte
Notes:
1.
2.
3.
4.
5.
6.
7.
92
COEFFA1[20:16]6
COEFFA1[15:8]6
Any register not listed is reserved and must not be written. Default hex value is loaded to register following any RESET. Only registers ID, MSTREN, MSTRSTAT, and IRQ0 are valid while the
PLL is not locked (MSTRSTAT[PLOCK]).
Reserved bit values are indeterminate.
Register address is in decimal.
Read only.
Protected bits.
Per channel bit(s).
Si3220 only.
Preliminary Rev. 0.91
Si3220/Si3225
Reg
Addr3
Mnemonic
Description
47
ZA1LO
Impedance Synthesis
Coeff A1—Low Byte
52
ZA2HI
Impedance Synthesis
Coeff A2—High Byte
51
ZA2MID
Impedance Synthesis
Coeff A2—Middle Byte
50
ZA2LO
37
Type
R/W
Def.
Hex
Init
R/W
0x00
Init
R/W
0x00
COEFFA2[15:8]6
Init
R/W
0x00
Impedance Synthesis
Coeff A2—Low Byte
COEFFA2[7:0]6
Init
R/W
0x00
ZB0HI
Impedance Synthesis
Coeff B0—High Byte
COEFFB0[23:16]6
Init
R/W
0x00
36
ZB0MID
Impedance Synthesis
Coeff B0—Middle Byte
COEFFB0[15:8]6
Init
R/W
0x00
35
ZB0LO
Impedance Synthesis
Coeff B0—Low Byte
COEFFB0[7:0]6
Init
R/W
0x00
40
ZB1HI
Impedance Synthesis
Coeff B1—High Byte
COEFFB1[23:16]6
Init
R/W
0x00
39
ZB1MID
Impedance Synthesis
Coeff B1—Middle Byte
COEFFB1[15:8]6
Init
R/W
0x00
38
ZB1LO
Impedance Synthesis
Coeff B1— Low Byte
COEFFB1[7:0]6
Init
R/W
0x00
43
ZB2HI
Impedance Synthesis
Coeff B2—High Byte
COEFFB2[23:16]6
Init
R/W
0x00
42
ZB2MID
Impedance Synthesis
Coeff B2—Middle Byte
COEFFB2[15:8]6
Init
R/W
0x00
41
ZB2LO
Impedance Synthesis
Coeff B2— Low Byte
COEFFB2[7:0]6
Init
R/W
0x00
46
ZB3HI
Impedance Synthesis
Coeff B3—High Byte
COEFFB3[23:16]6
Init
R/W
0x00
45
ZB3MID
Impedance Synthesis
Coeff B3—Middle Byte
COEFFB3[15:8]6
Init
R/W
0x00
44
ZB3LO
Impedance Synthesis
Coeff B3—Low Byte
COEFFB3[7:0]6
Init
R/W
0x00
33
ZRS
Impedance Synthesis
Analog Real Coeff
Init
R/W
0x00
34
ZZ
Impedance Synthesis
Analog Complex Coeff
Init
R/W
0x00
Notes:
1.
2.
3.
4.
5.
6.
7.
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
COEFFA1[7:0]
COEFFA2[20:16]6
RS[3:0]6
ZSDIS6
ZSOHT6
ZP[1:0]6
ZZ[1:0]6
Any register not listed is reserved and must not be written. Default hex value is loaded to register following any RESET. Only registers ID, MSTREN, MSTRSTAT, and IRQ0 are valid while the
PLL is not locked (MSTRSTAT[PLOCK]).
Reserved bit values are indeterminate.
Register address is in decimal.
Read only.
Protected bits.
Per channel bit(s).
Si3220 only.
Preliminary Rev. 0.91
93
Si3220/Si3225
16-Bit RAM Address Summary
All internal 16-bit RAM addresses can be assigned unique values for each SLIC channel and are accessed in a
similar manner as the 8-bit control registers except the data is twice as long. In addition, one more READ cycle is
required during READ operations to accommodate the one-deep pipeline architecture. See "SPI Control Interface"
on page 62 for more details. All internal RAM addresses are assigned a default value of 0 during initialization and
following a system reset. Unless otherwise noted, all RAM addresses use a 2’s complement, MSB first date format.
Refer to AN58 “Dual ProSLIC Programmer Guide” for detailed RAM location descriptions and recommended
settings.
Note: Any RAM address not listed is reserved and must not be written. (ordered alphabetically by mnemonic)
RAM
Addr
Mnemonic
Description
Bit
15
Bit
14
Bit
13
Bit
12
Bit
11
Bit
10
Bit
9
Bit
8
Bit
7
Bit
6
Bit
5
Bit
4
Bit
3
Bit
2
Bit
1
Bit
0
Type
Example
Hex
Example
Dec
Unit
Init
0E54
18
V
Init
0A08
10
Hz
Init
0D88
17
V
Init
0A08
10
Hz
Battery Selection and VOC Tracking
31
BATHTH
High Battery
Switch Threshold
34
BATLPF
Battery Tracking
Filter Coeff
32
BATLTH
Low Battery Switch
Threshold
33
BSWLPF
RING Voltage Filter Coeff
BATHTH[14:7]2
BATLPF[15:3]
BATLTH[14:7]2
BSWLPF[15:3]
Speedup
36
CMHITH
Speedup Threshold—
High Byte
CMHITH[15:0]
Init
0001
1
V
35
CMLOTH
Speedup Threshold—
Low Byte
CMLOTH[15:0]
Init
07F5
10
V
SLIC Diagnostics Filter
53
DIAGAC
SLIC Diags AC
Detector Threshold
54
DIAGACCO
SLIC Diags AC Filter Coeff
51
DIAGDC
SLIC Diags DC Output
52
DIAGDCCO
SLIC Diags DC Filter Coeff
55
DIAGPK
SLIC Diags Peak
Detector
DIAGAC[15:0]
DIAGACCO[15:3]
DIAGDC[15:0]
DIAGDCCO[15:3]
DIAGPK[15:0]
Diag
Diag
V
7FF8
127.3
Diag
Diag
Hz
V
0A08
Diag
10
Hz
V
DTMF Detection
118
DTCOL2HTH
DTMF Column Second
Harmonic Threshold
DTCOL2HTH[15:3]
Init
1013
116
DTCOLRTH
DTMF Column
Ratio Threshold
DTCOLRTH[15:3]
Init
0CC5
112
DTCOLTH
DTMF Column
Peak Threshold
DTCOLTH[15:3]
Init
1999
113
DTFTWTH
DTMF Forward
Twist Threshold
DTFTWTH[15:3]
Init
1013
120
DTHOTTH
DTMF Hot Limit
Threshold
DTHOTTH[15:0]
Init
0A1C
119
DTMINPTH
DTMF Minimum
Power Threshold
DTMINPTH[15:0]
Init
00E5
Notes:
1.
2.
3.
4.
5.
6.
7.
Any register not listed is reserved and must not be written.
Only positive input values are valid for these RAM addresses.
Si3225 only.
Si3220 only.
For the Si3220, the RINGFRHI RAM address location is used to store the high byte of the internal ringing signal frequency. For the Si3225, this address location stores the desired time delay
between when the relay opens and when the LFS register transitions out of the ringing state.
For the Si3220, the RINGAMP RAM address location is used to store the amplitude of the internal ringing signal. For the Si3225, this address location stores the desired time relay between the last
zero current crossing and the next opportunity to open the ringing relay.
RAM address in decimal.
94
Preliminary Rev. 0.91
Si3220/Si3225
RAM
Addr
Mnemonic
Description
108
DTROW0TH
DTMF Row 0
Peak Threshold
109
DTROW1TH
117
Bit
15
Bit
14
Bit
13
Bit
12
Bit
11
Bit
10
Bit
9
Bit
8
Bit
7
Bit
6
Bit
5
Bit
4
Bit
3
Bit
2
Bit
1
Bit
0
Type
Example
Hex
DTROW0TH[15:3]
Init
2AE1
DTMF Row 1
Peak Threshold
DTROW1TH[15:3]
Init
28FB
DTROW2HTH
DTMF Row Second
Harmonic Threshold
DTROW2HTH[15:3]
Init
308C
110
DTROW2TH
DTMF Row 2
Peak Threshold
DTROW2TH[15:3]
Init
25C2
111
DTROW3TH
DTMF Row 3
Peak Threshold
DTROW3TH[15:3]
Init
249B
115
DTROWRTH
DTMF Row Ratio
Threshold
DTROWRTH[15:3]
Init
0CC5
114
DTRTWTH
DTMF Reverse
Twist Threshold
DTRTWTH[15:3]
Init
1013
Example
Dec
Unit
Echo Cancellation
89
ECCO0
Echo Cancellation Coeff 0
ECCO0[15:3]
Init
01B0
1728
82
ECCO1
Echo Cancellation Coeff 1
ECCO1[15:3]
Init
FF20
–896
83
ECCO2
Echo Cancellation Coeff 2
ECCO2[15:3]
Init
1548
21792
84
ECCO3
Echo Cancellation Coeff 3
ECCO3[15:3]
Init
1E38
30944
85
ECCO4
Echo Cancellation Coeff 4
ECCO4[15:3]
Init
1238
18656
86
ECCO5
Echo Cancellation Coeff 5
ECCO5[15:3]
Init
01B8
1760
87
ECCO6
Echo Cancellation Coeff 6
ECCO6[15:3]
Init
FD08
–3040
88
ECCO7
Echo Cancellation Coeff 7
ECCO7[15:3]
Init
FFA0
–384
92
ECIIRA1
Echo Cancel IIR
Filter Coeff A1
ECIIRA1[15:3]
Init
0370
3520
93
ECIIRA2
Echo Cancel IIR
Filt Coeff A2
ECIIRA2[15:3]
Init
CB58
–53920
90
ECIIRB0
Echo Cancel IIR
Filt Coeff B0
ECIIRB0[15:3]
Init
0068
416
91
ECIIRB1
Echo Cancel IIR
Filt Coeff B1
ECIIRB1[15:3]
Init
FEA0
–1408
FSK Generation
102
FSKAMP0
FSK Amplitude for Space
FSKAMP0[15:3]
Init
0100
.22
Vrms
103
FSKAMP1
FSK Amplitude for Mark
FSKAMP1[15:3]
Init
01E0
.22
Vrms
100
FSKFREQ0
FSK Frequency for Space
FSKFREQ0[15:3]
Init
3CE0
1200
Hz
101
FSKFREQ1
FSK Frequency for Mark
FSKFREQ1[15:3]
Init
35B0
2200
Hz
104
FSK01HI
FSK 0-1 Transition Freq—
High
FSK01HI[15:3]
Init
3BE0
105
FSK01LO
FSK 0-1 Transition
Frequency—Low
FSK01LO[15:3]
Init
1330
106
FSK10HI
FSK 1-0 Transition
Frequency—High
FSK10HI[15:3]
Init
1118
107
FSK10LO
FSK 1-0 Transition
Frequency—Low
FSK10LO[15:3]
Init
1D88
Notes:
1.
2.
3.
4.
5.
6.
7.
Any register not listed is reserved and must not be written.
Only positive input values are valid for these RAM addresses.
Si3225 only.
Si3220 only.
For the Si3220, the RINGFRHI RAM address location is used to store the high byte of the internal ringing signal frequency. For the Si3225, this address location stores the desired time delay
between when the relay opens and when the LFS register transitions out of the ringing state.
For the Si3220, the RINGAMP RAM address location is used to store the amplitude of the internal ringing signal. For the Si3225, this address location stores the desired time relay between the last
zero current crossing and the next opportunity to open the ringing relay.
RAM address in decimal.
Preliminary Rev. 0.91
95
Si3220/Si3225
RAM
Addr
Mnemonic
Description
Bit
15
Bit
14
Bit
13
Bit
12
Bit
11
Bit
10
Bit
9
Bit
8
Bit
7
Bit
6
Bit
5
Bit
4
Bit
3
Bit
2
Bit
1
Bit
0
Type
Example
Hex
Example
Dec
Unit
Loop Currents
9
ILONG
Longitudinal Current Sense
Value
ILONG[15:0]2
DIag
8
ILOOP
Loop Current Sense Value
ILOOP[15:0]2
Diag
mA
18
IRING
Q5 Current Measurement
IRING[15:0]
Diag
mA
16
IRINGN
Q3 Current Measurement
IRINGN[15:0]
Diag
mA
15
IRINGP
Q2 Current Measurement
IRINGP[15:0]
Diag
mA
Diag
mA
mA
21
IRNGNG
External Ringing Generator Current Measurement
IRNGNG[15:0]3
19
ITIP
Q6 Current Measurement
ITIP[15:0]
Diag
mA
17
ITIPN
Q4 Current Measurement
ITIPN[15:0]
Diag
mA
14
ITIPP
Q1 Current Measurement
ITIPP[15:0]
Diag
mA
24
LCRDBI
Loop Closure Detection
Debounce Interval
25
LCRLPF
Loop Closure Filter
Coefficient
26
LCRMASK
Loop Closure Mask Interval
Coeff
166
LCRMSKPR
22
23
Loop Closure Detection
LCRDBI[15:0]2
Init
000C
15
ms
Init
0A10
10
Hz
LCRMASK[15:0]2
Init
0040
80
ms
LCR Mask During Polarity
Reversal
LCRMSKPR[15:0]
Init
0040
80
ms
LCROFFHK
Off-Hook Detect
Threshold
LCROFFHK[15:0]2
Init
0C0C
10
mA
LCRONHK
On-Hook Detect
Threshold
LCRONHK[15:0]2
Init
0DEO
11
mA
LCRLPF[15:3]
Longitudinal Current Detection
29
LONGDBI
Ground Key Detection
Debounce Interval
LONGDBI[15:0]2
Init
27
LONGHITH
Ground Key Detection
Threshold
LONGHITH[15:0]2
Init
08D4
7
mA
28
LONGLOTH
Ground Key Removal
Detection Threshold
LONGLOTH[15:0]2
Init
0A17
8
mA
30
LONGLPF
Ground Key Filter
Coefficient
Init
0A08
10
Hz
Init
004F
0.0775
Vrms
Init
3D98
350
Hz
LONGLPF[15:3]
ms
Oscillator Coefficients
95
OSC1AMP
Oscillator 1 Amplitude
94
OSC1FREQ
Oscillator 1 Frequency
OSC1AMP[15:0]
96
OSC1PHAS
Oscillator 1 Initial Phase
OSC1PHAS[15:0]
Init
0000
98
OSC2AMP
Oscillator 2 Amplitude
OSC2AMP[15:0]
Init
0063
0.0775
Vrms
97
OSC2FREQ
Oscillator 2 Frequency
Init
3C38
440
Hz
99
OSC2PHAS
Oscillator 2 Initial Phase
Init
0000
OSC1FREQ[15:3]
OSC2FREQ[15:3]
OSC2PHAS[15:0]
Power Calculations
Notes:
1.
2.
3.
4.
5.
6.
7.
Any register not listed is reserved and must not be written.
Only positive input values are valid for these RAM addresses.
Si3225 only.
Si3220 only.
For the Si3220, the RINGFRHI RAM address location is used to store the high byte of the internal ringing signal frequency. For the Si3225, this address location stores the desired time delay
between when the relay opens and when the LFS register transitions out of the ringing state.
For the Si3220, the RINGAMP RAM address location is used to store the amplitude of the internal ringing signal. For the Si3225, this address location stores the desired time relay between the last
zero current crossing and the next opportunity to open the ringing relay.
RAM address in decimal.
96
Preliminary Rev. 0.91
Si3220/Si3225
RAM
Addr
Mnemonic
Description
40
PLPFQ12
Q1/Q2 Thermal Low Pass
Filter Coeff
41
PLFPQ34
42
PLFPQ56
Bit
15
Bit
14
Bit
13
Bit
12
Bit
11
Bit
10
Bit
9
Bit
8
Bit
7
Bit
6
Bit
5
Bit
4
Bit
3
Bit
2
Bit
1
Bit
0
Type
Example
Hex
Example
Dec
Unit
PLFPQ12[15:3]
Init
0008
.3
s
Q3/Q4 Thermal Low Pass
Filter Coeff
PLFPQ34[15:3]
Init
0008
.3
s
Q5/Q6 Thermal Low Pass
Filter Coeff
PLFPQ56[15:3]
Init
0008
.3
s
Pulse Metering
68
PMAMPL
Pulse Metering Amplitude
PMAMPL[15:0]4
Init
4000
65536
V
70
PMAMPTH
Pulse Metering AGC
Amplitude Threshold
PMAMPTH[15:0]4
Init
00C8
798
V
67
PMFREQ
Pulse Metering
Frequency
Init
0000
0
Hz
69
PMRAMP
Pulse Metering Ramp Rate
Init
008A
550
s
PMFREQ[15:3]4
PMRAMP[15:0]4
Power Calculations
44
PQ1DH
Q1 Calculated Power
PQ1DH[15:0]
Diag
W
45
PQ2DH
Q2 Calculated Power
PQ2DH[15:0]
Diag
W
46
PQ3DH
Q3 Calculated Power
PQ3DH[15:0]
Diag
W
47
PQ4DH
Q4 Calculated Power
PQ4DH[15:0]
Diag
W
48
PQ5DH
Q5 Calculated Power
PQ5DH[15:0]
Diag
W
49
PQ6DH
Q6 Calculated Power
PQ6DH[15:0]
Diag
W
50
PSUM
Total Calculated Power
PSUM[15:0]
Diag
37
PTH12
Q1/Q2 Power Threshold
PTH12[15:0]2
Init
0007
.22
W
38
PTH34
Q3/Q4 Power Threshold
PTH34[15:0]2
Init
003C
17
W
2
Init
002A
1.28
W
39
PTH56
Q5/Q6 Power Threshold
43
RB56
Q5/Q6 Base Resistor
PTH56[15:0]
RB56[15:0]
W
Ω
Init
Ringing
59
RINGAMP
Ringing Amplitude/Zero
Crossing Delay
57
RINGFRHI
Ringing Frequency—High
Byte/Linefeed Status Delay
58
RINGFRLO
Ringing Frequency—
Low Byte
56
RINGOF
Ringing Waveform DC
Offset
60
RINGPHAS
Ringing Oscillator
Initial Phase
RINGAMP[15:0]6/ZERDELAY[15:0]
Init
00D5
47
Vrms
RINGFRHI[14:3]5/LFSDELAY[14:3]
Init
3F78
20
Hz
RINGFRLO[14:3]4
Init
6CE8
20
Hz
Init
0000
0
V
Init
0000
10
ms
RINGOF[15:0]4
RINGPHAS[15:3]4
Ring Trip Detection
66
RTACDB
AC Ring Trip
Debounce Interval
RTACDB[15:0]
Init
0008
64
RTACTH
AC Ring Trip
Detect Threshold
RTACTH[15:0]
Init
1086
61
RTCOUNT
Ring Trip
Timeout Counter
RTCOUNT[15:0]
Init
0400
Notes:
1.
2.
3.
4.
5.
6.
7.
mA
128
ms
Any register not listed is reserved and must not be written.
Only positive input values are valid for these RAM addresses.
Si3225 only.
Si3220 only.
For the Si3220, the RINGFRHI RAM address location is used to store the high byte of the internal ringing signal frequency. For the Si3225, this address location stores the desired time delay
between when the relay opens and when the LFS register transitions out of the ringing state.
For the Si3220, the RINGAMP RAM address location is used to store the amplitude of the internal ringing signal. For the Si3225, this address location stores the desired time relay between the last
zero current crossing and the next opportunity to open the ringing relay.
RAM address in decimal.
Preliminary Rev. 0.91
97
Si3220/Si3225
RAM
Addr
Mnemonic
Description
65
RTDCDB
DC Ring Trip
Debounce Interval
62
RTDCTH
63
RTPER
Bit
15
Bit
14
Bit
13
Bit
12
Bit
11
Bit
10
Bit
9
Bit
8
Bit
7
Bit
6
Bit
5
Bit
4
Bit
3
Bit
2
Bit
1
Bit
0
Type
Example
Hex
Example
Dec
Unit
RTDCDB[15:0]
Init
0008
10
ms
DC Ring Trip
Detect Threshold
RTDCTH[15:0]
Init
7FFF
Ring Trip Low Pass Filter
Coeff Period
RTPER[15:0]
Init
0028
20
mA
Hz
Receive Path Gain and Filters
81
RXIIRPOL
RX IIR Filter Pole Coeff
RXIIRPOL[15:3]
Init
3CCC
62256
80
RXEQCO0
RX Equalizer
Coeff 0
RXEQCO0[15:3]
Init
4000
65536
79
RXEQCO1
RX Equalizer
Coeff 1
RXEQCO1[15:3]
Init
0000
0
78
RXEQCO2
RX Equalizer
Coeff 2
RXEQCO2[15:3]
Init
0000
0
77
RXEQCO3
RX Equalizer
Coeff 3
RXEQCO3[15:3]
Init
0000
0
4000
1
0000
60
71
RXGAIN
RX Gain Setting
RXGAIN[15:3]
Init
123
RXMODPWR
RX Path Modem
Tone Power
RXMODPWR[15:3]
Init
121
RXPWR
RX Path Input
Signal Power
RXPWR[15:0]
Init
DC Speedup
168
SPEEDUP
DC Speedup Timer
SPEEDUP[15:0]
Init
132
TESTA1H1
TX Diag Filter Coeff A1H1
TESTA1H1[15:3]
Diag
142
TESTA1H2
TX Diag Filter Coeff A1H2
TESTA1H2[15:3]
Diag
152
TESTA1H3
TX Diag Filter Coeff A1H3
TESTA1H3[15:3]
Diag
131
TESTA1L1
TX Diag Filter Coeff A1L1
TESTA1L1[15:3]
Diag
141
TESTA1L2
TX Diag Filter Coeff A1L2
TESTA1L2[15:3]
Diag
151
TESTA1L3
TX Diag Filter Coeff A1L3
TESTA1L3[15:3]
Diag
134
TESTA2H1
TX Diag Filter Coeff A2H1
TESTA2H1[15:3]
Diag
144
TESTA2H2
TX Diag Filter Coeff A2H2
TESTA2H2[15:3]
Diag
154
TESTA2H3
TX Diag Filter Coeff A2H3
TESTA2H3[15:3]
Diag
133
TESTA2L1
TX Diag Filter Coeff A2L1
TESTA2L1[15:3]
Diag
143
TESTA2L2
TX Diag Filter Coeff A2L2
TESTA2L2[15:3]
Diag
153
TESTA2L3
TX Diag Filter Coeff A2L3
TESTA2L3[15:3]
156
TESTAVO
TX Diag Filter
Avg Output
158
TESTAVBW
TX Diag Filter
Avg Bandwidth
TESTAVBW[15:3]
Diag
160
TESTAVFL
TX Diag Filter
Average Flag
TESTAVFL[15:3]
Diag
ms
Test Diagnostic Filters
Notes:
1.
2.
3.
4.
5.
6.
7.
TESTAVO[15:0]
Diag
Diag
V
Any register not listed is reserved and must not be written.
Only positive input values are valid for these RAM addresses.
Si3225 only.
Si3220 only.
For the Si3220, the RINGFRHI RAM address location is used to store the high byte of the internal ringing signal frequency. For the Si3225, this address location stores the desired time delay
between when the relay opens and when the LFS register transitions out of the ringing state.
For the Si3220, the RINGAMP RAM address location is used to store the amplitude of the internal ringing signal. For the Si3225, this address location stores the desired time relay between the last
zero current crossing and the next opportunity to open the ringing relay.
RAM address in decimal.
98
Preliminary Rev. 0.91
Si3220/Si3225
RAM
Addr
Mnemonic
Description
Bit
15
Bit
14
Bit
13
Bit
12
Bit
11
Bit
10
Bit
9
Bit
8
Bit
7
Bit
6
Bit
5
Bit
4
Bit
3
Bit
2
Bit
1
Bit
0
162
TESTAVTH
TX Diag Filter
Avg Threshold
TESTAVTH[15:3]
Diag
126
TESTB0H1
TX Diag Filter Coeff B0H1
TESTB0H1[15:3]
Diag
136
TESTB0H2
TX Diag Filter Coeff B0H2
TESTB1H2[15:3]
Diag
146
TESTB0H3
TX Diag Filter Coeff B0H3
TESTB0H3[15:3]
Diag
125
TESTB0L1
TX Diag Filter Coeff B0L1
TESTB0L1[15:3]
Diag
135
TESTB0L2
TX Diag Filter Coeff B0L2
TESTB0L2[15:3]
Diag
145
TESTB0L3
TX Diag Filter Coeff B0L3
TESTB0L3[15:3]
Diag
128
TESTB1H1
TX Diag Filter Coeff B1H1
TESTB1H1[15:3]
Diag
138
TESTB1H2
TX Diag Filter Coeff B1H2
TESTB1H2[15:3]
Diag
148
TESTB1H3
TX Diag Filter Coeff B1H3
TESTB1H3[15:3]
Diag
127
TESTB1L1
TX Diag Filter Coeff B1L1
TESTB1L1[15:3]
Diag
137
TESTB1L2
TX Diag Filter Coeff B1L2
TESTB1L2[15:3]
Diag
147
TESTB1L3
TX Diag Filter Coeff B1L3
TESTB1L3[15:3]
Diag
130
TESTB2H1
TX Diag Filter Coeff B2H1
TESTB2H1[15:3]
Diag
140
TESTB2H2
TX Diag Filter Coeff B2H2
TESTB2H2[15:3]
Diag
150
TESTB2H3
TX Diag Filter Coeff B2H3
TESTB2H3[15:3]
Diag
129
TESTB2L1
TX Diag Filter Coeff B2L1
TESTB2L1[15:3]
Diag
139
TESTB2L2
TX Diag Filter Coeff B2L2
TESTB2L2[15:3]
Diag
149
TESTB2L3
TX Diag Filter Coeff B2L3
TESTB2L3[15:3]
Diag
159
TESTPKFL
TX Diag Filter
Peak Flag
TESTPKFL[15:3]
Diag
155
TESTPKO
TX Diag Filter
Peak Output
TESTPKO[15:3]
Diag
161
TESTPKTH
TX Diag Filter
Peak Threshold
TESTPKTH[15:3]
Diag
157
TESTWLN
TX Diag Filter
TESTWLN[15:3]
Diag
Type
Example
Hex
Example
Dec
Unit
V
Transmit Path Gain and Filters
76
TXEQCO0
TX Equalizer
Coefficient 0
TXEQCO0[15:3]
Init
4A6A
76201
75
TXEQCO1
TX Equalizer
Coefficient 1
TXEQCO1[15:3]
Init
F84C
–7888
74
TXEQCO2
TX Equalizer
Coefficient 2
TXEQCO2[15:3]
Init
012C
1199
73
TXEQCO3
TX Equalizer
Coefficient 3
TXEQCO3[15:3]
Init
004C
302
4000
1
72
TXGAIN
TX Gain Setting
TXGAIN[15:3]
Init
163
TXHPF1
TX HPF Coefficient 1
TXHPF1[15:3]
Diag
164
TXHPF2
TX HPF Coefficient 2
TXHPF2[15:3]
Diag
165
TXHPF3
TX HPF Coefficient 3
TXHPF3[15:3]
Diag
Notes:
1.
2.
3.
4.
5.
6.
7.
Any register not listed is reserved and must not be written.
Only positive input values are valid for these RAM addresses.
Si3225 only.
Si3220 only.
For the Si3220, the RINGFRHI RAM address location is used to store the high byte of the internal ringing signal frequency. For the Si3225, this address location stores the desired time delay
between when the relay opens and when the LFS register transitions out of the ringing state.
For the Si3220, the RINGAMP RAM address location is used to store the amplitude of the internal ringing signal. For the Si3225, this address location stores the desired time relay between the last
zero current crossing and the next opportunity to open the ringing relay.
RAM address in decimal.
Preliminary Rev. 0.91
99
Si3220/Si3225
RAM
Addr
Mnemonic
Description
124
TXMODPWR
TX Path Modem
Tone Power
122
TXPWR
TX Path Input
Signal Power
Bit
15
Bit
14
Bit
13
Bit
12
Bit
11
Bit
10
Bit
9
Bit
8
Bit
7
Bit
6
TXMODPWR[15:3]
TXPWR[15:0]
Bit
5
Bit
4
Bit
3
Bit
2
Bit
1
Bit
0
Type
Example
Hex
Example
Dec
Unit
Init
Init
Loop Voltages
13
VBAT
Scaled Battery Voltage
Measurement
4
VCM
Common Mode Voltage
7
VLOOP
Loop Voltage
0
VOC
Open Circuit Voltage
VOC[14:0]2
Init
2668
48
V
VOC Delta for Off-Hook
VOCDELTA[14:0]2
1
VOCDELTA
VBAT[15:0]
VCM[14:0]2
VLOOP[15:0]2
Diag
Init
V
0268
3
Diag
V
V
Init
059A
7
V
Init
0198
2
V
F9A2
–8
V
3
VOCHTH
VOC Delta Upper
Threshold
VOCHTH[15:0]2
2
VOCLTH
VOC Delta Lower
Threshold
VOCLTH[15:0]
Init
10
VOCTRACK
Battery Tracking Open
Circuit Voltage
VOCTRACK[15:0]2
Diag
5
VOV
Overhead Voltage
VOV[14:0]2
Init
0334
4
V
VOVRING[14:0]2
Init
0000
0
V
6
VOVRING
Ringing Overhead
Voltage
12
VRING
Scaled RING Voltage
Measurement
20
VRNGNG
External Ringing Generator Voltage Measurement
11
VTIP
Scaled TIP Voltage
Measurement
Notes:
1.
2.
3.
4.
5.
6.
7.
VRING[15:0]
VRNGNG[14:7]3
VTIP[15:0]
V
Diag
V
Diag
V
Diag
V
Any register not listed is reserved and must not be written.
Only positive input values are valid for these RAM addresses.
Si3225 only.
Si3220 only.
For the Si3220, the RINGFRHI RAM address location is used to store the high byte of the internal ringing signal frequency. For the Si3225, this address location stores the desired time delay
between when the relay opens and when the LFS register transitions out of the ringing state.
For the Si3220, the RINGAMP RAM address location is used to store the amplitude of the internal ringing signal. For the Si3225, this address location stores the desired time relay between the last
zero current crossing and the next opportunity to open the ringing relay.
RAM address in decimal.
100
Preliminary Rev. 0.91
Si3220/Si3225
64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49
BATSELa
TRD2a
TRD1a
RTRPa
BLKRNG
THERMa
IRINGPa
GND1
VDD1
ITIPPa
IRINGNa
ITIPNa
SRINGDCa
SRINGACa
STIPACa
STIPDCa
BATSELa
TRD2a
TRD1a
NC
NC
THERMa
IRINGPa
GND1
VDD1
ITIPPa
IRINGNa
ITIPNa
SRINGDCa
SRINGACa
STIPACa
STIPDCa
Pin Descriptions: Si3220/25
64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49
SVBATa
1
48
GPOa
SVBATa
1
RPOa
2
47
CS
48
RRDa
RPOa
2
47
RPIa
3
46
CS
SDITHRU
RPIa
3
46
RNIa
4
SDITHRU
45
SDI
RNIa
4
45
RNOa
SDI
5
44
SDO
RNOa
5
44
SDO
CAPPa
6
43
SCLK
CAPPa
6
43
SCLK
CAPMa
7
42
VDD4
CAPMa
7
42
VDD4
QGND
8
41
GND4
QGND
8
41
GND4
IREF
9
40
INT
IREF
9
40
INT
CAPMb
10
39
PCLK
CAPMb
10
39
PCLK
CAPPb
11
38
GND3
CAPPb
11
38
GND3
RNOb
12
37
VDD3
RNOb
12
37
VDD3
RNIb
13
36
DTX
RNIb
13
36
DTX
RPIb
14
35
DRX
RPIb
14
35
DRX
RPOb
15
34
FSYNC
RPOb
15
34
FSYNC
SVBATb
16
33
RESET
SVBATb
16
33
RESET
Pin Number(s)
Symbol
Input/
Output
BATSELb
RRDb
TRD2b
TRD1b
RTRPb
THERMb
IRINGPb
GND2
VDD2
ITIPPb
IRINGNb
ITIPNb
SRINGDCb
SRINGACb
STIPDCb
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
BATSELb
GPOb
TRD2b
TRD1b
NC
THERMb
IRINGPb
GND2
VDD2
ITIPPb
IRINGNb
ITIPNb
SRINGDCb
SRINGACb
STIPACb
STIPDCb
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
Si3225
64-Lead TQFP
(epad)
STIPACb
Si3220
64-Lead TQFP
(epad)
Description
Si3220
Si3225
1, 16
1, 16
SVBATa,
SVBATb
I
Battery Sensing Input.
Analog current input used to sense battery voltage.
2,15
2,15
RPOa, RPOb
O
Transconductance Amplifier External Resistor Connection.
3, 14
3, 14
RPIa,
RPIb
I
Transconductance Amplifier External Resistor Connection.
4, 13
4, 13
RNIa,
RNIb
I
Transconductance Amplifier Resistor Connection.
5, 12
5, 12
RNOa, RNOb
O
Transconductance Amplifier Resistor Connection.
6, 11
6, 11
CAPPa,
CAPPb
Differential Capacitor.
Capacitor used in low pass filter to stabilize SLIC feedback
loops.
7, 10
7, 10
CAPMa,
CAPMb
Common Mode Capacitor.
Capacitor used in low pass filter to stabilize SLIC feedback
loops.
8
8
QGND
Component Reference Ground.
Return path for differential and common mode capacitors. Do not
connect to system ground.
Preliminary Rev. 0.91
101
Si3220/Si3225
Pin Number(s)
Symbol
Input/
Output
Description
Si3220
Si3225
9
9
IREF
I
IREF Current Reference.
Connects to an external resistor to provide a high accuracy reference current. Return path for IREF resistor should be routed to
QGND pin.
17, 64
17, 64
STIPDCb,
STIPDCa
I
TIP Sense.
Analog current input senses dc voltage on TIP side of subscriber
loop.
18, 63
18, 63
STIPACb,
STIPACa
I
TIP Transmit Input.
Analog input senses ac voltage on TIP side of subscriber loop.
19, 62
19, 62
SRINGACb,
SRINGACa
I
RING Transmit Input.
Analog input senses ac voltage on RING side of subscriber loop.
20, 61
20, 61
SRINGDCb,
SRINGDCa
I
RING Sense.
Analog current input senses dc voltage on RING side of subscriber loop.
21, 60
21, 60
ITIPNb,
ITIPNa
O
Negative TIP Current Control.
Analog current output provides dc current return path to VBAT
from TIP side of the loop.
22, 59
22, 59
IRINGNb,
IRINGNa
O
Negative RING Current Control.
Analog current output provides dc current return path to VBAT
from RING side of loop.
23, 58
23, 58
ITIPPb,
ITIPPa
O
Positive TIP Current Control.
Analog current output drives dc current onto TIP side of subscriber loop in normal polarity. Also modulates ac current onto
TIP side of loop.
24, 37,
42, 57
24, 37,
42, 57
VDD2,VDD3,
VDD4,VDD1
Supply Voltage.
Power supply for internal analog and digital circuitry. Connect all
VDD pins to the same supply and decouple to adjacent GND pin
as close to the pins as possible.
25, 38,
41, 56
25, 38,
41, 56
GND2,GND3,
GND4,GND1
Ground.
Ground connection for internal analog and digital circuitry. Connect all pins to low-impedance ground plane.
26, 55
26, 55
IRINGPb,
IRINGPa
O
Positive RING Current Control.
Analog current output drives dc current onto RING side of subscriber loop in reverse polarity. Also modulates ac current onto
RING side of loop.
27,54
27,54
THERMb,
THERMa
I
Temperature Sensor.
Senses Internal temperature of Si3200.
29, 51
29, 51
TRD1b,
TRD1a
O
Test Relay Driver Output.
Drives test relays for connecting loop test equipment.
28, 52,
53
102
NC
No Internal Connection.
Leave unconnected or connect to ground plane.
Preliminary Rev. 0.91
Si3220/Si3225
Pin Number(s)
Si3220
30, 50
Symbol
Input/
Output
28, 52
RTRPb,
RTRPa
I
External Ring Trip Sensing Input.
Used to sense ring-trip condition when using centralized ring
generator. Connect to low side of ring sense resistor.
30, 50
TRD2b,
TRD2a
O
Test Relay Driver Output.
Drives test relays for connecting loop test equipment.
31, 48
RRDb, RRDa
O
Ring Relay Driver Output.
Connects an external centralized ring generator to the subscriber
loop.
GPOb, GPOa
O
General Purpose Output Driver.
Used as a relay driver or as a second battery select pin when
using a third battery supply.
Si3225
31, 48
Description
32, 49
32, 49
BATSELb,
BATSELa
O
Battery Voltage Select Pin.
Switches between high and low external battery supplies.
35
35
DRX
I
Receive PCM Data.
Input data from PCM/GCI bus.
36
36
DTX
O
Transmit PCM Data.
Output data to PCM/GCI bus.
39
39
PCLK
I
PCM Bus Clock.
Clock input for PCM/GCI bus timing.
33
33
RESET
I
Reset.
Active low. Hardware reset used to place all control registers in
known state. An internal pulldown resistor asserts this pin low
when not driven externally.
34
34
FSYNC
I
Frame Sync.
8 kHz frame synchronization signal for PCM/GCI bus. May be
short or long pulse format.
40
40
INT
O
Interrupt.
Maskable interrupt output. Open drain output for wire-ORed
operation.
43
43
SCLK
I
Serial Port Bit Clock Input.
Controls serial data on SDO and latches data on SDI.
44
44
SDO
O
Serial Port Data Out.
Serial port control data output.
45
45
SDI
I
Serial Port Data In.
Serial port control data input.
46
46
SDITHRU
O
Serial Data Daisy Chain.
Enables multiple devices to use a single CS for serial port control. Connect SDITHRU pin from master device to SDI pin of
slave device. An internal pullup resistor holds this pin high during
idle periods.
Preliminary Rev. 0.91
103
Si3220/Si3225
Pin Number(s)
Symbol
Input/
Output
Description
Si3220
Si3225
47
47
CS
I
Chip Select.
Active low. When inactive, SCLK and SDI are ignored and SDO
is high impedance. When active, serial port is operational.
53
BLKRNG
I
Ring Generator Sensing Input.
Senses ring-trip condition when using centralized ring generator.
Connect to high side of ring sense resistor. Shared by channel a
and b.
epad
GND
epad
104
Exposed Die Paddle Ground.
Connect to a low-impedance ground plane via top side PCB pad
directly under the part. See Package Outlines: 64-Pin TQFP for
PCB pad dimensions.
Preliminary Rev. 0.91
Si3220/Si3225
Pin Descriptions: Si3200
Si3200
16-Lead SOIC
(epad)
TIP
1
16
ITIPP
NC
2
15
RING
3
14
ITIPN
THERM
VBAT
4
13
IRINGP
VBATH
5
12
IRINGN
VBATL
6
11
GND
7
10
NC
NC
VDD
8
9
BATSEL
Pin #(s)
Symbol
Input/
Output
Description
1
TIP
I/O
TIP Output.
Connect to the TIP lead of the subscriber loop.
2, 10, 11
NC
—
No Internal Connection.
Do not connect to any electrical signal.
3
RING
I/O
RING Output.
Connect to the RING lead of the subscriber loop.
4
VBAT
—
Operating Battery Voltage.
Si3200 internal system battery supply. Connect SVBATa/b pin from Si3220/
25 and decouple with a 0.1 µF/100 V filter capacitor.
5
VBATH
—
High Battery Voltage.
Connect to the system ringing battery supply. Decouple with a 0.1 µF/100 V
filter capacitor.
6
VBATL
—
Low Battery Voltage.
Connect to lowest system battery supply for off-hook operation driving short
loops. An internal diode prevents leakage current when operating from
VBATH.
7
GND
—
Ground.
Connect to a low-impedance ground plane.
8
VDD
—
Supply Voltage.
Main power supply for all internal circuitry. Connect to a 3.3 V or 5 V supply.
Decouple locally with a 0.1 µF/10 V capacitor.
9
BATSEL
I
Battery Voltage Select.
Connect to the BATSEL pin of the Si3220 or Si3225 through an external
resistor to enable automatic battery switching. No connection is required
when used with the Si3225 in a single battery system configuration.
Preliminary Rev. 0.91
105
Si3220/Si3225
Pin #(s)
Symbol
Input/
Output
12
IRINGN
I
Negative RING Current Control.
Connect to the IRINGN lead of the Si3220 or Si3225.
13
IRINGP
I
Positive RING Current Drive.
Connect to the IRINGP lead of the Si3220 or Si3225.
14
THERM
O
Thermal Sensor.
Connection to internal temperature sensing circuit.
Connect to THERM pin of Si3220 or Si3225.
15
ITIPN
I
Negative TIP Current Control.
Connect to the ITIPN lead of the Si3220 or Si3225.
16
ITIPP
I
Positive TIP Current Control.
Connect to the ITIPP lead of the Si3220 or Si3225.
epad
GND
106
Description
Exposed Die Paddle Ground.
For adequate thermal management, the exposed die paddle should be soldered to a PCB pad that is connected to low-impedance inner and/or backside ground planes using multiple vias. See “Package Outline: 16-Pin
SOIC” for PCB pad dimensions.
Preliminary Rev. 0.91
Si3220/Si3225
Dual ProSLIC Selection Guide
Part
Number
Description
On-Chip
Ringing
External
Ringing
Support
Pulse
Metering
Temp
Range
Package
Si3200-KS
Linefeed interface
0 to 70 °C
SOIC-16
Si3200-BS
Linefeed interface
–40 to 85 °C
SOIC-16
Si3220-KQ
Dual ProSLIC
"
"
0 to 70 °C
TQFP-64
Si3220-BQ
Dual ProSLIC
"
"
–40 to 85 °C
TQFP-64
Si3225-KQ
Dual ProSLIC
"
0 to 70 °C
TQFP-64
Si3225-BQ
Dual ProSLIC
"
–40 to 85 °C
TQFP-64
Preliminary Rev. 0.91
107
Si3220/Si3225
Package Outline: 64-Pin TQFP
Figure 65 illustrates the package details for the Dual ProSLIC. Table 48 lists the values for the dimensions shown
in the illustration.
64
0° Min.
49
1
0.08/0.20 R
48
A2
0-7°
E1
E
0.08
R. Min.
A1
e
0.20 Min.
33
16
17
Detail A
32
D1
L
1.00 REF
Exposed Pad
6 x 6 mm
b
D
with lead finish
See Detail A
A
0.09/0.20
0.09/0.16
base metal
b1
See Detail B
Detail B
Figure 65. 64-Pin Thin Quad Flat Package (TQFP)
Table 48. 64-Pin Package Diagram Dimensions
Symbol
Millimeters
Min
Nom
Max
A
—
—
1.20
A1
0.05
—
0.15
A2
0.95
1.00
1.05
D
12.00 BSC
D1
10.00 BSC
E
12.00 BSC
E1
10.00 BSC
L
0.45
e
108
0.60
0.75
0.50 BSC
b
0.17
0.22
0.27
b1
0.17
0.20
0.23
Preliminary Rev. 0.91
GAUGE PLANE
Si3220/Si3225
Package Outline: 16-Pin SOIC
Figure 66 illustrates the package details for the Si3200. Table 49 lists the values for the dimensions shown in the
illustration.
16
9
h
E
H
0.010
θ
1
8
B
GAUGE PLANE
L
Exposed Pad
2.3 x 3.6 mm
Detail F
D
A2
e
C
A
L1
A1
See Detail F
Seating Plane
γ
Figure 66. 16-Pin Small Outline Integrated Circuit (SOIC) Package
Table 49. Package Diagram Dimensions
Symbol
A
A1
A2
B
C
D
E
e
H
h
L
L1
γ
θ
Millimeters
Min
Max
1.35
1.75
.10
.25
1.30
1.50
.33
.51
.19
.25
9.80
10.01
3.80
4.00
1.27 BSC
5.80
6.20
.25
.50
.40
1.27
1.07 BSC
—
0.10
0º
8º
Preliminary Rev. 0.91
109
Si3220/Si3225
Document Change List
Revision 0.9 to Revision 0.91
!
Table 8 on page 12
"
"
"
!
"Calculating Overhead Voltages" on page 27
"
!
TIP/RING Pulldown Transistor Saturation Voltage
updated.
TIP/RING Pullup Transistor Saturation Voltage updated.
Note added.
Second paragraph updated.
"Internal Trapezoidal Ringing" on page 42
"
110
RINGAMP equation updated.
Preliminary Rev. 0.91
Si3220/Si3225
Notes:
Preliminary Rev. 0.91
111
Si3220/Si3225
Contact Information
Silicon Laboratories Inc.
4635 Boston Lane
Austin, TX 78735
Tel: 1+(512) 416-8500
Fax: 1+(512) 416-9669
Toll Free: 1+(877) 444-3032
Email: productinfo@silabs.com
Internet: www.silabs.com
The information in this document is believed to be accurate in all respects at the time of publication but is subject to change without notice.
Silicon Laboratories assumes no responsibility for errors and omissions, and disclaims responsibility for any consequences resulting from
the use of information included herein. Additionally, Silicon Laboratories assumes no responsibility for the functioning of undescribed features
or parameters. Silicon Laboratories reserves the right to make changes without further notice. Silicon Laboratories makes no warranty, representation or guarantee regarding the suitability of its products for any particular purpose, nor does Silicon Laboratories assume any liability
arising out of the application or use of any product or circuit, and specifically disclaims any and all liability, including without limitation consequential or incidental damages. Silicon Laboratories products are not designed, intended, or authorized for use in applications intended to
support or sustain life, or for any other application in which the failure of the Silicon Laboratories product could create a situation where personal injury or death may occur. Should Buyer purchase or use Silicon Laboratories products for any such unintended or unauthorized application, Buyer shall indemnify and hold Silicon Laboratories harmless against all claims and damages.
Silicon Laboratories, Silicon Labs, ProSLIC, and Modfeed are trademarks of Silicon Laboratories Inc.
Other products or brandnames mentioned herein are trademarks or registered trademarks of their respective holders.
112
Preliminary Rev. 0.91