SILABS SI3050-KT

Si3050 +
Si 3 0 1 8/19
G L O B A L VO I C E D A A
Features












PCM highway data interface
µ-law/A-law companding
SPI control interface
GCI interface
80 dB dynamic range TX/RX
Line voltage monitor
Loop current monitor
+6 dBm or +3.2 dBm TX/RX level
mode
Parallel handset detection
3 µA on-hook line monitor current
Overload detection
Programmable line interface
AC termination
DC termination
Ring detect threshold
Ringer impedance














TIP/RING polarity detection
Integrated codec and 2- to 4-wire
analog hybrid
Programmable digital hybrid for
near-end echo reduction
Polarity reversal detection
Programmable digital gain in 0.1 dB
increments
Integrated ring detector
Type I and II caller ID support
Pulse dialing support
3.3 V power supply
Daisy-chaining for up to 16 devices
Greater than 5000 V isolation
Patented isolation technology
Ground start and loop start support
Available in Pb-free RoHS-compliant
packages
Ordering Information
See page 102.
Applications


DSL IADs
VoIP gateways


PBX and IP-PBX systems
Voice mail systems
Pin Assignments
Si3050
Description
The Si3050+Si3018/19 Voice DAA chipset provides a highly-programmable and
globally-compliant foreign exchange office (FXO) analog interface that is ideal for
DSL IADs, PBXs, IP-PBXs, and VoIP gateway products. The solution implements
Silicon Laboratories' patented isolation capacitor technology, which eliminates the
need for costly isolation transformers, relays, or opto-isolators, while providing
superior surge immunity for robust field performance. The Voice DAA is available
in one 20-pin TSSOP (Si3050) and one 16-pin TSSOP/SOIC (Si3018/19) and
requires minimal external components. The Si3050 interfaces directly to standard
telephony PCM interfaces.
Functional Block Diagram
Si3050
CS
SCLK
SDI
SDO
SDI THRU
DTX
Control
Data
Interface
DRX
Line
Data
Interface
FSYNC
RGDT
RG
TGD
TGDE
RESET
AOUT/INT
Rev. 1.31 5/09
Hybrid, AC
and DC
Terminations
Isolation
Interface
Isolation
Interface
Ring Detect
Control
Logic
CS
FSYNC
PCLK
DTX
DRX
RGDT
AOUT/INT
RG
1
2
3
4
5
6
7
8
9
Off-Hook
20
19
18
17
16
15
14
13
12
11
10
SDITHRU
SCLK
GND
VDD
VA
C1A
C2A
RESET
TGDE
TGD
Si3018/19
Si3018/19
RX
PCLK
SDO
SDI
IB
SC
DCT
VREG
VREG2
DCT2
DCT3
RNG1
RNG2
QB
QE
QE2
Copyright © 2009 by Silicon Laboratories
QE
DCT
RX
IB
C1B
C2B
VREG
RNG1
1
2
3
4
5
6
7
8
16
15
14
13
12
11
10
9
DCT2
IGND
DCT3
QB
QE2
SC
VREG2
RNG2
US Patent# 5,870,046
US Patent# 6,061,009
Other Patents Pending
Si3050 + Si3018/19
Si3050 + Si3018/19
2
Rev. 1.31
Si3050 + Si3018/19
TABLE O F C ONTENTS
Section
Page
1. Electrical Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
2. Typical Application Schematic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3. Bill of Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4. AOUT PWM Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5. Functional Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.1. Line-Side Device Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.2. Power Supplies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.3. Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
5.4. Isolation Barrier . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.5. Power Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
5.6. Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
5.7. In-Circuit Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
5.8. Exception Handling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.9. Revision Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.10. Transmit/Receive Full-Scale Level . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.11. Parallel Handset Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.12. Line Voltage/Loop Current Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
5.13. Off-Hook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.14. Ground Start Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
5.15. Interrupts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5.16. DC Termination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.17. AC Termination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
5.18. Ring Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5.19. Ring Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32
5.20. Ringer Impedance and Threshold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.21. Pulse Dialing and Spark Quenching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.22. Receive Overload Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33
5.23. Billing Tone Filter (Optional) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.24. On-Hook Line Monitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.25. Caller ID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5.26. Overload Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36
5.27. Gain Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36
5.28. Transhybrid Balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.29. Filter Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37
5.30. Clock Generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.31. Communication Interface Mode Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .38
5.32. PCM Highway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.33. Companding in PCM Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5.34. 16 kHz Sampling Operation in PCM Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
5.35. SPI Control Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
5.36. GCI Highway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
Rev. 1.31
3
Si3050 + Si3018/19
5.37. Companding in GCI Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.38. 16 kHz Sampling Operation in GCI Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.39. Monitor Channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
5.40. Summary of Monitor Channel Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56
5.41. Device Address Byte . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
5.42. Command Byte . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
5.43. Register Address Byte . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
5.44. SC Channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
5.45. Receive SC Channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
5.46. Transmit SC Channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
6. Control Registers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59
Appendix—UL1950 3rd Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
7. Pin Descriptions: Si3050 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98
8. Pin Descriptions: Si3018/19 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
9. Ordering Guide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
10. Product Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102
11. Package Outline: 20-Pin TSSOP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
12. Package Outline: 16-Pin SOIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
13. Package Outline: 16-Pin TSSOP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Silicon Labs Si3050 Support Documentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
Document Change List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Contact Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .112
4
Rev. 1.31
Si3050 + Si3018/19
1. Electrical Specifications
Table 1. Recommended Operating Conditions and Thermal Information
Parameter1
Symbol
Ambient Temperature
TA
Si3050 Supply Voltage, Digital
VD
Thermal Resistance (Si3018/19)3
JA
Thermal Resistance (Si3050)3
JA
Test Condition
Min2
Typ
Max2
F/K-Grade
0
25
70
B/G-Grade
–40
25
85
3.0
3.3
3.6
V
SOIC-16
—
77
—
°C/W
TSSOP-16
—
89
—
°C/W
TSSOP-20
—
84
—
°C/W
Unit
°C
Notes:
1. The Si3050 specifications are guaranteed when the typical application circuit (including component tolerance) and any
Si3050 and any Si3018/19 are used. See "2. Typical Application Schematic" on page 17 for the typical application
circuit.
2. All minimum and maximum specifications are guaranteed and apply across the recommended operating conditions.
Typical values apply at nominal supply voltages and an operating temperature of 25 °C unless otherwise stated.
3. Operation above 125 °C junction temperature may degrade device reliability.
Rev. 1.31
5
Si3050 + Si3018/19
Table 2. Loop Characteristics
(VD = 3.0 to 3.6 V, TA = 0 to 70 °C for K-Grade, see Figure 1 on page 6)
Parameter
Symbol
Test Condition
Min
Typ
Max
Unit
DC Termination Voltage
VTR
IL = 20 mA, ILIM = 0
DCV = 00, MINI = 11, DCR = 0
—
—
6.0
V
DC Termination Voltage
VTR
IL = 120 mA, ILIM = 0
DCV = 00, MINI = 11, DCR = 0
9
—
—
V
DC Termination Voltage
VTR
IL = 20 mA, ILIM = 0
DCV = 11, MINI = 00, DCR = 0
—
—
7.5
V
DC Termination Voltage
VTR
IL = 120 mA, ILIM = 0
DCV = 11, MINI = 00, DCR = 0
9
—
—
V
DC Termination Voltage
VTR
IL = 20 mA, ILIM = 1
DCV = 11, MINI = 00, DCR = 0
—
—
7.5
V
DC Termination Voltage
VTR
IL = 60 mA, ILIM = 1
DCV = 11, MINI = 00, DCR = 0
40
—
—
V
DC Termination Voltage
VTR
IL = 50 mA, ILIM = 1
DCV = 11, MINI = 00, DCR = 0
—
—
40
V
On-Hook Leakage Current
ILK
VTR = –48 V
—
—
5
µA
Operating Loop Current
ILP
MINI = 00, ILIM = 0
10
—
120
mA
Operating Loop Current
ILP
MINI = 00, ILIM = 1
10
—
60
mA
dc current flowing through ring
detection circuitry
—
1.5
3
µA
VRD
RT2 = 0, RT = 0
13.5
15
16.5
Vrms
VRD
RT2 = 0, RT = 1
19.35
21.5
23.65
Vrms
VRD
RT2 = 1, RT = 1
40.5
45
49.5
Vrms
Hz
DC Ring Current
Ring Detect Voltage*
Ring Detect Voltage
*
Ring Detect Voltage
*
Ring Frequency
Ringer Equivalence Number
FR
13
—
68
REN
—
—
0.2
*Note: The ring signal is guaranteed to not be detected below the minimum. The ring signal is guaranteed to be detected
above the maximum.
TIP
+
600 
IL
Si3018/19 VTR
10F
RING
–
Figure 1. Test Circuit for Loop Characteristics
6
Rev. 1.31
Si3050 + Si3018/19
Table 3. DC Characteristics, VD = 3.3 V
(VD = 3.0 to 3.6 V, TA = 0 to 70 °C for F/K-Grade)
Parameter
Symbol
Test Condition
Min
Typ
Max
Unit
High Level Input Voltage1
VIH
2.0
—
—
V
1
Low Level Input Voltage
VIL
—
—
0.8
V
High Level Output Voltage
VOH
IO = –2 mA
2.4
—
—
V
Low Level Output Voltage
VOL
IO = 2 mA
—
—
0.35
V
AOUT High Level Voltage
VAH
IO = 10 mA
2.4
—
—
V
AOUT Low Level Voltage
VAL
IO = 10 mA
—
—
0.35
V
–10
—
10
µA
Input Leakage Current
Power Supply Current, Digital
IL
2
Total Supply Current, Sleep Mode2
Total Supply Current, Deep Sleep
2,3
ID
VD pin
—
8.5
10
mA
ID
PDN = 1, PDL = 0
—
5.0
6.0
mA
ID
PDN = 1, PDL = 1
—
1.3
1.5
mA
Notes:
1. VIH/VIL do not apply to C1A/C2A.
2. All inputs at 0.4 or VD – 0.4 (CMOS levels). All inputs are held static except clock and all outputs unloaded
(Static IOUT = 0 mA).
3. RGDT is not functional in this state.
Rev. 1.31
7
Si3050 + Si3018/19
Table 4. AC Characteristics
(VD = 3.0 to 3.6 V, TA = 0 to 70 °C for F/K-Grade, Fs = 8000 Hz, see "2. Typical Application Schematic" on page 17)
Parameter
Symbol
Sample Rate
PCLK Input Frequency
Test Condition
Min
Typ
Max
Unit
Fs
8
—
16
kHz
PCLK
256
—
8192
kHz
Receive Frequency Response
Low –3 dBFS Corner, FILT = 0
—
5
—
Hz
Receive Frequency Response
Low –3 dBFS Corner, FILT = 1
—
200
—
Hz
FULL = 0 (0 dBm)
—
1.1
—
VPEAK
FULL = 1 (+3.2 dBm)2
—
1.58
—
VPEAK
dBm)2
—
2.16
—
VPEAK
VFS
Transmit Full-Scale Level1
FULL2 = 1 (+6.0
VFS
Receive Full-Scale
Level1,3
FULL = 0 (0 dBm)
—
1.1
—
VPEAK
dBm)2
—
1.58
—
VPEAK
FULL2 = 1 (+6.0 dBm)2
—
2.16
—
VPEAK
FULL = 1 (+3.2
Dynamic Range4,5,6
DR
ILIM = 0, DCV = 11, MINI=00
DCR = 0, IL = 100 mA
—
80
—
dB
Dynamic Range4,5,6
DR
ILIM = 0, DCV = 00, MINI=11
DCR = 0, IL = 20 mA
—
80
—
dB
Dynamic Range4,5,6
DR
ILIM = 1, DCV = 11, MINI=00
DCR = 0, IL = 50 mA
—
80
—
dB
Transmit Total Harmonic
Distortion6,7
THD
ILIM = 0, DCV = 11, MINI=00
DCR = 0, IL = 100 mA
—
–72
—
dB
Transmit Total Harmonic
Distortion6,7
THD
ILIM = 0, DCV = 00, MINI=11
DCR = 0, IL = 20 mA
—
–78
—
dB
Receive Total Harmonic
Distortion6,7
THD
ILIM = 0, DCV = 00, MINI=11
DCR = 0, IL = 20 mA
—
–78
—
dB
Receive Total Harmonic
Distortion6,7
THD
ILIM = 1,DCV = 11, MINI=00
DCR = 0, IL = 50 mA
—
–78
—
dB
Notes:
1. Measured at TIP and RING with 600 termination at 1 kHz, as shown in Figure 1 on page 6.
2. With FULL = 1, the transmit and receive full-scale level of +3.2 dBm can be achieved with a 600  ac termination. While
the transmit and receive level in dBm varies with reference impedance, the DAA will transmit and receive 1 dBV into all
reference impedances. With FULL2 = 1, the transmit and receive full-scale level of +6.0 dBm can be achieved with a
600  termination. In this mode, the DAA will transmit and receive +1.5 dBV into all reference impedances.
3. Receive full-scale level produces –0.9 dBFS at DTX.
4. DR = 20 x log (RMS VFS/RMS Vin) + 20 x log (RMS Vin/RMS noise). The RMS noise measurement excludes harmonics.
Here, VFS is the 0 dBm full-scale level per Note 1 above.
5. Measurement is 300 to 3400 Hz. Applies to both transmit and receive paths.
6. Vin = 1 kHz, –3 dBFS.
7. THD = 20 x log (RMS distortion/RMS signal).
8. DRCID = 20 x log (RMS VCID/RMS VIN) + 20 x log (RMS VIN/RMS noise). VCID is the 1.5 V full-scale level with the
enhanced caller ID circuit. With the typical CID circuit, the VCID full-scale level is 6 V peak, and the DRCID decreases to
50 dB.
9. Refer to Tables 10–11 for relative gain accuracy characteristics (passband ripple).
10. Analog hybrid only. ZACIM controlled by ACIM in Register 30.
8
Rev. 1.31
Si3050 + Si3018/19
Table 4. AC Characteristics (Continued)
(VD = 3.0 to 3.6 V, TA = 0 to 70 °C for F/K-Grade, Fs = 8000 Hz, see "2. Typical Application Schematic" on page 17)
Parameter
Symbol
Test Condition
Min
Typ
Max
Unit
DRCID
VIN = 1 kHz, –13 dBFS
—
62
—
dB
—
1.5
—
VPEAK
2-W to DTX,
TXG2, RXG2, TXG3,
and RXG3 = 0000
–0.5
0
0.5
dB
300–3.4 kHz, ZACIM = ZLINE
20
—
—
dB
1 kHz, ZACIM = ZLINE
—
30
—
dB
Two-Wire Return Loss
300–3.4 kHz, all ac
terminations
25
—
—
dB
Two-Wire Return Loss
1 kHz, all ac terminations
—
32
—
dB
Dynamic Range (Caller ID mode)
Caller ID Full-Scale Level
8
8
VCID
Gain Accuracy6,9
Transhybrid Balance10
Transhybrid Balance
10
Notes:
1. Measured at TIP and RING with 600 termination at 1 kHz, as shown in Figure 1 on page 6.
2. With FULL = 1, the transmit and receive full-scale level of +3.2 dBm can be achieved with a 600  ac termination. While
the transmit and receive level in dBm varies with reference impedance, the DAA will transmit and receive 1 dBV into all
reference impedances. With FULL2 = 1, the transmit and receive full-scale level of +6.0 dBm can be achieved with a
600  termination. In this mode, the DAA will transmit and receive +1.5 dBV into all reference impedances.
3. Receive full-scale level produces –0.9 dBFS at DTX.
4. DR = 20 x log (RMS VFS/RMS Vin) + 20 x log (RMS Vin/RMS noise). The RMS noise measurement excludes harmonics.
Here, VFS is the 0 dBm full-scale level per Note 1 above.
5. Measurement is 300 to 3400 Hz. Applies to both transmit and receive paths.
6. Vin = 1 kHz, –3 dBFS.
7. THD = 20 x log (RMS distortion/RMS signal).
8. DRCID = 20 x log (RMS VCID/RMS VIN) + 20 x log (RMS VIN/RMS noise). VCID is the 1.5 V full-scale level with the
enhanced caller ID circuit. With the typical CID circuit, the VCID full-scale level is 6 V peak, and the DRCID decreases to
50 dB.
9. Refer to Tables 10–11 for relative gain accuracy characteristics (passband ripple).
10. Analog hybrid only. ZACIM controlled by ACIM in Register 30.
Table 5. Absolute Maximum Ratings
Parameter
Symbol
Value
Unit
DC Supply Voltage
VD
–0.5 to 3.6
V
Input Current, Si3050 Digital Input Pins
IIN
±10
mA
VIND
–0.3 to (VD + 0.3)
V
TA
–40 to 100
°C
TSTG
–65 to 150
°C
Digital Input Voltage
Ambient Operating Temperature Range
Storage Temperature Range
Note: Permanent device damage can occur if the above Absolute Maximum Ratings are exceeded. Functional operation
should be restricted to the conditions as specified in the operational sections of this data sheet. Exposure to absolute
maximum rating conditions for extended periods might affect device reliability.
Rev. 1.31
9
Si3050 + Si3018/19
Table 6. Switching Characteristics—General Inputs
(VD = 3.0 to 3.6 V, TA = 0 to 70 °C for K-Grade, CL = 20 pF)
Parameter1
Symbol
Min
Typ
Max
Unit
Cycle Time, PCLK
tp
0.12207
—
3.90625
s
PCLK Duty Cycle
tdty
40
50
60
%
PCLK Jitter Tolerance
tjitter
—
—
2
ns
tr
—
—
25
ns
tf
—
—
25
ns
PCLK Before RESET 
tmr
10
—
—
cycles
RESET Pulse Width3
trl
250
—
—
ns
tmxr
20
—
—
ns
tr
—
—
25
ns
Rise Time, PCLK
Fall Time, PCLK
2
CS, SCLK Before RESET
Rise Time, Reset
Notes:
1. All timing (except Rise and Fall time) is referenced to the 50% level of the waveform. Input test levels are
VIH = VD – 0.4 V, VIL = 0.4 V. Rise and Fall times are referenced to the 20% and 80% levels of the waveform.
2. FSYNC/PCLK relationship must be fixed after RESET
3. The minimum RESET pulse width is the greater of 250 ns or 10 PCLK cycle times.
tr
tp
PC LK
V IH
V IL
tm r
RESET
t rl
C S, SC LK
tm xr
Figure 2. General Inputs Timing Diagram
10
tf
Rev. 1.31
Si3050 + Si3018/19
Table 7. Switching Characteristics—Serial Peripheral Interface
(VIO = 3.0 to 3.6 V, TA = 0 to 70 °C for K-Grade, CL = 20 pF)
Parameter*
Symbol
Test
Conditions
Min
Typ
Max
Unit
Cycle Time SCLK
tc
61.03
—
—
ns
Rise Time, SCLK
tr
—
—
25
ns
Fall Time, SCLK
tf
—
—
25
ns
Delay Time, SCLK Fall to SDO Active
td1
—
—
20
ns
Delay Time, SCLK Fall to SDO
Transition
td2
—
—
20
ns
Delay Time, CS Rise to SDO Tri-state
td3
—
—
20
ns
Setup Time, CS to SCLK Fall
tsu1
25
—
—
ns
Hold Time, SCLK to CS Rise
th1
20
—
—
ns
Setup Time, SDI to SCLK Rise
tsu2
25
—
—
ns
Hold Time, SCLK Rise to SDI Transition
th2
20
—
—
ns
Delay time between chip selects
tcs
220
—
—
ns
—
6
—
ns
Propagation Delay, SDI to SDITHRU
*Note: All timing (except Rise and Fall time) is referenced to the 50% level of the waveform. Input test levels are
VIH = VD – 0.4 V, VIL = 0.4 V. Rise and Fall times are referenced to the 20% and 80% levels of the waveform.
tr
SCLK
tf
tc
tsu1
th1
CS
tsu2
th2
tcs
SDI
SDO
td1
td2
td3
Figure 3. SPI Timing Diagram
Rev. 1.31
11
Si3050 + Si3018/19
Table 8. Switching Characteristics—PCM Highway Serial Interface
(VD = 3.0 to 3.6 V, TA = 0 to 70 °C for K-Grade, CL = 20 pF)
Parameter1
Test
Conditions
Symbol
Cycle Time PCLK
Min
Typ
Max
Units
122
—
3906
ns
—
—
—
—
—
—
—
—
256
512
768
1.024
1.536
2.048
4.096
8.192
—
—
—
—
—
—
—
—
kHz
kHz
kHz
MHz
MHz
MHz
MHz
MHz
tfp
—
125
—
s
tp
Valid PCLK Inputs
FSYNC Period2
PCLK Duty Cycle
tdty
40
50
60
%
PCLK Jitter-Tolerance
tjitter
—
—
2
ns
FSYNC Jitter Tolerance
tjitter
—
—
±120
ns
tr
—
—
25
ns
Rise Time, PCLK
tf
—
—
25
ns
td1
—
—
20
ns
Delay Time, PCLK Rise to DTX Transition
td2
—
—
20
ns
3
Delay Time, PCLK Rise to DTX Tri-State
td3
—
—
20
ns
Setup Time, FSYNC Rise to PCLK Fall
tsu1
25
—
—
ns
Hold Time, PCLK Fall to FSYNC Fall
th1
20
—
—
ns
Setup Time, DRX Transition to PCLK Fall
tsu2
25
—
—
ns
Hold Time, PCLK Falling to DRX Transition
th2
20
—
—
ns
Fall Time, PCLK
Delay Time, PCLK Rise to DTX Active
Notes:
1. All timing is referenced to the 50% level of the waveform. Input test levels are VIH = VO – 0.4 V, VIL = 0.4 V, rise and fall
times are referenced to the 20% and 80% levels of the waveform.
2. FSYNC must be 8 kHz under all operating conditions.
3. Specification applies to PCLK fall to DTX tri-state when that mode is selected.
tp
PCLK
th1
t fp
tsu1
FSYNC
tsu2
th2
DRX
td1
td2
td3
DTX
Figure 4. PCM Highway Interface Timing Diagram (RXS = TXS = 1)
12
Rev. 1.31
Si3050 + Si3018/19
Table 9. Switching Characteristics—GCI Highway Serial Interface
(VD = 3.0 to 3.6 V, TA = 0 to 70 °C for K-Grade, CL = 20 pF)
Parameter1
Symbol
Test
Conditions
Min
Typ
Max
Units
Cycle Time PCLK (Single Clocking Mode)
tp
—
488
—
ns
Cycle Time PCLK (Double Clocking Mode)
tp
—
244
—
ns
—
—
2.048
4.096
—
—
MHz
MHz
Valid PCLK Inputs
FSYNC Period2
tfp
—
125
—
µs
PCLK Duty Cycle
tdty
40
50
60
%
PCLK Jitter Tolerance
tjitter
—
—
2
ns
FSYNC Jitter Tolerance
tjitter
—
—
±120
ns
Rise Time, PCLK
tr
—
—
25
ns
Fall Time, PCLK
tf
—
—
25
ns
td1
—
—
20
ns
Delay Time, PCLK Rise to DTX Transition
td2
—
—
20
ns
3
td3
—
—
20
ns
Setup Time, FSYNC Rise to PCLK Fall
tsu1
25
—
—
ns
Hold Time, PCLK Fall to FSYNC Fall
th1
20
—
—
ns
Setup Time, DRX Transition to PCLK Fall
tsu2
25
—
—
ns
Hold Time, PCLK Falling to DRX Transition
th2
20
—
—
ns
Delay Time, PCLK Rise to DTX Active
Delay Time, PCLK Rise to DTX Tri-State
Notes:
1. All timing is referenced to the 50% level of the waveform. Input test levels are VIH = VO – 0.4 V, VIL = 0.4 V, rise and fall
times are referenced to the 20% and 80% levels of the waveform.
2. FSYNC must be 8 kHz under all operating conditions.
3. Specification applies to PCLK fall to DTX tri-state when that mode is selected.
tr
tf
tp
PCLK
th1
t su1
tfp
FSYNC
t su2
t
h2
DRX
t d1
t d2
t d3
DTX
Figure 5. GCI Highway Interface Timing Diagram (1x PCLK Mode)
Rev. 1.31
13
Si3050 + Si3018/19
tr
tf
PCLK
tfp
th1
tsu2
FSYNC
tsu2
th2
DRX
td2
td1
td3
DTX
Figure 6. GCI Highway Interface Timing Diagram (2x PCLK Mode)
Table 10. Digital FIR Filter Characteristics—Transmit and Receive
(VD = 3.0 to 3.6 V, Sample Rate = 8 kHz, TA = 0 to 70 °C for K-Grade)
Parameter
Symbol
Min
Typ
Max
Unit
Passband (0.1 dB)
F(0.1 dB)
0
—
3.3
kHz
F(3 dB)
0
—
3.6
kHz
–0.1
—
0.1
dB
—
4.4
—
kHz
–74
—
—
dB
—
12/Fs
—
s
Passband (3 dB)
Passband Ripple Peak-to-Peak
Stopband
Stopband Attenuation
Group Delay
tgd
Note: Typical FIR filter characteristics for Fs = 8000 Hz are shown in Figures 7, 8, 9, and 10.
Table 11. Digital IIR Filter Characteristics—Transmit and Receive
(VD = 3.0 to 3.6 V, Sample Rate = 8 kHz, TA = 0 to 70 °C for K-Grade)
Parameter
Passband (3 dB)
Symbol
Min
Typ
Max
Unit
F(3 dB)
0
—
3.6
kHz
–0.2
—
0.2
dB
—
4.4
—
kHz
–40
—
—
dB
—
1.6/Fs
—
s
Passband Ripple Peak-to-Peak
Stopband
Stopband Attenuation
Group Delay
tgd
Note: Typical IIR filter characteristics for Fs = 8000 Hz are shown in Figures 11, 12, 13, and 14. Figures 15 and 16 show
group delay versus input frequency.
14
Rev. 1.31
Si3050 + Si3018/19
Figure 7. FIR Receive Filter Response
Figure 9. FIR Transmit Filter Response
Figure 8. FIR Receive Filter Passband Ripple
Figure 10. FIR Transmit Filter Passband Ripple
For Figures 7–10, all filter plots apply to a sample rate of
Fs = 8 kHz.
For Figures 11–14, all filter plots apply to a sample rate of
Fs = 8 kHz.
Rev. 1.31
15
Si3050 + Si3018/19
Figure 11. IIR Receive Filter Response
Figure 14. IIR Transmit Filter Passband Ripple
Figure 12. IIR Receive Filter Passband Ripple
Figure 15. IIR Receive Group Delay
Figure 13. IIR Transmit Filter Response
Figure 16. IIR Transmit Group Delay
16
Rev. 1.31
Ground Start
PCM Highway
SPI Control
/RG
/TGD
/TGDE
/RGDT
/INT
/RESET
FSYNC
PCLK
DTX
DRX
SCLK
SDITHRU
SDO
SDI
/CS
R51
Rev. 1.31
R53
1
2
3
4
5
6
7
8
9
10
U1
C50
Si3050
SDO
SDI_THRU
SDI
SCLK
CS
GND
FSYNC
VDD
PCLK
VA
DTX
C1A
DRX
C2A
RGDT
RST
AOUT/INT
TGDE
RG
TGD
R302
20
19
18
17
16
15
14
13
12
11
C51
R13
R12
C2
C1
R9
C4
C5
+
R1
1
2
3
4
5
6
7
8
Si3019
QE
DCT2
DCT
IGND
RX
DCT3
IB
QB
C1B
QE2
C2B
SC
VREG VREG2
RNG1 RNG2
U2
16
15
14
13
12
11
10
9
R8
C6
R2
C31
R32
R31
R7
R30
C30
Optional CID Population
R33
R3
R10
Q5
R11
Q4
-
R4
C3
+
D1
C7
R5
No Ground Plane In DAA Section
Z1
Q1
FB1
FB2
Q2
Q3
R6
C10
C8
C9
R15
R16
Figure 17. Typical Application Circuit for the Si3050 and Si3018/19
(Refer to “AN67: Si3050/52/54/56 Layout Guidelines” for Recommended Layout Guidelines)
R52
VDD
RV1
TIP
RING
Si3050 + Si3018/19
2. Typical Application Schematic
17
Si3050 + Si3018/19
3. Bill of Materials
Component
Value
Supplier(s)
C1, C2
33 pF, Y2, X7R, ±20%
Panasonic, Murata, Vishay
C31
3.9 nF, 250 V, X7R, ±20%
Venkel, SMEC
C4
1.0 µF, 50 V, Elec/Tant, ±20%
Panasonic
C5, C6, C50, C51
0.1 µF, 16 V, X7R, ±20%
Venkel, SMEC
C7
2.7 nF, 50 V, X7R, ±20%
Venkel, SMEC
C8, C9
680 pF, Y2, X7R, ±10%
Panasonic, Murata, Vishay
C10
0.01 µF, 16 V, X7R, ±20%
Venkel, SMEC
120 pF, 250 V, X7R, ±10%
Venkel, SMEC
Dual Diode, 225 mA, 300 V, (CMPD2004S)
Central Semiconductor
FB1, FB2
Ferrite Bead, BLM18AG601SN1
Murata
Q1, Q3
NPN, 300 V, MMBTA42
Central OnSemi, Fairchild
Q2
PNP, 300 V, MMBTA92
Central OnSemi, Fairchild
Q4, Q5
NPN, 80 V, 330 mW, MMBTA06
Central OnSemi, Fairchild
RV1
Sidactor, 275 V, 100 A
Teccor, Diodes Inc., Shindengen
R1
1.07 k, 1/2 W, 1%
Venkel, SMEC, Panasonic
R2
150 , 1/16 W, 5%
Venkel, SMEC, Panasonic
R3
3.65 k, 1/2 W, 1%
Venkel, SMEC, Panasonic
R4
2.49 k, 1/2 W, 1%
Venkel, SMEC, Panasonic
R5, R6
100 k, 1/16 W, 5%
Venkel, SMEC, Panasonic
Not Installed, 20 M, 1/8 W, 5%
Venkel, SMEC, Panasonic
R9
1 M, 1/16 W, 1%
Venkel, SMEC, Panasonic
R10
536 , 1/4 W, 1%
Venkel, SMEC, Panasonic
R11
C30,
C311
D1, D2
R7,
2
R81
73.2 , 1/2 W, 1%
Venkel, SMEC, Panasonic
R133
0 , 1/16 W
Venkel, SMEC, Panasonic
4
0 , 1/16 W
Venkel, SMEC, Panasonic
1
15 M, 1/8 W, 5%
Venkel, SMEC, Panasonic
R331
5.1 M, 1/8 W, 5%
Venkel, SMEC, Panasonic
R51, R52, R53
4.7 k, 1/16 W, 5%
Venkel, SMEC, Panasonic
U1
Si3050
Silicon Labs
U2
Si3018/19
Silicon Labs
Z1
Zener Diode, 43 V, 1/2 W
General Semi, On Semi, Diodes Inc.
R12,
R15, R16
R30, R32
R31,
Notes:
1. R7–R8 may be substituted for R30–R33 and C30–C31 for lower cost, but reduced CID performance.
2. Several diode bridge configurations are acceptable. Parts, such as a single HD04, a DF-04S, or four 1N4004 diodes,
may be used (suppliers include General Semiconductor, Diodes Inc., etc.).
3. 56 , 1/16, 1% resistors may be substituted for R12–R13 (0 ) to decrease emissions. (See AN81.)
4. Murata BLM18AG601SN1 may be substituted for R15–R16 (0 ) to decrease emissions. (See AN81.)
18
Rev. 1.31
Si3050 + Si3018/19
4. AOUT PWM Output
Figure 18 illustrates an optional circuit to support the pulse width modulation (PWM) output capability of the Si3050
for call progress monitoring purposes.To enable this mode, the INTE bit (Register 2) should be set to 0, the PWME
bit (Register 1) set to 1, and the PWMM bits (Register 2) set to 00.
+5VA
LS1
R41
Q6
AOUT
C41
Figure 18. AOUT PWM Circuit for Call Progress
Table 12. Component Values—AOUT PWM
Component
Value
Supplier
LS1
Speaker BRT1209PF-06
Intervox
Q6
NPN KSP13
Fairchild
C41
0.1 µF, 16 V, X7R, ±20%
Venkel, SMEC
R41
150 1/10 W, ±5%
Venkel, SMEC, Panasonic
Registers 20 and 21 allow the receive and transmit paths to be attenuated linearly. When these registers are set to
all 0s, the transmit and receive paths are muted. These registers affect the call progress output only and do not
affect transmit and receive operations on the telephone line.
The PWMM[1:0] bits (Register 1, bits 5:4) select one of three different PWM output modes for the AOUT signal,
including a delta-sigma data stream, a 32 kHz return to 0 PWM output, and a balanced 32 kHz PWM output.
Rev. 1.31
19
Si3050 + Si3018/19
5. Functional Description
The Si3050 is an integrated direct access arrangement
(DAA) providing a programmable line interface that
meets global telephone line requirements. The Si3050
implements Silicon Laboratories’ patented isolation
capacitor technology, which offers the highest level of
integration by replacing an analog front end (AFE), an
isolation transformer, relays, opto-isolators, and a 2- to
4-wire hybrid with two highly-integrated ICs.
The Si3050 DAA is fully software programmable to meet
global requirements and is compliant with FCC, TBR21,
JATE, and other country-specific PTT specifications as
shown in Table 13. In addition, the Si3050 meets the
most stringent global requirements for out-of-band
energy, emissions, immunity, high-voltage surges, and
safety, including FCC Parts 15 and 68, EN55022,
EN55024, and many other standards.
5.1.1. Si3018

Selectable
dc terminations.
selectable ac terminations to increase return loss
and trans-hybrid loss performance.
+6 dBm TX/RX level mode (600 )
Four
5.1.2. Si3019

5.1. Line-Side Device Support
Two different line-side devices are available for use with
the Si3050 system-side device. Both line-side devices
support the following features:
Global compliance.
Selectable 5 Hz or 200 Hz RX low-pass filter pole.
 –16.5 to 13.5 dB digital gain/attenuation adjustment
in 0.1 dB increments for the transmit and receive
paths.


20
Globally-compliant line-side device—targets global
DAA requirements for voice applications. This
line-side device supports both FCC-compliant
countries and non-FCC-compliant countries.
Rev. 1.31
Globally-compliant, enhanced features line-side
device—targets global DAA requirements for voice
applications.
Selectable
dc terminations
selectable ac terminations to further increase
return loss and trans-hybrid loss performance.
Line voltage monitoring in on- and off-hook modes to
enable line in-use/parallel handset detection.
Programmable line current / voltage threshold interrupt.
Polarity reversal interrupt.
+3.2 dBm TX/RX level mode (600 )
+6 dBm TX/RX level mode (600 )
Higher resolution (1.1 mA/bit) loop current
measurement.
Sixteen
Si3050 + Si3018/19
Table 13. Country Specific Register Settings
Register
16
31
16
16
26
26
26
Country
OHS
OHS2
RZ
RT
ILIM
Argentina
0
0
0
0
0
11
00
0000
Australia1
1
0
0
0
0
01
01
0011
Austria
0
1
0
0
1
11
00
0010
Bahrain
0
1
0
0
1
11
00
0010
Belgium
0
1
0
0
1
11
00
0010
Brazil
0
0
0
0
0
11
00
0001
Bulgaria
0
1
0
0
1
11
00
0011
Canada
0
0
0
0
0
11
00
0000
Chile
0
0
0
0
0
11
00
0000
China
0
0
0
0
0
11
00
1010
Colombia
0
0
0
0
0
11
00
0000
Croatia
0
1
0
0
1
11
00
0010
Cyprus
0
1
0
0
1
11
00
0010
Czech Republic
0
1
0
0
1
11
00
0010
Denmark
0
1
0
0
1
11
00
0010
Ecuador
0
0
0
0
0
11
00
0000
Egypt
0
1
0
0
1
11
00
0010
El Salvador
0
0
0
0
0
11
00
0000
Finland
0
1
0
0
1
11
00
0010
France
0
1
0
0
1
11
00
0010
Germany
0
1
0
0
1
11
00
0010
Greece
0
1
0
0
1
11
00
0010
Guam
0
0
0
0
0
11
00
0000
Hong Kong
0
0
0
0
0
11
00
0000
Hungary
0
1
0
0
1
11
00
0010
Iceland
0
1
0
0
1
11
00
0010
India
0
0
0
0
0
11
00
0000
Indonesia
0
0
0
0
0
11
00
0000
DCV[1:0] MINI[1:0]
30
ACIM[3:0]
Note:
1. See "5.16. DC Termination" on page 30 for DCV and MINI settings.
2. Supported for loop current  20 mA.
3. TBR21 includes the following countries: Austria, Belgium, Denmark, Finland, France, Germany, Greece,
Iceland, Ireland, Italy, Luxembourg, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, and the
United Kingdom.
Rev. 1.31
21
Si3050 + Si3018/19
Table 13. Country Specific Register Settings (Continued)
Register
16
31
16
16
26
Country
OHS
OHS2
RZ
RT
ILIM
Ireland
0
1
0
0
1
11
00
0010
Israel
0
1
0
0
1
11
00
0010
Italy
0
1
0
0
1
11
00
0010
Japan
0
0
0
0
0
10
01
0000
Jordan
0
0
0
0
0
01
01
0000
Kazakhstan
0
0
0
0
0
11
00
0000
Kuwait
0
0
0
0
0
11
00
0000
Latvia
0
1
0
0
1
11
00
0010
Lebanon
0
1
0
0
1
11
00
0010
Luxembourg
0
1
0
0
1
11
00
0010
Macao
0
0
0
0
0
11
00
0000
0
0
0
0
0
01
01
0000
Malta
0
1
0
0
1
11
00
0010
Mexico
0
0
0
0
0
11
00
0000
Morocco
0
1
0
0
1
11
00
0010
Netherlands
0
1
0
0
1
11
00
0010
New Zealand
0
0
0
0
0
11
00
0100
Nigeria
0
1
0
0
1
11
00
0010
Norway
0
1
0
0
1
11
00
0010
Oman
0
0
0
0
0
01
01
0000
Pakistan
0
0
0
0
0
01
01
0000
Peru
0
0
0
0
0
11
00
0000
Philippines
0
0
0
0
0
01
01
0000
Poland
0
1
0
0
1
11
00
0010
Portugal
0
1
0
0
1
11
00
0010
Romania
0
1
0
0
1
11
00
0010
Russia
0
0
0
0
0
11
00
0000
Saudi Arabia
0
0
0
0
0
11
00
0000
Singapore
0
0
0
0
0
11
00
0000
Malaysia
2
26
26
DCV[1:0] MINI[1:0]
30
ACIM[3:0]
Note:
1. See "5.16. DC Termination" on page 30 for DCV and MINI settings.
2. Supported for loop current  20 mA.
3. TBR21 includes the following countries: Austria, Belgium, Denmark, Finland, France, Germany, Greece,
Iceland, Ireland, Italy, Luxembourg, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, and the
United Kingdom.
22
Rev. 1.31
Si3050 + Si3018/19
Table 13. Country Specific Register Settings (Continued)
Register
16
31
16
16
26
26
26
Country
OHS
OHS2
RZ
RT
ILIM
Slovakia
0
1
0
0
1
11
00
0010
Slovenia
0
1
0
0
1
11
00
0010
South Africa
0
0
1
0
0
11
00
0011
South Korea
0
0
1
0
0
11
00
0000
Spain
0
1
0
0
1
11
00
0010
Sweden
0
1
0
0
1
11
00
0010
Switzerland
0
1
0
0
1
11
00
0010
Taiwan
0
0
0
0
0
11
00
0000
TBR213
0
0
0
0
1
11
00
0010
Thailand
0
0
0
0
0
01
01
0000
UAE
0
0
0
0
0
11
00
0000
United Kingdom
0
1
0
0
1
11
00
0101
USA
0
0
0
0
0
11
00
0000
Yemen
0
0
0
0
0
11
00
0000
DCV[1:0] MINI[1:0]
30
ACIM[3:0]
Note:
1. See "5.16. DC Termination" on page 30 for DCV and MINI settings.
2. Supported for loop current  20 mA.
3. TBR21 includes the following countries: Austria, Belgium, Denmark, Finland, France, Germany, Greece,
Iceland, Ireland, Italy, Luxembourg, Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, and the
United Kingdom.
Rev. 1.31
23
Si3050 + Si3018/19
5.2. Power Supplies
The Si3050 operates from a 3.3 V power supply. The
Si3050 input pins can only accept 3.3 V CMOS signal
levels. If support of 5 V signal levels is necessary, a
level shifter is required. The Si3018/19 derives its power
from two sources: the Si3050 and the telephone line.
The Si3050 supplies power over the patented isolation
capacitor link between the two devices, allowing the
Si3019 to communicate with the Si3050 while on-hook
and perform other on-hook functions, such as line
voltage monitoring. When off-hook, the Si3018/19 also
derives power from the line current supplied from the
telephone line. This feature is exclusive to DAAs from
Silicon Labs and allows the most cost-effective
implementation for a DAA while still maintaining robust
performance over all line conditions.
5.3. Initialization
Each time the Si3050 is powered up, assert the RESET
pin. When the RESET pin is deasserted, the registers
have default values to guarantee the line-side device
(Si3018/19) is powered down without the possibility of
loading the line (i.e., off-hook). An example initialization
procedure follows:
1. Power up and de-assert RESET.
2. Wait until the PLL is locked. This time is less than
1 ms from the application of PCLK.
3. Enable PCM (Register 33) or GCI (Register 42)
mode.
4. Set the desired line interface parameters (i.e.,
DCV[1:0], MINI[1:0], ILIM, DCR, ACIM[3:0], OHS,
RT, RZ, TGA2, and TXG2[3:0]) shown in Table 13 on
page 21.
The communications link is disabled by default. To
enable it, the PDL bit (Register 6, bit 4) must be
cleared. No communication between the Si3050 and
Si3018/19 can occur until this bit is cleared. Allow the
PLL to lock to the PCLK and FSYNC input signals
before clearing the PDL bit.
5.5. Power Management
The Si3050 supports four basic power management
operation modes. The modes are normal operation,
reset operation, sleep mode, and full powerdown mode.
The power management modes are controlled by the
PDN and PDL bits (Register 6).
On powerup, or following a reset, the Si3050 is in reset
operation. The PDL bit is set, and the PDN bit is
cleared. The Si3050 is operational, except for the
communications link. No communication between the
Si3050 and line-side device (Si3018/19) can occur
during reset operation. Bits associated with the line-side
device are invalid in this mode.
In typical applications, the DAA will predominantly be
operated in normal mode. In normal mode, the PDL and
PDN bits are cleared. The DAA is operational and the
communications link passes information between the
Si3050 and the Si3018 or Si3019.
The Si3050 supports a low-power sleep mode that
supports ring validation and wake-up-on-ring features.
To enable the sleep mode, the PDN bit must be set.
When the Si3050 is in sleep mode, the PCLK signal
must remain active. In low-power sleep mode, the
Si3050 is non-functional except for the communications
link and the RGDT signal. To take the Si3050 out of
sleep mode, pulse the reset pin (RESET) low.
5. Set the FULL (or FULL2) + IIRE bits as required.
In summary, the powerdown/up sequence for sleep
mode is as follows:
6. Write a 0x00 into Register 6 to power up the
line-side device (Si3018/19).
1. Ensure the PDL bit (Register 6, bit 4) is cleared.
When this procedure is complete, the Si3018/19 is
ready for ring detection and off-hook operation.
5.4. Isolation Barrier
The Si3050 achieves an isolation barrier through
low-cost, high-voltage capacitors in conjunction with
Silicon Laboratories’ patented signal processing
techniques. Differential capacitive communication
eliminates
signal
degradation
from
capacitor
mismatches, common mode interference, or noise
coupling. As shown in the "2. Typical Application
Schematic" on page 17, the C1, C2, C8, and C9
capacitors isolate the Si3050 (system-side) from the
Si3018/19 (line-side). Transmit, receive, control, ring
detect, and caller ID data are passed across this barrier.
24
2. Set the PDN bit (Register 6, bit 3).
3. The device is now in sleep mode. PCLK must remain
active.
4. To exit sleep mode, reset the Si3050 by pulsing the
RESET pin.
5. Program registers to desired settings.
The Si3050 also supports an additional Powerdown
mode. When both the PDN (Register 6, bit 3) and PDL
(Register 6, bit 4) bits are set, the chipset enters a
complete powerdown mode and draws negligible
current (deep sleep mode). In this mode, the Si3050 is
non-functional. The RGDT pin does not function and the
Si3050 will not detect a ring. Normal operation can be
restored using the same process for taking the Si3050
out of sleep mode.
Rev. 1.31
Si3050 + Si3018/19
5.6. Calibration
The Si3050 initiates two auto-calibrations by default
when the device goes off-hook or experiences a loss of
line power. A 17 ms resistor calibration is performed to
allow circuitry internal to the DAA to adjust to the exact
line conditions present at the time of going off-hook.
This resistor calibration can be disabled by setting the
RCALD bit (Register 25, bit 5). A 256 ms ADC
calibration is also performed to remove offsets that
might be present in the on-chip A/D converter, which
could affect the A/D dynamic range. The ADC
auto-calibration is initiated after the DAA dc termination
stabilizes and the resistor calibration completes. Due to
the large variation in line conditions and line card
behavior presented to the DAA, it might be beneficial to
use manual ADC calibration instead of auto-calibration.
Manual ADC calibration should be executed as close as
possible to 256 ms before valid transmit/receive data is
expected.
The following steps should be taken to implement
manual ADC calibration:
1. The CALD bit (auto-calibration disable—Register 17)
must be set to 1.
2. The MCAL bit (manual calibration) must be toggled
to one and then 0 to begin and complete the
calibration.
3. The calibration is completed in 256 ms.
5.7. In-Circuit Testing
The Si3050’s advanced design provides the designer
with an increased ability to determine system
functionality during production line tests and support for
end-user diagnostics. Six loopback modes allow
increased coverage of system components. For four of
the test modes, a line-side power source is needed.
While a standard phone line can be used, the test circuit
in Figure 1 on page 6 is adequate. In addition, an
off-hook sequence must be performed to connect the
power source to the line-side device.
For the start-up loopback test mode, no line-side power
is necessary, and no off-hook sequence is required. The
start-up test mode is enabled by default. When the PDL
bit (Register 6, bit 4) is set (the default case), the line
side is in a powerdown mode, and the system-side is in
a digital loopback mode. In this mode, data received on
DRX passes through the internal filters and is
transmitted on DTX. This path introduces approximately
0.9 dB of attenuation on the DRX signal received. The
group delay of both transmit and receive filters exists
between DRX and DTX. Clearing the PDL bit disables
this mode, and the DTX data switches to the receive
data from the line side. When the PDL bit is cleared, the
FDT bit (Register 12, bit 6) becomes active to indicate
that successful communication between the line side
and system side is established. This provides
verification that the communications link is operational.
The digital data loop-back mode offers a way to input
data on the DRX pin and have the identical data output
on the DTX pin through bypassing the transmit and
receive filters. Setting the DDL bit (Register 10, bit 0)
enables this mode, which provides an easy way to verify
communication between the host processor/DSP and
the DAA. No line-side power or off-hook sequence is
required for this mode.
The remaining test modes require an off-hook sequence
to operate. The following sequence lists the off-hook
requirements:
1. Powerup or reset.
2. Allow the internal PLL to lock on PCLK and FSYNC.
3. Enable line-side by clearing PDL bit.
4. Issue an off-hook command.
5. Delay 402.75 ms for calibration to occur.
6. Set desired test mode.
The communications link digital loopback mode allows
the host processor to provide a digital input test pattern
on DRX and receive that digital test pattern back on
DTX. To enable this mode, set the IDL bit (Register 1,
bit 1). The communications link is tested in this mode.
The digital stream is delivered across the isolation
capacitors, C1 and C2, of the "2. Typical Application
Schematic" on page 17, to the line-side device and
returned across the same path. In this digital loopback
mode, the 0.9 dB attenuation and filter group delays
also exist.
The PCM analog loopback mode extends the signal
path of the analog loopback mode. In this mode, an
analog signal is driven from the line into the line-side
device. This analog signal is converted to digital data
and then passed across the communications link to the
system-side device. The data passes through the
receive filter, through the transmit filter, and is then
passed across the communications link and sent back
out onto the line as an analog signal. Set the PCML bit
(Register 33, bit 7) to enable this mode.
With the final testing mode, internal analog loopback,
the system can test the operation of the transmit and
receive paths on the line-side device and the external
components in the "2. Typical Application Schematic"
on page 17. The host provides a digital test waveform
on DRX. Data passes across the isolation barrier, is
transmitted to and received from the line, passes back
across the isolation barrier, and is presented to the host
on DTX. Clear the HBE bit (Register 2, bit 1) to enable
this mode.
Rev. 1.31
25
Si3050 + Si3018/19
Note: All test modes are mutually exclusive. If more than one
test mode is enabled concurrently, the results are
unpredictable.
mode (or 2x full scale) is enabled by setting the FULL2
bit in Register 30. With FULL2 = 1, the full-scale signal
level increases to +6.0 dBm into a 600  load or
1.5 dBV into all reference impedances. The full-scale
and enhanced full-scale modes provide the ability to
trade off TX power and TX distortion for a peak signal.
By using the programmable digital gain registers in
conjunction with the enhanced full-scale signal level
mode, a specific power level (+3.2 dBm for example)
can be achieved across all ACT settings.
5.8. Exception Handling
5.11. Parallel Handset Detection
The Si3050 can determine if an error occurs during
operation. Through the secondary frames of the serial
link, the controlling DSP can read several status bits.
The bit of highest importance is the frame detect bit
(FDT, Register 12, bit 6) which indicates that the
system-side (Si3050) and line-side (Si3018 or Si3019)
devices are communicating. During normal operation,
the FDT bit can be checked before reading the bits that
indicate information about the line side. If FDT is not
set, the following bits related to the line side are
invalid—RDT, RDTN, RDTP, LCS[4:0], LSID[1:0],
REVB[3:0], LVS[7:0], LCS2[7:0], ROV, BTD, DOD, and
OVL; the RGDT operation is also non-functional.
The Si3050 can detect a parallel handset going
off-hook. When the Si3050 is off-hook, the loop current
can be monitored with the LCS or LCS2 bits. A
significant drop in loop current signals a parallel handset
going off-hook. If a parallel handset going off-hook
causes the loop current to drop to 0, the LCS and LCS2
bits will read all 0s. Additionally, the Drop-Out Detect
(DOD) bit will fire (and generate an interrupt if the
DODM bit is set) indicating that the line-derived power
supply has collapsed.
When the HBE bit is cleared, it produces a dc offset that
affects the signal swing of the transmit signal. Silicon
Laboratories recommends that the transmit signal be
12 dB lower than normal transmit levels. A lower level
eliminates clipping from the dc offset that results from
disabling the hybrid. It is assumed in this test that the
line ac impedance is nominally 600 
Following powerup and reset, the FDT bit is not set
because the PDL bit (Register 6 bit 4) defaults to 1. In
this state, the ISOcap is not operating and no
information about the line side can be determined. The
user must provide a valid PCLK and FSYNC to the
system and clear the PDL bit to activate the ISOcap
link. Communication with the line-side device takes less
than 10 ms to establish.
5.9. Revision Identification
The Si3050 provides information to determine the
revision of the Si3050 and/or the Si3018/19. The
REVA[3:0] bits (Register 11) identify the revision of the
Si3050, where 0101b denotes revision E. The
REVB[3:0] bits (Register 13) identify the revision of the
line-side device, where 0110b denotes revision F.
5.10. Transmit/Receive Full-Scale Level
The Si3050 supports programmable maximum transmit
and receive levels. The default signal level supported by
the Si3050 is 0 dBm into a 600  load. Two additional
modes of operation offer increased transmit and receive
level capability to enable use of the DAA in applications
that require higher signal levels. The full-scale mode is
enabled by setting the FULL bit in Register 31. With
FULL = 1 (Si3019 only), the full-scale signal level
increases to +3.2 dBm into a 600  load or 1 dBV into
all reference impedances. The enhanced full-scale
26
With the Si3019 line side, the LVS bits also can be read
when on- or off-hook to determine the line voltage.
Significant drops in line voltage can signal a parallel
handset. For the Si3050 to operate in parallel with
another handset, the parallel handset must have a
sufficiently high dc termination to support two off-hook
DAAs on the same line. Improved parallel handset
operation can be achieved by changing the dc
impedance from 50  to 800  and reducing the DCT
pin voltage with the DCV[1:0] bits.
5.12. Line Voltage/Loop Current Sensing
The Si3050 can measure loop current with either the
Si3018 or the Si3019 line-side device. The 5-bit
LCS[4:0] register reports loop current measurements
when off-hook. The Si3019 offers an additional register
to report loop current to a finer resolution (LCS2[7:0]).
The Si3050 can only measure line voltage when used
with the Si3019 line-side device. The LVS[7:0] register
available with the Si3019 monitors voltage both on and
off-hook. These registers can be used to help determine
the following line conditions:
When on-hook, detect if a line is connected.
When on-hook, detect if a parallel phone is off-hook.
 When off-hook, detect if a parallel phone goes on or
off-hook.
 Detect if enough loop current is available to operate.
 When used in conjunction with the OPD bit, detect if
an overload condition exists. (See "5.26. Overload
Detection" on page 36.)


Rev. 1.31
Si3050 + Si3018/19
5.12.1. Line Voltage Measurement
(Si3019 Line Side Only)
The Si3050 reports line voltage with the LVS[7:0] bits
(Register 29) in both on- and off-hook states with a
resolution of 1 V per bit. The accuracy of these bits is
approximately ±10%. Bits 0 through 7 of this 8-bit
signed number indicate the value of the line voltage in
2s complement format. Bit 7 indicates the polarity of the
TIP/RING voltage.
If the INTE bit (Register 2, bit 7) and the POLM bit
(Register 3, bit 0) are set, a hardware interrupt is
generated on the AOUT/INT pin when Bit 7 of the LVS
register changes state. The edge-triggered interrupt is
cleared by writing 0 to the POLI bit (Register 4, bit 0).
The POLI bit is set each time bit 7 of the LVS register
changes state, and must be written to 0 to clear it. The
default state of the LVS register forces the LVS[7:0] bits
to 0 when the line voltage is 3 V or less. The LVFD bit
(Register 31, bit 0) disables this force-to-zero function
and allows the LVS register to display non-zero values
of 3 V and below. This register may display
unpredictable values at line voltages between 0 to 2 V.
At 0 V, the LVS register displays all 0s.
Possible Overload
30
25
20
LCS
BITS
15
10
5
0
0
3.3
6.6
9.9 13.2 16.5 19.8 23.1 26.4 29.7
33
36.3 39.6 42.9 46.2 49.5 52.8 56.1 59.1 62.7 66
69.3 72.6 75.9 79.2 82.5 85.8 89.1 92.4 95.7 99 102.3
Loop Current (mA)
127
Figure 19. Typical Loop Current LCS Transfer Function (ILIM = 0)
Rev. 1.31
27
Si3050 + Si3018/19
5.12.2. Loop Current Measurement
When the Si3050 is off-hook, the LCS[4:0] bits measure
loop current in 3.3 mA/bit resolution. With the LCS[4:0]
bits, a user can detect another phone going off-hook by
monitoring the dc loop current. The line current sense
transfer function is shown in Figure 19 and is detailed in
Table 14. The LCS and LCS2 bits report loop current
down to the minimum operating loop current for the
DAA. Below this threshold, the reported value of loop
current is unpredictable. The minimum operating loop
current of the DAA is set by the MINI[1:0] bits in
Register 26.
When the LCS bits reach max value, the Loop Current
Sense Overload Interrupt bit (Register 4) fires. LCSOI
firing however, does not necessarily imply that an
overcurrent situation has occurred. An overcurrent
situation in the DAA is determined by the status of the
OPD bit (Register 19). After the LCSOI interrupt fires,
the OPD bit should be checked to determine if an
overcurrent situation exists. The OPD bit indicates an
overcurrent situation when loop current exceeds either
160 mA (ILIM = 0) or 60 mA (ILIM = 1), depending on
the setting of the ILIM bit (Register 26).
Table 14. Loop Current Transfer Function
LCS[4:0]
Condition
00000
Insufficient line current for normal operation.
Use the DOD bit (Register 19, bit 1) to
determine if a line is still connected.
00100
Minimum line current for normal operation.
(MINI[1:0] = 01)
11111
Loop current may be excessive. Use the
OPD bit to determine if an overload condition exists.
The LCS2 register also reports loop current in the
off-hook state. This register has a resolution of 1.1 mA
per bit.
5.13. Off-Hook
The communication system generates an off-hook
command by setting the OH bit (Register 5, bit 0). This
off-hook state seizes the line for incoming/outgoing
calls. It also can be used for pulse dialing.
28
With the OH bit at logic 0, negligible dc current flows
through the hookswitch. When a logic 1 is written to the
OH bit, the hookswitch transistor pair, Q1 and Q2, turn
on. A termination impedance across TIP and RING is
applied and causes dc loop current to flow. The
termination impedance has both an ac and a dc
component.
Several events occur in the DAA when the OH bit is set.
There is a 250 µs latency for the off-hook command to
be communicated to the line-side device. When the
line-side device goes off-hook, an off-hook counter
forces a delay to allow line transients to settle before
transmission or reception can occur. The off-hook
counter time is controlled by the FOH[1:0] bits
(Register 31, bits 6:5). The default setting for the
off-hook counter time is 128 ms, but can be adjusted up
to 512 ms or down to 64 or 8 ms.
After the off-hook counter expires, a resistor calibration
is performed for 17 ms to allow the DAA internal
circuitry to adjust to the exact conditions present at the
time of going off-hook. This resistor calibration can be
disabled by setting the RCALD bit (Register 25, bit 5).
After the resistor calibration is performed, an ADC
calibration is performed for 256 ms. This calibration
helps to remove offset in the A/D sampling the
telephone line. ADC calibration can be disabled by
setting the CALD bit (Register 17, bit 5). See "5.6.
Calibration" on page 25 for more information on
automatic and manual calibration.
Silicon Laboratories recommends that the resistor and
the ADC calibrations not be disabled except when a fast
response is needed after going off-hook, such as when
responding to a Type II Caller-ID signal. See "5.25.
Caller ID" on page 34 for detailed information.
To calculate the total time required to go off-hook and
start transmission or reception, include the digital filter
delay (typically 1.5 ms with the FIR filter) in the
calculation.
5.14. Ground Start Support
The Si3050 DAA supports loop-start applications by
default. It can also support ground-start applications
with the RG, TGD, and TGDE pins and the schematic
shown in Figure 20. The component values are listed in
Table 15.
Rev. 1.31
Si3050 + Si3018/19
(Register 32, bit 2). The DAA may then be taken
off-hook and the relay in series with RING opened (clear
the RG bit). The call continues as in loop-start mode.
VD
R106
5.14.3. CO Requests Line Seizure
TGDb
-24V
In a normal on-hook state, the relay in series with TIP
should be closed, connecting the –24 V isolated supply.
The CO grounds TIP to request line seizure, causing
current to flow. The opto-isolator U3 (see Figure 20 on
page 29) detects this current and sets the TGD bit
(Register 32, bit 2). This bit remains high as long as
current is detected. The TGDI bit (Register 4, bit 1) is a
sticky bit, and remains high until cleared. A hardware
interrupt on the AOUT/INT can be made to occur when
TIP current begins to flow by enabling the TGDM bit
(Register 3, bit 1). Clear the interrupt by writing 0 to the
TGDI bit (Register 4 bit 1). The DAA may then be taken
off-hook and the call continued as in loop-start mode.
R105
U3
1
1
4
4
2
2
3
3
Opto-Isolator
VD
R104
R102
R103
RL1
TGDEb
RGb
1
1
8
8
2
2
7
7
TIP
3
3
6
6
RING
4
4
5
5
Opto-Relay
R101
5.15. Interrupts
Figure 20. Typical Application Circuit for
Ground Start Support on the SI3050
Table 15. Component Values for the Ground
Start Support Schematic
Symbol
Value
Supplier(s)
R101
200 , 3 W, ±5%
Venkel, SMEC,
Panasonic
R102, R103,
R106
1 k, 1/10 W, ±5%
Venkel, SMEC,
Panasonic
R104
1.5 k, 1/10 W, ±5% Venkel, SMEC,
Panasonic
R105
10 k, 1/2 W, ±5%
Venkel, SMEC,
Panasonic
RL1
AQW210S
Aromat, NEC
U3
PS2501L-1
NEC, Fairchild
5.14.1. Ground Start Idle
Ensure the relay in series with TIP is closed by clearing
the TGOE bit (Register 32, bit 1). This enables the DAA
to sense if the CO grounds TIP. Set RG to 1
(Register 32, bit 0) so that no current flows through the
relay connecting RING to ground.
5.14.2. DAA Requests Line Seizure
With TGOE set to zero, seize the line by closing the
relay in series with RING (clear the RG bit,
Register 32, bit 0). The CO detects this current flowing
on RING and grounds TIP. This sets the TGD bit
The AOUT/INT pin can be used as a hardware interrupt
pin by setting the INTE bit (Register 2, bit 7). When this
bit is set, the analog output used for call progress
monitoring is not available. The default state of this
interrupt output pin is active low, but active high
operation can be enabled by setting the INTP bit
(Register 2, bit 6). This pin is an open-drain output
when the INTE bit is set and requires a 4.7 k pullup or
pulldown for correct operation. If multiple INT pins are
connected to a single input, the combined pullup or
pulldown resistance should equal 4.7 k Bits 7–0 in
Register 3 and bit 1 in Register 44 can be set to enable
hardware interrupt sources (bit 0 is available with the
Si3019 line-side only). When one or more of these bits
is set, the AOUT/INT pin goes into an active state and
stays active until the interrupts are serviced. If more
than one hardware interrupt is enabled in Register 3,
use software polling to determine the cause of the
interrupts. Register 4 and bit 3 of Register 44 contain
sticky interrupt flag bits. Clear these bits after servicing
the interrupt.
Registers 43 and 44 contain the line current/voltage
threshold interrupt. These line current/voltage registers
and interrupt are only available with the Si3019 line-side
device. This interrupt is triggered when the measured
line voltage or current in the LVS or LCS2 registers, as
selected by the CVS bit (Register 44, bit 2), crosses the
threshold programmed into the CVT[7:0] bits. With the
CVP bit, the interrupt can be programmed to occur
when the measured value rises above or falls below the
threshold. Only the magnitude of the measured value is
used for comparison to the threshold programmed into
the CVT[7:0] bits. Therefore, only positive numbers
should be used as a threshold.
Rev. 1.31
29
Si3050 + Si3018/19
5.16. DC Termination
The DAA has programmable settings for the dc
impedance, current limiting, minimum operational loop
current and TIP/RING voltage. The dc impedance of the
DAA is normally represented with a 50  slope as
shown in Figure 21, but can be changed to an 800 
slope by setting the DCR bit. This higher dc termination
presents a higher resistance to the line as loop current
increases.
Voltage Across DAA (V)
12
FCC DCT Mode
8
Finally, Australia has separate dc termination
requirements for line seizure versus line hold. Japan
mode may be used to satisfy both requirements.
However, if a higher transmit level for modem operation
is desired, switch to FCC mode 500 ms after the initial
off-hook. This satisfies the Australian dc termination
requirements.
7
5.17. AC Termination
11
10
9
6
.01 .02 .03 .04 .05 .06 .07 .08 .09 .1 .11
Loop Current (A)
Figure 21. FCC Mode I/V Characteristics,
DCV[1:0] = 11, MINI[1:0] = 00, ILIM = 0
For applications requiring current limiting per the TBR21
standard, the ILIM bit may be set to select this mode. In
this mode, the dc I/V curve is changed to a 2000 
slope above 40 mA, as shown in Figure 22. This allows
the DAA to operate with a 50 V, 230  feed, which is the
maximum linefeed specified in the TBR21 standard.
Voltage Across DAA (V)
45
The Si3018 provides four ac termination impedances
when used with the Si3050. The ACIM[3:0] bits in
Register 30 are used to select the ac impedance setting
on the Si3018. The four available settings for the Si3018
are listed in Table 16. If an ACIM[3:0] setting other than
the four listed in Table 16 is selected, the ac termination
is forced to 600  (ACIM[3:0] = 0000). The
programmable digital hybrid can be used to further
reduce near-end echo for each of the four listed ac
termination settings. See "5.28. Transhybrid Balance"
on page 37 for details.
Table 16. AC Termination Settings for the
Si3018 Line-Side Device
TBR21 DCT Mode
ACIM[3:0]
AC Termination
0000
600 
0011
220  + (820  || 120 nF) and 220  +
(820  || 115 nF)
0100
370  + (620  || 310 nF)
1111
Global complex impedance
40
35
30
25
20
15
The Si3019 provides sixteen ac termination
impedances when used with the Si3050. The ACIM[3:0]
bits in Register 30 are used to select the ac impedance
setting on the Si3019. The sixteen available settings for
the Si3019 are listed in Table 17.
10
5
.015 .02 .025 .03 .035 .04 .045 .05 .055 .06
Loop Current (A)
Figure 22. TBR21 Mode I/V Characteristics,
DCV[1:0] = 11, MINI[1:0] = 00, ILIM = 1
30
The MINI[1:0] bits select the minimum operational loop
current for the DAA, and the DCV[1:0] bits adjust the
DCT pin voltage, which affects the TIP/RING voltage of
the DAA. These bits allow important trade-offs to be
made between signal headroom and minimum
operational loop current. Increasing TIP/RING voltage
increases signal headroom, whereas decreasing the
TIP/RING voltage allows compliance to PTT standards
in low-voltage countries, such as Japan. Increasing the
minimum operational loop current above 10 mA also
increases signal headroom and prevents degradation of
the signal level in low-voltage countries.
The most widely used ac terminations are available as
register options to satisfy various global PTT
requirements. The real 600  impedance satisfies the
requirements of FCC Part 68, JATE, and other country
requirements. The 270  + (750  || 150 nF) satisfies
Rev. 1.31
Si3050 + Si3018/19
the requirements of TBR21.
5.18. Ring Detection
There are two selections useful for satisfying
non-standard ac termination requirements. The
350  + (1000  || 210 nF) impedance selection in
Register 30 is the ANSI/EIA/TIA 464 compromise
impedance network for trunks. The last ac termination
selection, ACIM[3:0] = 1111, is designed to satisfy
minimum return loss requirements for every country that
requires a complex termination. By selecting this
setting, the system is ensured to meet minimum PTT
requirements.
The ring signal is resistively coupled from TIP and RING
to the RNG1 and RNG2 pins. The Si3050 supports
either full- or half-wave ring detection. With full-wave
ring detection, the designer can detect a polarity
reversal of the ring signal. See “5.25.Caller ID” on
page 34. The ring detection threshold is programmable
with the RT bit (Register 16, bit 0) and RT2 bit
(Register 17, bit 4). The ring detector output can be
monitored in three ways. The first method uses the
RGDT pin. The second method uses the register bits,
RDTP, RDTN, and RDT (Register 5). The final method
uses the DTX output.
For each of the sixteen ac termination settings, the
programmable digital hybrid can be used to further
reduce near-end echo. See "5.28. Transhybrid Balance"
on page 37 for details.
Table 17. AC Termination Settings for the
Si3019 Line-Side Device
ACIM[3:0]
AC Termination
0000
600 
0001
900 
0010
270  + (750  || 150 nF) and
275  + (780  || 150 nF)
0011
220  + (820  || 120 nF) and 220 
+ (820  || 115 nF)
0100
370  + (620  || 310 nF)
0101
320  + (1050  || 230 nF)
0110
370  + (820  || 110 nF)
0111
275  + (780  || 115 nF)
1000
120  + (820  || 110 nF)
1001
350  + (1000  || 210 nF)
1010
200  + (680  || 100 nF)
1011
600  + 2.16 µF
1100
900  + 1 µF
1101
900  + 2.16 µF
1110
600  + 1 µF
1111
Global complex impedance
The ring detector mode is controlled by the RFWE bit
(Register 18, bit 1). When the RFWE bit is 0 (default
mode), the ring detector operates in half-wave rectifier
mode. In this mode, only positive ring signals are
detected. A positive ring signal is defined as a voltage
greater than the ring threshold across RNG1-RNG2.
Conversely, a negative ring signal is defined as a
voltage less than the negative ring threshold across
RNG1-RNG2. When the RFWE bit is 1, the ring detector
operates in full-wave rectifier mode. In this mode, both
positive and negative ring signals are detected.
The first method to monitor ring detection output uses
the RGDT pin. When the RGDT pin is used, it defaults
to active low, but can be changed to active high by
setting the RPOL bit (Register 14, bit 1). This pin is an
open-drain output, and requires a 4.7 k pullup or
pulldown for correct operation. If multiple RGDT pins
are connected to a single input, the combined pullup or
pulldown resistance should equal 4.7 k
When the RFWE bit is 0, the RGDT pin is asserted
when the ring signal is positive, which results in an
output signal frequency equal to the actual ring
frequency. When the RFWE bit is 1, the RGDT pin is
asserted when the ring signal is positive or negative.
The output then appears to be twice the frequency of
the ring waveform.
The second method to monitor ring detection uses the
ring detect bits (RDTP, RDTN, and RDT). The RDTP
and RDTN behavior is based on the RNG1-RNG2
voltage. When the signal on RNG1-RNG2 is above the
positive ring threshold, the RDTP bit is set. When the
signal on RNG1-RNG2 is below the negative ring
threshold, the RDTN bit is set. When the signal on
RNG1-RNG2 is between these thresholds, neither bit is
set.
Rev. 1.31
31
Si3050 + Si3018/19
The RDT behavior is also based on the RNG1-RNG2
voltage. When the RFWE bit is 0, a positive ring signal
sets the RDT bit for a period of time. When the RFWE
bit is 1, a positive or negative ring signal sets the RDT
bit.
The RDT bit acts like a one shot. When a new ring
signal is detected, the one shot is reset. If no new ring
signals are detected prior to the one shot counter
reaching 0, then the RDT bit clears. The length of this
count is approximately 5 seconds. The RDT bit is reset
to 0 by an off-hook event. If the RDTM bit
(Register 3, bit 7) is set, a hardware interrupt occurs on
the AOUT/INT pin when RDT is triggered. This interrupt
can be cleared by writing to the RDTI bit
(Register 4, bit 7). When the RDI bit (Register 2, bit 2) is
set, an interrupt occurs on both the beginning and end
of the ring pulse as defined by the RTO bits
(Register 23, bits 6:3). Ring validation may be enabled
when using the RDI bit.
The third method to monitor detection uses the DTX
data samples to transmit ring data. If the ISOcap is
active (PDL=0) and the device is not off-hook or in
on-hook line monitor mode, the ring data is presented
on DTX. The waveform on DTX depends on the state of
the RFWE bit.
When RFWE is 0, DTX is –32768 (0x8000) while the
RNG1-RNG2 voltage is between the thresholds. When
a ring is detected, DTX transitions to +32767 when the
ring signal is positive, then goes back to –32768 when
the ring is near 0 and negative. Thus a near square
wave is presented on DTX that swings from –32768 to
+32767 in cadence with the ring signal.
When RFWE is 1, DTX sits at approximately +1228
while the RNG1-RNG2 voltage is between the
thresholds. When the ring becomes positive, DTX
transitions to +32767. When the ring signal goes near 0,
DTX remains near 1228. As the ring becomes negative,
the DTX transitions to –32768. This repeats in cadence
with the ring signal.
To observe the ring signal on DTX, watch the MSB of
the data. The MSB toggles at the same frequency as
the ring signal independent of the ring detector mode.
This method is adequate for determining the ring
frequency.
5.19. Ring Validation
Ring validation prevents false triggering of a ring
detection by validating the ring parameters. Invalid
signals, such as a line-voltage change when a parallel
handset goes off-hook, pulse dialing, or a high-voltage
line test are ignored. Ring validation can be enabled
during normal operation and in low-power sleep mode
when a valid external PCLK signal is supplied.
The ring validation circuit operates by calculating the
time between alternating crossings of positive and
negative ring thresholds to validate that the ring
frequency is within tolerance. High and low frequency
tolerances are programmable in the RAS[5:0] and
RMX[5:0] fields. The RCC[2:0] bits define how long the
ring signal must be within tolerance.
Once the duration of the ring frequency is validated by
the RCC bits, the circuitry stops checking for frequency
tolerance and begins checking for the end of the ring
signal, which is defined by a lack of additional threshold
crossings for a period of time configured by the
RTO[3:0] bits. When the ring frequency is first validated,
a timer defined by the RDLY[2:0] bits is started. If the
RDLY[2:0] timer expires before the ring timeout, then
the ring is validated and a valid ring is indicated. If the
ring timeout expires before the RDLY[2:0] timer, a valid
ring is not indicated.
Ring validation requires the following five parameters:
Timeout parameter to place a lower limit on the
frequency of the ring signal (the RAS[5:0] bits in
Register 24). The frequency is measured by
calculating the time between crossings of positive
and negative ring thresholds.
 Minimum count to place an upper limit on the
frequency (the RMX[5:0] bits in Register 22).
 Time interval over which the ring signal must be the
correct frequency (the RCC[2:0] bits in Register 23).
 Timeout period that defines when the ring pulse has
ended based on the most recent ring threshold
crossing.
 Delay period between when the ring signal is
validated and when a valid ring signal is indicated to
accommodate distinctive ringing.
The RNGV bit (Register 24, bit 7) enables or disables
the ring validation feature in both normal operating
mode and low-power sleep mode.

Ring validation affects the behavior of the RDT status
bit, the RDTI interrupt, the INT pin, and the RGDT pin.
1. When ring validation is enabled, the status bit seen
in the RDT read-only bit (r5.2), represents the
detected envelope of the ring. The ring validation
parameters are configurable so that this envelope
32
Rev. 1.31
Si3050 + Si3018/19
may remain high throughout a distinctive-ring
sequence.
2. The RDTI interrupt fires when a validated ring
occurs. If RDI is zero (default), the interrupt occurs
on the rising edge of RDT. If RDI is set, the interrupt
occurs on both rising and falling edges of RDT.
3. The INT pin follows the RDTI bit with configurable
polarity.
4. The RGDT pin can be configured to follow the
ringing signal envelope detected by the ring
validation circuit by setting RFWE to 0. If RFWE is
set to 1, the RGDT pin follows an unqualified ring
detect one-shot signal initiated by a ring-threshold
crossing and terminated by a fixed counter timeout
of approximately 5 seconds. (This information is
shown in Register 18).
5.20. Ringer Impedance and Threshold
The ring detector in a typical DAA is ac coupled to the
line with a large 1 F, 250 V decoupling capacitor. The
ring detector on the Si3018/19 is resistively coupled to
the line. This coupling produces a high ringer
impedance to the line of approximately 20 M to meet
the majority of country PTT specifications including FCC
and TBR21.
Several countries including Poland, South Africa, and
Slovenia require a maximum ringer impedance that can
be met with an internally-synthesized impedance by
setting the RZ bit (Register 16). Certain countries also
specify ringer thresholds differently. The RT and RT2
bits (Register 16 and Register 17, respectively) select
between three different ringer thresholds: 15 V ±10%,
21 V ±10%, and 45 V ±10%. These three settings
enable satisfaction of global ringer threshold
requirements. Thresholds are set so that a ring signal is
guaranteed to not be detected below the minimum, and
a ring signal is guaranteed to be detected above the
maximum.
5.21. Pulse Dialing and Spark Quenching
Pulse dialing is accomplished by going off- and on-hook
to generate make and break pulses. The nominal rate is
10 pulses per second. Some countries have strict
specifications for pulse fidelity including make and
break times, make resistance, and rise and fall times. In
a traditional, solid-state dc holding circuit, there are a
number of issues in meeting these requirements.
The Si3050 dc holding circuit has active control of the
on- and off-hook transients to maintain pulse dialing
fidelity.
Spark quenching requirements in countries, such as
Italy, the Netherlands, South Africa, and Australia, deal
with the on-hook transition during pulse dialing. These
tests provide an inductive dc feed resulting in a large
voltage spike. This spike is caused by the line
inductance and the sudden decrease in current through
the loop when going on-hook. The traditional way of
dealing with this problem is to put a parallel RC shunt
across the hookswitch relay. The capacitor is large
(~1 µF, 250 V) and relatively expensive. In the Si3050,
loop current can be controlled to achieve three distinct
on-hook speeds to pass spark quenching tests without
additional BOM components. Through the settings of
four bits in three registers, OHS (Register 16), OHS2
(Register 31), SQ0, and SQ1 (Register 59), a slow ramp
down of loop current can be achieved which induces a
delay between the time the OH bit is cleared and the
time the DAA actually goes on-hook.
To ensure proper operation of the DAA during pulse
dialing, disable the automatic resistor calibration that is
performed each time the DAA enters the off-hook state
by setting the RCALD bit (Register 25, bit 5).
5.22. Receive Overload Detection
The Voice DAA chipset is capable of monitoring and
reporting receive overload conditions on the line. Billing
tones, parallel phone off-hook events, polarity reversals
and other disturbances on the line may trigger multiple
levels of overload detection as described below.
Transient events less than 1.1 VPK on the line are
filtered out by the low-pass digital filter on the
Si3050+Si3019. The ROV and ROVI bits are set when
the received signal is greater than 1.1 VPK. Both bits
will continue to indicate an overload condition until a
zero is written to clear. The OVL mirrors the function of
the ROV and ROVI bits but it automatically clears after
the overload condition has been removed. When the
OVL bit returns to 0, the DAA initiates an
auto-calibration sequence that must complete before
data can be transmitted. An external interrupt can
optionally be triggered by the ROVI bit by setting the
ROVM and INTE bits.
Certain events such as billing tones can be sufficiently
large to disrupt the line-derived power supply of the
Voice DAA line side device (Si3018 or Si3019.) To
ensure that the device maintains the off-hook line state
during these events, the BTE bit should be set. If such
an event occurs while the BTE bit is set, the BTD and
BTDI bits will be asserted. A zero must be written to the
BTE bit to clear the BTD and BTDI bits. An external
interrupt can optionally be triggered by the BTDI bit by
setting the BTDM and INTE bits.
In the event that a line disturbance causes the loop
current to collapse below the minimum required
operating current of the Voice DAA, the DOD and DODI
Rev. 1.31
33
Si3050 + Si3018/19
bits will be set. An external interrupt can optionally be
triggered by the DODI bit by setting the DODM and
INTE bits.
5.23. Billing Tone Filter (Optional)
Optionally, a billing tone filter may be inserted between
the line and the voice DAA to minimize disruptions
caused by large billing tones. The notch filter design
requires two notches, one at 12 kHz and one at 16 kHz.
Because these components are expensive and few
countries utilize billing tones, this filter is typically placed
in an external dongle or added as a population option.
Figure 23 shows a billing tone filter example. Table 18
gives the component values.
L1 must carry the entire loop current. The series
resistance of the inductors is important to achieve a
narrow and deep notch. This design has more than
25 dB of attenuation at both 12 kHz and 16 kHz.
The on-hook line monitor mode allows the Si3050 to
receive line activity when in an on-hook state. This
mode is typically used to detect caller ID data (see
“5.25.Caller ID”) and is enabled by setting the ONHM bit
(Register 5, bit 3). Caller ID data can be gained up or
attenuated using the receive gain control bits in
Registers 39 and 41.
5.25. Caller ID
The Si3050 can pass caller ID data from the phone line
to a caller ID decoder connected to the DAA.
5.25.1. Type I Caller ID
Type I Caller ID sends the CID data when the phone is
on-hook.
C2
In systems where the caller ID data is passed on the
phone line between the first and second rings, utilize the
following method to capture the caller ID data:
L1
1. After identifying a ring signal using one of the
methods described in "5.18. Ring Detection" on
page 31, determine when the first ring is complete.
L2
2. Assert the ONHM bit (Register 5, bit 3) to enable
caller ID data detection. The caller ID data is passed
across the RNG 1/2 pins and presented to the host
via the DTX pin.
To
DAA
C3
3. Clear the ONHM bit after the caller ID data is
received.
RING
In systems where the caller ID data is preceded by a
line polarity (battery) reversal, use the following method
to capture the caller ID data:
Figure 23. Billing Tone Filter
Table 18. Component Values—Optional Billing
Tone Filters
Component
Value
C1,C2
0.027 µF, 50 V, ±10%
C3
0.01 µF, 250 V, ±10%
L1
3.3 mH, >120 mA, <10 , ±10%
L2
10 mH, >40 mA, <10 , ±10%
The billing tone filter affects the DAA’s ac termination
and return loss. The global compromise complex ac
termination as selected by ACIM[3:0] = 1111 passes
global return loss specifications with and without the
billing tone filter by at least 3 dB. This ac termination is
optimized for frequency response and hybrid
34
5.24. On-Hook Line Monitor
C1
TIP
From Line
cancellation and has greater than 4 dB of margin with or
without the dongle for South Africa, Australia, TBR21,
Germany,
and
Switzerland
country-specific
specifications.
1. Enable full wave rectified ring detection (RFWE,
Register 18, bit 1).
2. Monitor the RDTP and RDTN register bits (or the
POLI bit with the Si3019 line-side) to identify if a
polarity reversal or a ring signal has occurred. A
polarity reversal trips either the RDTP or RDTN ring
detection bits, therefore the full-wave ring detector
must be used to distinguish a polarity reversal from a
ring. The lowest specified ring frequency is 15 Hz;
so, if a battery reversal occurs, the DSP should wait
a minimum of 40 ms to verify that the event is a
battery reversal and not a ring signal. This time is
greater than half the period of the longest ring signal.
If another edge is detected during this 40 ms pause,
this event is characterized as a ring signal and not a
battery reversal.
Rev. 1.31
Si3050 + Si3018/19
f. Set the OH bit to return to an off-hook state.
Immediately after returning to an off-hook state,
the off-hook counter must be allowed to expire.
This allows the line voltage to settle before
transmitting or receiving data. After 8 ms normal
data transmission and reception can begin. If a
non-compliant parallel device is present, then a
reply tone is not sent by the host tone generator
and the CO does not send the CID data. If all
devices on the line are Type II CID compliant,
then the host must mute its upstream data output
to avoid the propagation of its reply tone and the
subsequent CID data. When muting its upstream
data output, the host processor should return an
acknowledgement (ACK) tone to the CO
requesting transmission of CID data.
3. Assert the ONHM bit (Register 5, bit 3) to enable
caller ID data detection. The caller ID data is passed
across the RNG 1/2 pins and presented to the host
via the DTX pin.
4. Clear the ONHM bit after the caller ID data is
received.
5.25.2. Type II Caller ID (Si3019 Line-Side Device
Only)
Type II Caller ID sends the CID data while the phone is
off-hook. This mode is often referred to as caller ID/
call waiting (CID/CW). To receive the CID data when
off-hook, use the following procedure (also see
Figure 24):
1. The Caller Alert Signal (CAS) tone is sent from the
central office (CO) and is digitized along with the line
data. The host processor detects the presence of
this tone.
2. The DAA must check if there is another parallel
device on the same line, which is accomplished by
briefly going on-hook, measuring the line voltage,
and returning to an off-hook state.
a. Set the CALD bit (Register 17, bit 5) to disable
the calibration that automatically occurs when
going off-hook.
3. The CO then responds with CID data after receiving
the CID data, the host processor unmutes the
upstream data output and continues with normal
operation.
4. The muting of the upstream data path by the host
processor mutes the handset in a telephone
application so the user cannot hear the
acknowledgement tone and CID data being sent.
b. Set the RCALD bit (Register 25, bit 5) to disable
the resistor calibration that automatically occurs
when going off-hook
5. The CALD and the RCALD bits can be cleared to
re-enable the automatic calibrations when going
off-hook. The FOH[1:0] bits also can be programmed
to 01 to restore the default off-hook counter time.
c. Set the FOH[1:0] bits (Register 31 bits 6:5) to 11
to reduce the time period for the off-hook counter
to 8 ms allowing compliance to the Type II CID
timing requirements.
Because of the nature of the low-power ADC, the data
presented on DTX can have up to a 10% dc offset. The
caller ID decoder must either use a high-pass or a
band-pass filter to accurately retrieve the caller ID data.
d. Clear the OH bit (Register 5, bit 0). This puts the
DAA into an on-hook state. The RXM bit
(Register 15, bit 3) also can be set to mute the
receive path.
e. Read the LVS bits to determine the state of the
line. If the LVS bits read the typical on-hook line
voltage, then there are no parallel devices active
on the line, and CID data reception can be
continued. If the LVS bits read well below the
typical on-hook line voltage, then there are one or
more devices present and active on the same line
that are not compliant with Type II CID. Do not
continue CID data reception.
Rev. 1.31
35
Si3050 + Si3018/19
1
LINE
On-Hook
Off-Hook Counter
and Calibration
(402.75 ms nominally)
CA S Tone
Of f -Hook
Received
2
3
On-Hook
5
4
Off-Hook Counter
(8 ms)
Of f -Hook
A ck
FOH[1] Bit
FOH[0] Bit
RCALD Bit
CALD Bit
OH Bit
Notes:
1. The off-hook counter and calibrations prevent transmission or reception of data for 402.75 ms (default) for the line
voltage to settle.
2. The caller alert signal (CAS) tone transmits from the CO to signal an incoming call.
3. The device is taken on-hook to read the line voltage in the LVS bits to detect parallel handsets. In this mode, no data is
transmitted on the DTX pin.
4. When the device returns off-hook, the normal off-hook counter is reduced to 8 ms. If the CALD and RCALD bits are set,
then the automatic calibrations are not performed.
5. After allowing the off-hook counter to expire (8 ms), normal transmission and reception can continue. If CID data
reception is required, send the appropriate signal to the CO at this time.
Figure 24. Implementing Type II Caller ID on the Si3050+Si3019
5.26. Overload Detection
5.27. Gain Control
The Si3050 can be programmed to detect an overload
condition that exceeds the normal operating power
range of the DAA circuit. To use the overload detection
feature, the following steps should be followed:
1. Set the OH bit (Register 5, bit 0) to go off-hook, and
wait 25 ms to allow line transients to settle.
The Si3050 supports multiple levels of gain and
attenuation for the transmit and receive paths.
The TXG2 and RXG2 bits (Registers 38–39) enable
gain or attenuation in 1 dB increments for the transmit
and receive paths (up to 12 dB gain and 15 dB
attenuation). The TGA2 and RGA2 bits select either
gain or attenuation. The TXG3 and RXG3 bits
(Registers 40–41) enable gain or attenuation in 0.1 dB
increments up to 1.5 dB for the transmit and receive
paths. The TGA3 and RGA3 bits select either gain or
attenuation. The transmit and receive paths can be
individually muted with the TXM and RXM bits
(Register 15). The signal flow through the Si3050 and
the Si3018/19 is shown in Figures 25–26.
2. Enable overload detection by then setting the OPE
bit (Register 17, bit 3).
If the DAA senses an overload situation it automatically
presents an 800  impedance to the line to reduce the
hookswitch current. At this time, the DAA also sets the
OPD bit (Register 19, bit 0) to indicate that an overload
condition exists. The line current detector within the
DAA has a threshold that is dependent on the ILIM bit
(Register 26). When ILIM = 0, the overload detection
threshold equals 160 mA. When ILIM = 1, the overload
detection threshold equals 60 mA. The OPE bit should
always be cleared before going off-hook.
36
Rev. 1.31
Si3050 + Si3018/19
DAC
To
Si3050
ACT
TX
Analog
Hybrid
Link
CO
0.6 Hz
HPF
ADC
Figure 25. Si3018/19 Signal Flow Diagram
DRX
IIRE
Digital
Filter
TXG3
TXA3
TXG2
1 dB
Gain
Steps
TXA2
1 dB
Attenuation
Steps
0.1 dB
Gain/ATT
Steps
Digital
Hybrid
0.1 dB
Gain/ATT
Steps
1 dB
Attenuation
Steps
DTX
RXA2
IIRE
Digital
Filter
RXG3
RXA3
1 dB
Gain
Steps
RXG2
Link
To
Si3018/19
Selectable
200 Hz
HPF
Figure 26. Si3050 Signal Flow Diagram
5.28. Transhybrid Balance
5.29. Filter Selection
The Si3050 contains an on-chip analog hybrid that
performs the 2- to 4-wire conversion and near-end echo
cancellation. This hybrid circuit is adjusted for each ac
termination setting selected to achieve a minimum
transhybrid balance of 20 dB when the line impedance
matches the impedance set by ACIM.
The Si3050 supports additional filter selections for the
receive and transmit signals as defined in Tables 10 and
11. The IIRE bit (Register 16, bit 4) selects between the
IIR and FIR filters. The IIR filter provides a shorter, but
non-linear, group delay alternative to the default FIR
filter, and only operates with an 8 kHz sample rate. The
FILT bit (Register 31, bit 1) selects a –3 dB low
frequency pole of 5 Hz when cleared and a –3 dB low
frequency pole of 200 Hz (per EIA/TIA 464) when set.
The FILT bit affects the receive path only.
The Si3050 also offers a digital hybrid stage for
additional near-end echo cancellation. For each ac
termination setting, the eight programmable hybrid
registers (Registers 45–52) can be programmed with
coefficients to increase cancellation of real-world line
impedances. This digital filter can produce 10 dB or
greater of near-end echo cancellation in addition to the
trans-hybrid loss from the analog hybrid circuitry.
Coefficients are 2s complement, where unity is
represented as binary 0100 0000b, the maximum value
as binary 0111 1111b, and the minimum value as binary
1000 000b. See AN84 for a more detailed description of
the digital hybrid and how to use it.
5.30. Clock Generation
The Si3050 generates the necessary 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, 768 kHz,
1.024 MHz, 1.53 MHz, 2.048 MHz, 4.09 MHz, or
8.192 MHz. The ratio of the PCLK rate to the FSYNC
rate is determined internally by the DAA and is
transferred into internal registers after a reset. These
internal registers are not accessible through register
reads or writes. Figure 27 shows the operation of the
Si3050 clock circuitry.
Rev. 1.31
37
Si3050 + Si3018/19
The PLL clock synthesizer settles quickly after powerup.
However, the settling time depends on the PCLK
frequency and it can be approximately predicted by the
following equation:
Tsettle = 64/FPCLK
For all valid PCLK frequencies listed above, the default
line sample rate is 8 kHz. This sample rate can be
increased to 16 kHz by setting the HSSM bit
(Register 7, bit 3). Regardless of the sample rate
frequency, the serial data communication rate of the
PCM and GCI highways remains 8 kHz. When the
16 kHz sample rate is selected, additional timeslots in
the PCM or GCI highway are used to transfer the
additional data.
5.31. Communication Interface Mode
Selection
The Si3050 supports two communication interface
protocols:
PCLK
N
PCM/SPI mode where data and control information
transmission/reception occurs across separate
buses (PCM highway for data, and SPI port for
control).
 GCI mode where data and control information is
multiplexed and transmission/reception occurs
across the GCI highway bus.
A pin-strapping method (specifically, the state of SCLK
on power-up [reset]) is used to select between the two
communication interface protocols. Tables 18 and 19
specify how to select a communication mode, and how
the various pins are used in each mode.

When operating in PCM/SPI mode, the GCI control
register should not be written (i.e., Register 42 must
each remain set at 0000_0000 when using the PCM/
SPI highway mode). Similarly, when operating in GCI
highway mode the PCM registers should not be written
(i.e., Registers 33–37 must remain set to 0000_0000
when using the GCI highway mode).
PFD
VCO
2
DIV M
Internal PLL
Register
Figure 27. PLL Clock Synthesizer
Table 19. PCM or GCI Highway Mode Selection
SCLK
SDI
Mode Selected
1
0
X
0
0
1
PCM Mode
GCI Mode,
B2 Channel used
GCI Mode,
B1 Channel used
Note: Values shown are the states of the pins at the rising edge of
RESET.
38
Rev. 1.31
2
16.384 MHz
Si3050 + Si3018/19
Table 20. Pin Functionality in PCM or GCI Highway Mode
Pin Name
PCM Mode
GCI Mode
SDI_THRU
SPI Data Throughput pin for Daisy Chaining
Operation (Connects to the SDI pin of the
subsequent device in the daisy chain)
Sub-frame
Selector, bit 2
SCLK
SPI Clock Input
PCM/GCI Mode Selector
SDI
SPI Serial Data Input
B1/B2 Channel Selector
SDO
SPI Serial Data Output
Sub-frame
Selector, bit 1
CS
SPI Chip Select
Sub-frame
Selector, bit 0
FSYNC
PCM Frame Sync Input
GCI Frame Sync Input
PCLK
PCM Input Clock
GCI Input Clock
DTX
PCM Data Transmit
GCI Data Transmit
DRX
PCM Data Receive
GCI Data Receive
Note: This table denotes pin functionality after the rising edge of RESET and mode selection.
5.32. PCM Highway
The Si3050 contains a flexible programmable interface
for the transmission and reception of digital PCM
samples. PCM data transfer is controlled via the PCLK
and FSYNC inputs, the PCM Transmit and Receive
Start Count registers (Registers 34–37), and the PCM
Mode Select register (Register 33). The interface can
be configured to support from 4 to 128 8-bit timeslots in
each frame, which corresponds to PCLK frequencies of
256 kHz to 8.192 MHz in power of 2 increments. Time
slot assignment and data delay from FSYNC edge are
handled via the TXS and RXS registers. These 10-bit
values are programmed with the number of PCLK
cycles following the rising edge of FSYNC until the data
transfer begins. Because the Si3050 looks for the rising
edge of FSYNC, both long and short FSYNC pulse
widths can be accommodated. A value of 0 in the PCM
Transmit and Receive Start Count registers signifies
that the MSB of the data should occur in the same cycle
as the rising edge of FSYNC.
By setting the correct starting point of the data, the
Si3050 can operate with buses having multiple devices
requiring different time slots. The DTX pin is high
impedance except during transmission of an 8-bit PCM
sample. DTX returns to high impedance either on the
negative edge of PCLK during the LSB or on the
positive edge of PCLK following the LSB. This behavior
is based on the setting of the TRI bit in 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, a 16-bit linear mode is also
provided. This mode can be activated via the PCMF bits
in the PCM Mode Select register. Double-clocked timing
also is supported in which the duration of a data bit is
two PCLK cycles. This mode is activated via the PHCF
bit in the PCM Mode Select register. Setting the TXS or
RXS registers greater than the number of PCLK cycles
in a sample period stops data transmission or reception.
Figures 28–31 illustrate the usage of the PCM highway
interface to adapt to common PCM standards.
Rev. 1.31
39
Si3050 + Si3018/19
PCLK
FSYNC
PCLK_CNT
0
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 28. PCM Highway Transmission, Short FSYNC, Single Clock Cycle Delayed Transmission
(TXS = RXS = 0, PHCF = 0, TRI = 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
DTX
MSB
LSB
MSB
LSB
H I-Z
H I-Z
Figure 29. PCM Highway Transmission, Long FSYNC (TXS = RXS = 0, PHCF = 0, TRI = 1)
40
Rev. 1.31
Si3050 + Si3018/19
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 30. PCM Highway Transmission, Long FSYNC, Delayed Data Transfer
(TXS = RXS = 10, PHCF = 0, TRI = 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
DTX
M SB
LSB
M SB
LSB
HI-Z
HI-Z
Figure 31. PCM Highway Double Clocked Transmission, Short FSYNC
(TXS = RXS = 0, PHCF = 1, TRI = 1)
Rev. 1.31
41
Si3050 + Si3018/19
5.33. Companding in PCM Mode
5.34. 16 kHz Sampling Operation in PCM
Mode
The Si3050 supports both µ-Law and A-Law
companding formats in addition to 16-bit linear data.
The 8-bit companding schemes follow a segmented
curve formatted as a sign bit, three chord bits, and four
step bits. µ-Law is commonly used in North America
and Japan, while A-Law is primarily used in Europe.
Data format is selected via the PCMF bits (Register 33).
Table 21 on page 43 and Table 22 on page 44 define
the µ-Law and A-Law encoding formats. If linear mode
is used the resulting 16-bit data is transmitted in two
consecutive 8-bit PCM highway timeslots as shown in
Figure 32.
The Si3050 can be configured to support a 16 kHz
sampling rate and transmit the data on an 8 kHz PCM or
GCI highway bus. By setting the HSSM bit
(Register 7, bit 3) to 1, the DAA changes its sampling
rate, Fs, to 16 kHz if it was originally configured for an
8 kHz sampling rate. If µ-law or A-law companding is
used, the resulting 8-bit samples are transmitted in two
consecutive 8-bit PCM highway timeslots. If linear mode
is used, the resulting 16-bit samples are transmitted in
four consecutive 8-bit PCM highway timeslots as shown
in Figure 33.
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
DTX
LSB
HI-Z
HI-Z
MSB
LSB
Figure 32. PCM Highway Transmission, Single Clock Cycle, 16-bit linear mode
(TXS = RXS = 0, PHCF = 0, TRI = 1, PCMF = 11)
PCLK
FSYNC
PCLK_CNT
0
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21
22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
DRX
MSB
LSB
DTX
HI-Z
HI-Z
MSB
LSB
Sample 1
Sample 2
Figure 33. PCM Highway Transmission, Single Clock Cycle, 16-bit linear mode
(TXS = RXS = 0, PHCF = 0, TRI = 1, PCMF = 11, HSSM = 1)
42
Rev. 1.31
Si3050 + Si3018/19
Table 21. µ-Law Encode-Decode Characteristics1,2
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
__________________
1x1
Notes:
1. Characteristics are symmetrical about analog 0 with sign bit = 1 for negative analog values.
2. Digital code includes inversion of both sign and magnitude bits.
Rev. 1.31
43
Si3050 + Si3018/19
Table 22. A-Law Encode-Decode Characteristics1,2
Segment
Number
#Intervals x interval size
Value at segment endpoints
7
16 x 128
4096
3968
.
.
2143
2015
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
.
.
.
1055
991
10110101b
1056
.
.
.
511
479
10000101b
528
.
.
.
239
223
10010101b
264
.
.
.
103
95
11100101b
132
.
.
.
35
31
11110101b
66
.
.
.
2
0
11010101b
1
Notes:
1. Characteristics are symmetrical about analog 0 with sign bit = 1 for negative analog values.
2. Digital code includes inversion of all even numbered bits.
44
Rev. 1.31
Si3050 + Si3018/19
5.35. SPI Control Interface
The control interface to the Si3050 is a 4-wire interface
modeled on commonly available micro-controller and
serial peripheral devices. The interface consists of four
pins: clock (SCLK), chip select (CS), serial data input
(SDI), and serial data output (SDO). In addition, the
Si3050 includes a serial data through output pin
(SDITHRU) to support daisy chain operation of up to 16
devices. The device can operate with 8-bit and 16-bit
SPI controllers. Each SPI operation consists of a control
byte, an address byte (of which only the six LSBs are
used internally), and either one or two data bytes
depending on the width of the controller. Bytes are
transmitted MSB first.
There are a number of variations of usage on this
four-wire interface as follows:

Continuous clocking. During continuous clocking,
assertion of the CS pin controls the data transfers.
The CS pin 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 edge of SCLK
after the transfer.
 Clock only during transfer. The clock is active during
the actual byte transfers only. Each byte transfer
consists of eight clock cycles in a return to 1 format.
 SDI/SDO wired operation. Independent of the
clocking options described, the SDI and SDO pins
can be treated as two separate lines or wired
together if the master can tri-state its output during
the data byte transfer of a read operation.
The SPI state machine resets when the CS pin is
asserted during an operation on an SCLK cycle that is
not a multiple of eight. This provides a mechanism for
the controller to force the state machine to a known
state in the case where the controller and the device are
not synchronized.
The control byte has the following structure and is
presented on the SDI pin MSB first.
7
6
5
4
3
2
1
0
BRCT
R/W
1
0
CID[0]
CID[1]
CID[2]
CID[3]
The bits are defined as follows:
Indicates a broadcast operation that is intended for all devices in the daisy chain. This is only
valid for write operations as it causes contention on the SDO pin during a read.
7
BRCT
6
R/W
5
1
This bit must be 1 at all times.
4
0
This bit must be 0 at all times.
3:0
CID[0:3]
Read/Write Bit.
1 = Read operation.
0 = Write operation.
This field 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 away in the SDI/SDITHRU chain. See Figure 34.
As the CID information propagates down the daisy chain, each channel decrements the CID by
1. The device that receives a value of 0 in the CID field responds to the SPI transaction. See
Figure 35. If a broadcast to all devices connected to the chain is requested, the CID do not
decrement. In this case, the same 8- or 16-bit data is presented to all channels regardless of
the CID values.
Rev. 1.31
45
Si3050 + Si3018/19
SDO
SCLK
SCLK
CPU
CS
CS
SDI
SDO
SDI
Si3050 #1
Channel 0
SDITHRU
SDI
SCLK
CS
Si3050 #2
Channel 1
SDO
SDITHRU
SDI
SCLK
CS
SDO
Si3050 #16
Channel 15
SDITHRU
Figure 34. SPI Daisy Chain Control Architecture
SPI Control Byte
BRCT
R/W
1
0
CID[0]
CID[1]
CID[2]
CID[3]
SDI0
0
0 or 1
1
0
0
0
0
0
SDI1
0
0 or 1
1
0
1
0
0
0
SDI2
0
0 or 1
1
0
0
1
0
0
SDI3
0
0 or 1
1
0
1
1
0
0
SDI14
0
0 or 1
1
0
0
1
1
1
SDI15
0
0 or 1
1
0
1
1
1
1
Figure 35. Sample SPI Control Byte to Access Channel 0
46
Rev. 1.31
Si3050 + Si3018/19
1
SDI0-15
0
1
0
X
X
X
X
Figure 36. Sample SPI Control Byte for Broadcast Mode (Write Only)
In Figure 35 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 BRCT and R/W bits remain unchanged as the control word passes through the entire chain. A
unique CID is presented to each device, and the device receiving a CID value of 0 is the target of the operation
(channel 0 in this case). Figure 36 illustrates that in broadcast mode, all bits pass through the chain without
permutation.
CS B
S CLK
S DI
CONTROL
A DDRE S S
DA TA [7:0]
Hi-Z
S DO
Figure 37. Write Operation via an 8-bit SPI Port
CSB
SCLK
SDI
CONTROL
ADDRESS
XXXXXXXXXXXX
Data [7:0]
SDO
Figure 38. Read Operation via an 8-bit SPI Port
Figure 37 and Figure 38 illustrate WRITE and READ operations via an 8-bit SPI controller. Each of these
operations are performed as a 3-byte transfer. The CS pin is asserted between each byte. The CS pin must be
asserted before the first falling edge of SCLK after the DATA byte to indicate to the state machine that only one
byte should be transferred. The state of the SDI pin is ignored during the DATA byte of a read operation.
CS B
S CLK
S DI
CONTROL
A DDRE S S
Data [7:0]
XXXXXXXX
Hi - Z
S DO
Figure 39. Write Operation via a 16-bit SPI Port
Rev. 1.31
47
Si3050 + Si3018/19
CS B
S CLK
S DI
CONTROL
A DDRE S S
S DO
XXXXXXXX
XXXXXXXX
Data [7:0]
Data [7:0]
S am e byte repeated twice.
Figure 40. Read Operation via a 16-bit SPI Port
Figures 39 and 40 illustrate WRITE and READ
operations via a 16-bit SPI controller. These operations
require a 4-byte transfer arranged as two 16-bit words.
The CS pin does not go high when the eighth bit of data
is received, which indicates to the SPI state machine
that eight more SCLK pulses follow to complete the
operation. In the case of a WRITE operation, the last
eight bits are ignored. In a read operation, the 8-bit data
value is repeated so that the data may be captured
during the last half of a data transfer if required by the
controller. The Si3050 autodetects the SPI mode (16-bit
or 8-bit mode).
5.36. GCI Highway
The Si3050 contains an alternate communication
interface to the SPI and PCM highway control and data
interface. The general circuit interface (GCI) can be
used for the transmission and reception of control and
data information onto a GCI highway bus. The PCM and
GCI highways are 4-wire interfaces and share the same
pins. The SPI control interface is not used as a
communication interface in the GCI highway mode, but
rather as hardwired channel selector pins.
When GCI mode is selected, the sub-frame selection
pins must be tied to the correct state to select one of
eight sub-frame timeslots in the GCI frame (Table 23).
These pins must remain in this state when the Si3050 is
operating. Selecting a particular subframe automatically
causes that individual Si3050 to transmit and receive on
the appropriate sub-frame in the GCI frame, which is
initiated by an FSYNC pulse. No more register settings
are needed to select which sub-frame a device uses,
and the sub-frame for a particular device cannot be
changed when in operation. Only one Si3050 DAA can
be assigned per sub-frame, which allows a total of eight
DAAs to be connected to the same GCI highway bus.
GCI mode supports a 1x and a 2x PCLK rate as shown
in Figures 5 and 6 on pages 13 and 14, respectively.
The PCLK rate is autodetected and no internal register
settings are needed to support either 1x or 2x PCLK
mode.
48
Table 23. GCI Mode Sub-Frame Selection
SDI_THRU 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
GCI Subframe 3 Selected
(Voice channels 7–8)
1
0
0
GCI Subframe 4 Selected
(Voice channels 9–10)
0
1
1
GCI Subframe 5 Selected
(Voice channels 11–12)
0
1
0
GCI Subframe 6 Selected
(Voice channels 13–14)
0
0
1
GCI Subframe 7 Selected
(Voice channels 15–16)
0
0
0
The GCI highway requires either a 2.048 or 4.096 MHz
clock frequency on the PCLK pin, and an 8 kHz frame
sync input on the FSYNC pin. The overall unit of data
used to communicate on the GCI highway is a frame,
which is 125 µs in length. Each frame is initiated by a
pulse on the FSYNC pin and the rising edge signifies
the beginning of the next frame. In 2x PCLK mode,
there are twice as many PCLK cycles during each
125 µs frame versus 1x PCLK mode. Each frame
consists of eight fixed timeslot sub-frames that are
assigned using the Sub-Frame Select pins as described
in Table 20 on page 39 (SDI_THRU, SDO, and CS).
Within each sub-frame are four channels (bytes) of
data, including the two voice data channels (B1 and
B2), one Monitor channel (M) for initialization and setup
of the device, and one Signaling and Control channel
Rev. 1.31
Si3050 + Si3018/19
(SC) for communicating status of the device and for
initiating commands. Within the SC channel are six
Command/Indicate (C/I) bits and two handshaking bits
(MR and MX). The C/I bits are used for status and
command communication, whereas the handshaking
bits Monitor Receive (MR) and Monitor Transmit (MX)
are used for data exchanges in the Monitor channel.
Figure 41 illustrates the contents of a GCI highway
frame.
5.37. Companding in GCI Mode
The Si3050 supports µ-Law and A-Law companding
formats in addition to 8-bit or 16-bit linear data. The 8-bit
companding schemes are described in "5.33.
Companding in PCM Mode" on page 42 and are shown
in Table 21 and Table 22. If 16-bit linear mode is used,
the resulting 16-bit samples are transmitted in both the
B1 and B2 channels of a single subframe. For proper
operation, select all Si3050 DAAs to use the B1 channel
with only one DAA per subframe.
5.38. 16 kHz Sampling Operation in GCI
Mode
The Si3050 can be configured to support a 16 kHz
sampling rate (as described in "5.34. 16 kHz Sampling
Operation in PCM Mode" on page 42) and transmit the
data on an 8 kHz GCI Highway bus. If 8-bit samples are
used with a 16 kHz sample rate, the samples are
transmitted in both the B1 and B2 channels of a single
subframe. If 16-bit linear mode is used, the resulting
16-bit samples are transmitted in both the B1 and B2
channels of two consecutive subframes. In this case,
assign one DAA per two subframes.
5.39. Monitor Channel
The Monitor channel is used for initialization and setup
of the Si3050. It also can be used for general
communication with the Si3050 by allowing read and
write access to the Si3050’s 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 below. For purposes of this
specification, “downstream” is identified to be the data
sent by a host to the Si3050. “Upstream” is identified to
be the data sent by the Si3050 to a host.
Figure 41 illustrates the Monitor channel communication
protocol. For successful communication with the
Si3050, 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 kbytes/second.
125 s
FSYNC
SF0
SF1
SF2
SF3
SF4
SF5
SF6
SF7
Sub-Frame
8
B1
8
B2
8
M
C/I
0
1
2
6
MR MX
1
1
Channel
SC Channel
Figure 41. Time-Multiplexed GCI Highway Frame Structure
Rev. 1.31
49
Si3050 + Si3018/19
1st Byte
2nd Byte
3rd Byte
MX
Transm itter
MX
MR
Receiver
MR
ACK
1st Byte
ACK
2nd Byte
ACK
3rd Byte
125  s
Figure 42. Monitor Handshake Timing
The Idle state is achieved by the MX and MR bits being
held inactive (signal is high) for two or more frames.
When a transmission is initiated by a host device, an
active state (signal is low) is present on the downstream
MX bit. This signals to the Si3050 that a transmission
has begun on the Monitor channel and the Si3050
should begin accepting data from host device. The
Si3050, 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. The end of message is signaled by the
downstream MX being held inactive for two or more
consecutive frames. Receipt of initial data is signaled by
the upstream MR bit’s transitioning from an inactive to
an active state. Upon receiving acknowledgement from
the Si3050 that the initial data is received, 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.
50
When the host is performing a write command, the host
only manipulates the downstream MX bit, and the
Si3050 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 Si3050 responding with the requested
data. Similarly, the Si3050 initially manipulates its
upstream MR bit to receive the read command, and
then manipulates its upstream MX bit to respond with
the requested data. If the host is transmitting data, the
Si3050 always transmits a $FF value on its Monitor data
byte. While the Si3050 is transmitting data, the host
should always transmit a $FF value on its Monitor byte.
If the Si3050 is transmitting data and detects a value
other than a $FF on the downstream Monitor byte, the
Si3050 signals an Abort.
For read and write commands, an initial address must
be specified. The Si3050 responds to a read or a write
command at this address, and then subsequently
increment this address after every register access.
Rev. 1.31
Si3050 + Si3018/19
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 downstream MX and MR bits inactive for two
consecutive frames. The transmission also can be
stopped by the Si3050 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 Si3050 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 Si3050 is transmitting data
When the Si3050 aborts because of an invalid
command sequence, the state of the Si3050 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 EOM is detected
before a valid command sequence is communicated,
the Si3050 returns to the idle state and remains
unchanged.

The data presented to the Si3050 in the downstream
Monitor bits must be present for two consecutive frames
to be considered valid data. The Si3050 checks to
ensure it receives the same data in two consecutive
frames. If 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 that it 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 Si3050 is still waiting to receive a valid
data byte transmission of two consecutive identical data
bytes, the Si3050 signals an abort and ends the
transmission. Figure 43 shows a state diagram for the
Receiver Monitor channel for the Si3050. Figure 44 on
page 53 shows a state diagram for the Transmitter
Monitor channel for the Si3050.
Rev. 1.31
51
Si3050 + Si3018/19
Idle
MR = 1
MX x LL
Initial
State
MX
MX
1st Byte
Received
MR = 0
Abort
MR = 1
MX
MX
ABT
Any
State
MX x LL
MX x LL
Byte
Valid
MR = 0
MX
Wait
for LL
MR = 0
MX x LL
MX
MX x LL
MX
New Byte
MR = 1
MX
nth Byte
received
MR = 1
MX x LL
Wait
for LL
MR = 0
MR: MR bit calculated and transmitted on DTX line.
MX: MX bit received data downstream (DRX line).
LL: Last look of monitor byte received on DRX line.
ABT: Abort indication to internal source.
Figure 43. Si3050 Monitor Receiver State Diagram
52
Rev. 1.31
MX
Si3050 + Si3018/19
MR x MXR
Idle
MR = 1
MR x MXR
MXR
Wait
MX = 1
MR x MXR
Abort
MX = 1
Initial
State
MR x RQT
MR
1s t Byte
MX = 0
MR x RQT
EOM
MX = 1
MR
MR x RQT
nth Byte
ack
MX = 1
MR
Wait for
ack
MX = 0
MR
MR x RQT
CLS/
ABT
MR x RQT
MR: MR bit received on DRX line.
MX: MX bit calculated and expected on DTX line.
MXR: MX bit s am pled on DTX line.
CLS: Collis ion within the m onitor data byte on DTX line.
RQT: Reques t for trans m is s ion from internal s ource.
ABT: Abort reques t/indication.
Any
State
Figure 44. Si3050 Monitor Transmitter State Diagram
Rev. 1.31
53
54
125 s
1 Frame
$FF
$91
Device Address
$91
Rev. 1.31
$FF
$FF
$FF
$FF
$81
R/W
$FF
$81
$10
$FF
$FF
Register Address
$10
$FF
$FF
$91
$FF
<product> sends
address before
data
Device Address
$91
$FF
$FF
$FF
$FF
EOM Signalled
$FF
$FF
$FF
EOM
Acknowledge
Contents of Contents of Contents of Contents of Contents of
Register
Register
Register
Register
Register
$10
$10
$11
$11
$12
(ignored by
host)
$FF
Figure 45. Example Read of Registers $10 and $11 in Subframe 0 of the Si3050
= Acknowledgement of data reception
MR Upstream Bit
MX Upstream Bit
$FF
Monitor Data Upstream
MR Downstream Bit
MX Downstream Bit
$FF
Monitor Data Downstream
Si3050 + Si3018/19
125 s
1 Frame
$FF
$91
Device Address
$91
Rev. 1.31
$FF
$FF
$FF
$FF
$01
R/W
$FF
$01
$10
$FF
$FF
Register Address
$10
$FF
Data to be
written to
$10
$FF
Data to be
written to
$10
$FF
Data to be
written to
$11
$FF
Data to be
written to
$11
$FF
$FF
EOM
Acknowledge
$FF
EOM Signalled
$FF
Figure 46. Example Write to Registers $10 and $11 in Subframe 0 of the Si3050
= Acknowledgement of data reception
MR Upstream Bit
MX Upstream Bit
$FF
Monitor Data Upstream
MR Downstream Bit
MX Downstream Bit
$FF
Monitor Data Downstream
Si3050 + Si3018/19
55
Si3050 + Si3018/19
5.40. Summary of Monitor Channel Commands
A = 1: Response to CID command from the device
using channel B2 is placed in Monitor Data.
Communication with the Si3050 should be in the
following format:
When C = 1, bits A and B are channel enable bits.
When these bits are set to 1, the individual
corresponding channels receives the command in the
next
command
byte.
The
channels
whose
corresponding bits are set to 0 ignores the subsequent
command byte.
Byte 1: Device Address Byte
Byte 2: Command Byte
Byte 3: Register Address Byte
Bytes 4-n: Data Bytes
Bytes n+1, n+2: EOM
A = 1: Channel B1 receives the command.
5.41. Device Address Byte
A = 0: Channel B1 does not receive the command.
The Device Address byte identifies which device
connected to the GCI highway receives the particular
message. This address should be the first byte sent to
the Si3050 at the beginning of every transmission
sequence. For Read commands, the address sent to
the Si3050 is the first byte transmitted in response to
the Read command before register data is transmitted.
This Device Address byte has the following structure:
1
0
0
A
B
0
0
C
B = 1: Channel B2 receives the command.
B = 0: Channel B2 does not receive the command.
5.42. Command Byte
The Command byte has the following structure:
RW
CMD[6:0]
The RW bit is a register read/write bit.
RW = 0: A write is performed to the Si3050’s register.
The lowest programmable bit, C, has a special function.
This bit enables a register read or write, or enables a
special Channel Identification Command (CID).
RW = 1: A read is performed on the Si3050’s register.
C = 1: Normal command follows.
CMD[6:0] = 0000001: Read or write a register on the
Si3050.
C = 0: Channel Identification Command.
The CID is a special command to identify themselves by
software. For this special command, the subsequent
command byte transmitted by the host processor must
be $00 (binary), and have no address or data bytes. The
Si3050 in turn responds with a fixed 2-byte identification
code:
CMD[6:0] = 0000010 – 1111111: Reserved.
5.43. Register Address Byte
The Register Address byte has the following structure:
ADDRESS[7:0]
1
0
0
A
0
0
0
0
This byte contains the actual 8-bit address of the
register to be read or written.
1
0
1
1
1
1
1
0
5.44. SC Channel
Upon sending the 2-byte identification code, the Si3050
sends an EOM (MR = MX = 1) for two consecutive
frames. When A = 0, B must be 0 or the Si3050 signals
an abort due to an invalid command. In this mode, bit C
is the only other programmable bit.
A = 0: Response to CID command from the device
using channel B1 is placed in Monitor Data.
56
The CMD[6:0] bits specify the actual command to be
performed.
The SC channel consists of six C/I bits and two
handshaking bits, MR and MX. One of these channels is
contained in every 4-byte sub-frame and is transmitted
every frame. The handshaking bits are described in the
above Monitor Channel section. The definition of the six
C/I bits depends on the direction the bits are being sent,
either transmitted to the GCI highway bus via the DTX
pin or received from the GCI highway bus via the DRX
pin.
Rev. 1.31
Si3050 + Si3018/19
5.45. Receive SC Channel
:
MSB
7
6
5
4
3
2
1
0
CIR6
CIR5
CIR4
CIR3
CIR2
CIR1
MR
MX
LSB
C/I Bits
These bits are defined as follows:
CIR6: Reserved
CIR5: Reserved
CIR4: ONHM
CIR3: TGDE
CIR2: RG
CIR1: OH
Data that is received must be consistent and match for
at least two consecutive frames to be considered valid.
When a new command or status is communicated via
the C/I bits, the data must be sent for at least two
consecutive frames to be recognized by the Si3050.
The following steps describe the protocol of how C/I bits
are stored, detected, and validated. This is illustrated in
Figure 47.
1. The current state of the C/I bits are stored in a
primary register P. If the received C/I bits are
identical to this current state, no action is taken.
2. Upon receipt of an SC channel with C/I bits that differ
from the current state, these new C/I bits are
immediately latched into a secondary register S.
3. The C/I bits in the SC channel received in the frame
immediately after the SC channel just stored in S are
compared with the C/I bits in the S register.
a. If the C/I bits in these two channels are identical,
then the C/I bits in the S register are loaded into
the P register and are considered a valid change
of C/I bits. The Si3050 then responds accordingly
to the changed C/I bits.
b. If a set of C/I bits is latched into the S register
and the subsequent set of C/I bits received does
not match either the S or P registers, then the
newly received set of C/I bits are latched into the
S register. This continues to occur as long as the
subsequent set of C/I bits received differs from
the C/I bits in the S and P registers.
c. If the C/I bits in the SC channel received
immediately after the SC channel just stored in S
do not match the C/I bits stored in S, but DO
match the C/I bits stored in P, then the single set
of C/I bits stored in the S latch are invalidated,
and the current state of the C/I bits in P remains
unchanged.
Rev. 1.31
57
Si3050 + Si3018/19
Receive New
CI Code
Yes
= P?
No
P: C/I Primary Register Contents
Store in S
S: C/I Secondary Register Contents
Receive New
C/I Code
Load C/I Register
With New C/I Bits
Yes
= S?
No
Yes
= P?
No
Figure 47. Protocol for Receiving C/I Bits in the Si3050
5.46. Transmit SC Channel
The following diagram shows the definition of the transmitted SC channel, which is transmitted MSB first.
MSB
7
6
5
4
3
2
1
0
CIT6
CIT5
CIT4
CIT3
CIT2
CIT1
MR
MX
C/I Bits
These bits are defined as follows:
CIT6: Reserved
CIT5: CVI
CIT4: DOD
CIT3: INT (represents the state of the INT pin)
CIT2: Battery Reversal (represents the state of bit 7
of the LVS register)
CIT1: TGD
58
Rev. 1.31
LSB
Si3050 + Si3018/19
6. Control Registers
Note: Registers not listed here are reserved and must not be written.
Table 24. Register Summary
Register
Name
Bit 7
Bit 6
1
Control 1
SR
2
Control 2
INTE
INTP
3
Interrupt Mask
RDTM
ROVM
4
Interrupt Source
RDTI
5
DAA Control 1
6
DAA Control 2
7
Sample Rate Control
8
Reserved
9
Reserved
10
DAA Control 3
11
System- and Line-Side Device Revision
12
Line-Side Device Status
13
Line-Side Device Revision
14
DAA Control 4
15
TX/RX Gain Control 1
16
International Control 1
17
International Control 2
18
International Control 3
19
International Control 4
20
Call Progress RX Attenuation
Bit 5
Bit 4
Bit 3
PWMM[1:0]
Bit 2
Bit 1
RDI
HBE
RXE
DODM
LCSOM
TGDM
POLM*
DODI
LCSOI
TGDI
POLI*
ONHM
RDT
PWME
IDL
WDTEN
FDTM
BTDM
ROVI
FDTI
BTDI
RDTN
RDTP
PDL
OH
PDN
HSSM
DDL
LSID[3:0]
REVA[3:0]
FDT
LCS[4:0]
1
REVB[3:0]
RPOL
TXM
RXM
OHS
CALZ
MCAL
IIRE
CALD
RT2
OPE
BTE
RZ
RT
ROV
BTD
RFWE
OVL
DOD
Call Progress TX Attenuation
22
Ring Validation Control 1
23
Ring Validation Control 2
RDLY[2]
24
Ring Validation Control 3
RNGV
25
Resistor Calibration
RCALS
26
DC Termination Control
27
Reserved
ATM[7:0]
RDLY[1:0]
RMX[5:0]
RTO[3:0]
RCC[2:0]
RAS[5:0]
RCALM
RCALD
DCV[1:0]
RCAL[3:0]
0
0
OHS2
0
FILT
TGD
TGDE
RG
0
PHCF
TRI
MINI[1:0]
28
Loop Current Status
LCS2[7:0]*
29
Line Voltage Status
LVS[7:0]*
AC Termination Control
31
DAA Control 5
32
Ground Start Control
33
PCM/SPI Mode Select
34
PCM Transmit Start Count—Low Byte
35
PCM Transmit Start Count—High Byte
36
PCM Receive Start Count—Low Byte
37
PCM Receive Start Count—High Byte
38
TX Gain Control 2
TGA2
TXG2[3:0]
39
RX Gain Control 2
RGA2
RXG2[3:0]
40
TX Gain Control 3
TGA3
TXG3[3:0]
41
RX Gain Control 3
RGA3
RXG3[3:0]
GCI Control
Line Current/Voltage Threshold Interrupt
44
Line Current/Voltage Threshold Interrupt
Control
45–52
Programmable Hybrid Register 1–8
53–58
Reserved
59
FULL2
ILIM
30
43
OPD
ARM[7:0]
21
42
Bit 0
FULL*
FOH[1:0]
PCML
Spark Quenching Control
PCME
0
DCR
ACIM[3:0]
PCMF[1:0]
LVFD*
TXS[7:0]
TXS[1:0]
RXS[7:0]
RXS[1:0]
GCIF[1:0]
B2D
B1D
CVS*
CVM*
CVP*
RG1
GCE
CVT[7:0]*
CVI*
HYB1–8[7:0]
SQ1
SQ0
*Note: Bit is available for Si3019 line-side device only.
Rev. 1.31
59
Si3050 + Si3018/19
Register 1. Control 1
Bit
D7
D6
D5
D4
D3
D2
D1
Name
SR
PWMM[1:0]
PWME
IDL
Type
R/W
R/W
R/W
R/W
D0
Reset settings = 0000_0000
Bit
Name
7
SR
Function
Software Reset.
0 = Enables the DAA for normal operation.
1 = Sets all registers to their reset value.
Note: Bit automatically clears after being set.
6
5:4
60
Reserved
Read returns zero.
PWMM[1:0] Pulse Width Modulation Mode.
Used to select the type of signal output on the call progress AOUT pin.
00 = PWM output is clocked at 16.384 MHz as a delta-sigma data stream. A local density of
1s and 0s tracks the combined transmit and receive signals. Use this setting with the optional
call progress circuit shown in Figure 18 on page 19.
01 = Balanced conventional PWM output signal has high and low portions of the modulated
pulse that are centered on the 16 kHz sample clock.
10 = Conventional PWM output signal returns to logic 0 at regular 32 kHz intervals and rises
at a time in the 32 kHz period proportional to its instantaneous amplitude.
11 = Reserved.
3
PWME
2
Reserved
1
IDL
0
Reserved
Pulse Width Modulation Enable.
0 = Pulse width modulation mode disabled (AOUT).
1 = Enable pulse width modulation mode for the call progress analog output (AOUT). This
mode sums the transmit and receive audio paths and presents this as a CMOS digital-level
output of PWM data. The circuit in Figure 18 on page 19 should be used.
Read returns zero.
Isolation Digital Loopback.
0 = Digital loopback across the isolation barrier is disabled.
1 = Enables digital loopback mode across the isolation barrier. The line-side device must be
enabled and off-hook prior to setting this mode. The data path includes the TX and RX filters.
Read returns zero.
Rev. 1.31
Si3050 + Si3018/19
Register 2. Control 2
Bit
D7
D6
Name
INTE
INTP
Type
R/W
R/W
D5
D4
D3
D2
D1
D0
WDTEN
RDI
HBE
RXE
R/W
R/W
R/W
R/W
Reset settings = 0000_0011
Bit
Name
Function
7
INTE
Interrupt Pin Enable.
0 = The AOUT/INT pin functions as an analog output for call progress monitoring purposes.
1 = The AOUT/INT pin functions as a hardware interrupt pin.
6
INTP
Interrupt Polarity Select.
0 = The AOUT/INT pin, when used in hardware interrupt mode, is active low.
1 = The AOUT/INT pin, when used in hardware interrupt mode, is active high.
5
Reserved
4
WDTEN
3
Reserved
2
RDI
Read returns zero.
Watchdog Timer Enable.
0 = Watchdog timer disabled.
1 = Watchdog timer enabled. When set, this bit can be cleared only by a hardware reset. The
watchdog timer monitors register access. If no register access occurs within a 4 s window, the
DAA is put into an on-hook state. A read or write of a DAA register restarts the watchdog timer
counter. If the watchdog timer times out, the OH bit is cleared, placing the DAA into an
on-hook state. Setting the OH bit places the DAA back into an off-hook state.
Read returns zero.
Ring Detect Interrupt Mode.
This bit operates in conjunction with the RDTM and RDTI bits. This bit selects whether one or
two interrupts are generated for every ring burst.
0 = An interrupt is generated at the beginning of every ring burst.
1 = An interrupt is generated at the beginning and end of every ring burst. The interrupt at the
beginning of the ring burst must be serviced (by writing 0 to the RDTI bit) before the end of the
ring burst in order for both interrupts to occur.
1
HBE
Hybrid Enable.
0 = Disconnects hybrid in transmit path.
1 = Connects hybrid in transmit path.
0
RXE
Receive Enable.
0 = Receive path disabled.
1 = Enables receive path.
Rev. 1.31
61
Si3050 + Si3018/19
Register 3. Interrupt Mask
Bit
D7
D6
D5
D4
D3
D2
D1
D0
Name
RDTM
ROVM
FDTM
BTDM
DODM
LCSOM
TGDM
POLM
Type
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Reset settings = 0000_0000
Bit
Name
Function
7
RDTM
Ring Detect Mask.
0 = A ring signal does not cause an interrupt on the AOUT/INT pin.
1 = A ring signal causes an interrupt on the AOUT/INT pin.
6
ROVM
Receive Overload Mask.
0 = A receive overload does not cause an interrupt on the AOUT/INT pin.
1 = A receive overload causes an interrupt on the AOUT/INT pin.
5
FDTM
Frame Detect Mask.
0 = The ISOcap losing frame lock does not cause an interrupt on the AOUT/INT pin.
1 = The ISOcap losing frame lock causes an interrupt on the AOUT/INT pin.
4
BTDM
Billing Tone Detect Mask.
0 = A detected billing tone does not cause an interrupt on the AOUT/INT pin.
1 = A detected billing tone causes an interrupt on the AOUT/INT pin.
3
DODM
Drop Out Detect Mask.
0 = A line supply dropout does not cause an interrupt on the AOUT/INT pin.
1 = A line supply dropout causes an interrupt on the AOUT/INT pin.
2
LCSOM
Loop Current Sense Overload Mask.
0 = An interrupt does not occur when the LCS bits are all 1s.
1 = An interrupt occurs when the LCS bits are all 1s.
1
TGDM
TIP Ground Detect Mask.
0 = The TGD bit going active does not cause an interrupt on the AOUT/INT pin.
1 = The TGD bit going active causes an interrupt on the AOUT/INT pin.
0
62
POLM
Polarity Reversal Detect Mask (Si3019 line-side only).
This interrupt is generated from bit 7 of the LVS register. When this bit transitions, it indicates
that the polarity of TIP and RING is switched.
0 = A polarity change on TIP and RING does not cause an interrupt on the AOUT/INT pin.
1 = A polarity change on TIP and RING causes an interrupt on the AOUT/INT pin.
Rev. 1.31
Si3050 + Si3018/19
Register 4. Interrupt Source
Bit
D7
D6
D5
D4
D3
D2
D1
D0
Name
RDTI
ROVI
FDTI
BTDI
DODI
LCSOI
TGDI
POLI
Type
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
Reset settings = 0000_0000
Bit
Name
Function
7
RDTI
Ring Detect Interrupt.
0 = A ring signal is not occurring.
1 = A ring signal is detected. If the RDTM bit (Register 3) and INTE bit (Register 2) are set, a
hardware interrupt occurs on the AOUT/INT pin. This bit must be written to a 0 to be cleared.
The RDI bit (Register 2) determines if this bit is set only at the beginning of a ring pulse, or at
the both the beginning and end of a ring pulse. This bit should be cleared after clearing the
PDL bit (Register 6) as powering up the line-side device can cause this interrupt to be triggered.
6
ROVI
Receive Overload Interrupt.
0 = Normal operation.
1 = An excessive input level on the receive pin is detected. If the ROVM bit (Register 3) and
INTE bit (Register 2) are set, a hardware interrupt occurs on the AOUT/INT pin. This bit must
be written to 0 to clear it. This bit is identical in function to the ROV bit (Register 17) and clearing this bit also clears the ROV bit.
5
FDTI
Frame Detect Interrupt.
0 = Frame detect is established on the ISOcap link.
1 = This bit is set when the ISOcap link does not have frame lock. If the FDTM bit (Register 3)
and INTE bit (Register 2) are set, a hardware interrupt occurs on the AOUT/INT pin. When
set, this bit must be written to 0 to be cleared.
4
BTDI
Billing Tone Detect Interrupt.
0 = Normal operation.
1 = The line-side power supply has been disrupted. If the BTDM bit (Register 3) and INTE bit
(Register 2) are set, a hardware interrupt occurs on the AOUT/INT pin. This bit must be written to 0 to clear it.
3
DODI
Drop Out Detect Interrupt.
0 = Normal operation.
1 = The line-side power supply has collapsed. (The DOD bit in Register 19 has fired.) If the
DODM bit (Register 3) and INTE bit (Register 2) are set, a hardware interrupt occurs on the
AOUT/INT pin. This bit must be written to 0 to be cleared. This bit should be cleared after
clearing the PDL bit (Register 6) as powering up the line-side device can cause this interrupt
to be triggered.
2
LCSOI
Loop Current Sense Overload Interrupt.
0 = Normal operation.
1 = The LCS bits have reached max value. If the LCSOM bit (Register 3) and the INTE bit are
set, a hardware interrupt occurs on the AOUT/INT pin. This bit must be written to 0 to clear it.
Note: LCSOI does not necessarily imply that an overcurrent situation has occurred. An overcurrent
situation in the DAA is determined by the status of the OPD bit (Register 19). After the LCSOI
interrupt fires, the OPD bit should be checked to determine if an overcurrent situation exists.
Rev. 1.31
63
Si3050 + Si3018/19
Bit
Name
Function
1
TGDI
TIP Ground Detect Interrupt.
This bit is reverse logic as compared to the TGD bit.
0 = The CO has not grounded TIP causing current to flow.
1 = The CO has grounded TIP, causing current to flow. Once set, this bit must be written to 0
to clear it. If the TDGM bit (Register 3) and INTE bit (Register 3) are set, a hardware interrupt
occurs on the AOUT/INT pin. To clear the interrupt, write this bit to 0.
0
POLI
Polarity Reversal Detect Interrupt (Si3019 line-side only).
0 = Bit 7 of the LVS register has not changed states.
1 = Bit 7 of the LVS register has transitioned from 0 to 1, or from 1 to 0, indicating the polarity
of TIP and RING is switched. If the POLM bit (Register 3) and INTE bit (Register 2) are set, a
hardware interrupt occurs on the AOUT/INT pin. To clear the interrupt, write this bit to 0.
64
Rev. 1.31
Si3050 + Si3018/19
Register 5. DAA Control 1
Bit
D7
D6
D5
Name
RDTN
Type
R
D4
D3
D2
D1
D0
RDTP
ONHM
RDT
OH
R
R/W
R
R/W
Reset settings = 0000_0000
Bit
Name
Function
7
Reserved
6
RDTN
Ring Detect Signal Negative.
0 = No negative ring signal is occurring.
1 = A negative ring signal is occurring.
5
RDTP
Ring Detect Signal Positive.
0 = No positive ring signal is occurring.
1 = A positive ring signal is occurring.
4
Reserved
3
ONHM
2
RDT
1
Reserved
0
OH
Read returns zero.
Read returns zero.
On-Hook Line Monitor.
0 = Normal on-hook mode.
1 = Enables low-power on-hook monitoring mode allowing the host to receive line activity
without going off-hook. This mode is used for caller-ID detection.
Ring Detect.
0 = Reset 5 seconds after last positive ring is detected or when the system executes an
off-hook. Only a positive ring sets this bit when RFWE = 0. When RFWE = 1, either a positive
or negative ring sets this bit.
1 = Indicates a ring is occurring.
Read returns zero.
Off-Hook.
0 = Line-side device on-hook.
1 = Causes the line-side device to go off-hook.
Rev. 1.31
65
Si3050 + Si3018/19
Register 6. DAA Control 2
Bit
D7
D6
D5
D4
D3
Name
PDL
PDN
Type
R/W
R/W
D2
D1
D0
Reset settings = 0001_0000
Bit
Name
Function
7:5
Reserved
4
PDL
Powerdown Line-Side Device.
0 = Normal operation. Program the clock generator before clearing this bit.
1 = Places the line-side device in lower power mode.
3
PDN
Powerdown System-Side Device.
0 = Normal operation.
1 = Powers down the system-side device. A pulse on RESET is required to restore normal
operation.
2:0
Reserved
Read returns zero.
Read returns zero.
Register 7. Sample Rate Control
Bit
D7
D6
D5
D4
D3
Name
HSSM
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
66
Bit
Name
7:4
Reserved
3
HSSM
2:0
Reserved
Function
Read returns zero.
High-Speed Sampling Mode.
0 = Sample Rate is 8 kHz.
1 = Sample Rate is 16 kHz. The PCM or the GCI highway continues to be at 8 kHz; thus,
twice as many samples are generated per device timeslot. Samples are transmitted in adjacent timeslots.
Read returns zero.
Rev. 1.31
Si3050 + Si3018/19
Register 8-9. Reserved
Bit
D7
D6
D5
D4
D3
D2
D1
D0
D2
D1
D0
Name
Type
Reset settings = 0000_0000
Bit
Name
7:0
Reserved
Function
Read returns zero.
Register 10. DAA Control 3
Bit
D7
D6
D5
D4
D3
Name
DDL
Type
R/W
Reset settings = 0000_0000
Bit
Name
7:1
Reserved
0
DDL
Function
Read returns zero.
Digital Data Loopback.
0 = Normal operation.
1 = Takes data received on DRX and loops it back out to DTX before the TX and RX filters.
Output data is identical to the input data.
Rev. 1.31
67
Si3050 + Si3018/19
Register 11. System-Side and Line-Side Device Revision
Bit
D7
D6
D5
D4
D3
D2
D1
Name
LSID[3:0]
REVA[3:0]
Type
R
R
D0
Reset settings = xxxx_xxxx
Bit
Name
7:4
LSID[3:0]
Function
Line-Side ID Bits.
These four bits will always read one of the following values, depending on which line-side
device is used:
Device
LSID[3:0]
Si3018
0001
Si3019
0011
3:0
REVA[3:0] System-Side Revision.
Four-bit value indicating the revision of the Si3050 (system-side) device.
Register 12. Line-Side Device Status
Bit
D7
D6
D5
D4
D3
D2
Name
FDT
LCS[4:0]
Type
R
R
Reset settings = 0000_0000
Bit
Name
7
Reserved
6
FDT
5
Reserved
Read returns zero.
4:0
LCS[4:0]
Off-Hook Loop Current Monitor (3.3 mA/bit).
00000 = Loop current is less than required for normal operation.
00100 = Minimum loop current for normal operation.
11111 = Loop current is >127 mA, and an overload condition may exist.
68
Function
Read returns zero.
Frame Detect.
0 = Indicates ISOcap link has not established frame lock.
1 = Indicates ISOcap link frame lock is established.
Rev. 1.31
D1
D0
Si3050 + Si3018/19
Register 13. Line-Side Device Revision
Bit
D7
D6
D5
D4
D3
Name
1
REVB[3:0]
Type
R
R
D2
D1
D0
D2
D1
D0
Reset settings = xxxx_xxxx
Bit
Name
Function
7
Reserved
Read returns zero.
6
Reserved
This bit always reads a one.
5:2
REVB[3:0] Line-Side Device Revision.
Four-bit value indicating the revision of the line-side device.
1:0
Reserved
Read returns zero.
Register 14. DAA Control 4
Bit
D7
D6
D5
D4
D3
Name
RPOL
Type
R/W
Reset settings = 0000_0000
Bit
Name
7:2
Reserved
1
RPOL
Function
Read returns zero.
Ring Detect Polarity.
0 = The RGDT pin is active low.
1 = The RGDT pin is active high.
0
Reserved
Read returns zero.
Rev. 1.31
69
Si3050 + Si3018/19
Register 15. TX/RX Gain Control 1
Bit
D7
D6
D5
D4
D3
Name
TXM
RXM
Type
R/W
R/W
Reset settings = 0000_0000
70
Bit
Name
7
TXM
6:4
Reserved
3
RXM
2:0
Reserved
Function
Transmit Mute.
0 = Transmit signal is not muted.
1 = Mutes the transmit signal.
Read returns zero.
Receive Mute.
0 = Receive signal is not muted.
1 = Mutes the receive signal.
Read returns zero.
Rev. 1.31
D2
D1
D0
Si3050 + Si3018/19
Register 16. International Control 1
Bit
D7
D6
D5
D4
D3
D2
D1
D0
Name
OHS
IIRE
RZ
RT
Type
R/W
R/W
R/W
R/W
Reset settings = 0000_0000
Bit
7
6
5
Name
Function
Reserved These bits may be written to a zero or one.
OHS
On-Hook Speed.
This bit, in combination with the OHS2 bit (Register 31) and the SQ[1:0] bits (Register 59), sets
the amount of time for the line-side device to go on-hook. The on-hook speeds specified are
measured from the time the OH bit is cleared until loop current equals zero.
OHS
OHS2
SQ[1:0]
Mean On-Hook Speed
0
0
00
Less than 0.5 ms
0
1
00
3 ms ±10% (meets ETSI standard)
1
X
11
26 ms ±10% (meets Australia spark quenching spec)
Reserved These bits may be written to a zero or one.
4
IIRE
3:2
Reserved
1
RZ
Ringer Impedance.
0 = Maximum (high) ringer impedance.
1 = Synthesized ringer impedance used to satisfy a maximum ringer impedance specification
in countries, such as Poland, South Africa, and Slovenia.
0
RT
Ringer Threshold Select.
This bit, in combination with the RT2 bit, is used to satisfy country requirements on ring detection. Signals below the lower level do not generate a ring detection; signals above the upper
level are guaranteed to generate a ring detection.
RT
RT2
RT Lower level
RT Upper level
0
0
13.5 Vrms
16.5 Vrms
0
1
Reserved, do not use this setting.
23.65 VRMS
1
0
19.35 Vrms
1
1
40.5 Vrms
49.5 VRMS
IIR Filter Enable.
0 = FIR filter enabled for transmit and receive filters. (See Figures 7–10 on page 15.)
1 = IIR filter enabled for transmit and receive filters. (See Figures 11–16 on page 16.)
Read returns zero.
Rev. 1.31
71
Si3050 + Si3018/19
Register 17. International Control 2
Bit
D7
D6
D5
D4
D3
D2
D1
D0
Name
CALZ
MCAL
CALD
RT2
OPE
BTE
ROV
BTD
Type
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R
Reset settings = 0000_0000
Bit
7
Name
CALZ
6
MCAL
5
CALD
4
RT2
3
OPE
2
BTE
72
Function
Clear ADC Calibration.
0 = Normal operation.
1 = Clears the existing ADC calibration data. This bit must be written back to 0 after being set.
Manual ADC Calibration.
0 = No calibration.
1 = Initiate manual ADC calibration.
Auto-Calibration Disable.
0 = Enable auto-calibration.
1 = Disable auto-calibration.
Ringer Threshold Select 2.
This bit, in combination with the RT bit, is used to satisfy country requirements on ring detection. Signals below the lower level do not generate a ring detection; signals above the upper
level are guaranteed to generate a ring detection.
RT
RT2
RT Lower level
RT Upper level
0
0
13.5 Vrms
16.5 Vrms
0
1
Reserved, do not use this setting.
23.65 VRMS
1
0
19.35 Vrms
1
1
40.5 Vrms
49.5 VRMS
Overload Protect Enable.
0 = Disabled.
1 = Enabled.
The OPE bit should always be cleared before going off-hook.
Billing Tone Detect Enable.
The DAA can detect events, such as billing tones, that can cause a disruption in the line-side
power supply. When this bit is set, the device will maintain off-hook during such events. If a
billing tone is detected, the BTD bit (Register 17, bit 0) is set to indicate the event. Writing this
bit to zero clears the BTD bit.
0 = Billing tone detection disabled. The BTD bit is not functional.
1 = Billing tone detection enabled. The BTD bit is not functional.
Rev. 1.31
Si3050 + Si3018/19
Bit
1
Name
ROV
0
BTD
Function
Receive Overload.
This bit is set when the receive input has an excessive input level (i.e., receive pin goes below
ground). Writing a 0 to this location clears this bit and the ROVI bit (Register 4, bit 6).
0 = Normal receive input level.
1 = Excessive receive input level.
Billing Tone Detected.
This bit is set if an event, such as a billing tone, causes a disruption in the line-side power
supply. Writing a zero to BTE clears this bit.
0 = No billing tone detected.
1 = Billing tone detected.
Register 18. International Control 3
Bit
D7
D6
D5
D4
D3
D2
D1
Name
RFWE
Type
R/W
D0
Reset settings = 0000_0000
Bit
7:3
2
1
0
Name
Function
Reserved Read returns zero.
Reserved This bit may be written to a zero or one.
RFWE
Ring Detector Full-Wave Rectifier Enable.
When RNGV (Register 24) is disabled, this bit controls the ring detector mode and the assertion of the RGDT pin. When RNGV is enabled, this bit configures the RGDT pin to either follow
the ringing signal detected by the ring validation circuit, or to follow an unqualified ring detect
one-shot signal initiated by a ring-threshold crossing and terminated by a fixed counter timeout
of approximately 5 seconds.
RNGV
RFWE
RGDT
0
0
Half-Wave
0
1
Full-Wave
1
0
Validated Ring Envelope
1
1
Ring Threshold Crossing One-Shot
Reserved Read returns zero.
Rev. 1.31
73
Si3050 + Si3018/19
Register 19. International Control 4
Bit
D7
D6
D5
D4
D3
D2
D1
D0
Name
OVL
DOD
OPD
Type
R
R
R
Reset settings = 0000_0000
Bit
Name
7:3
Reserved
2
OVL
Receive Overload Detect.
This bit has the same function as ROV (Register 17), but clears itself after the overload is
removed. See “5.22.Receive Overload Detection” on page 33. This bit is only masked by the
off-hook counter and is not affected by the BTE bit.
0 = Normal receive input level.
1 = Excessive receive input level.
1
DOD
Recal/Dropout Detect.
When the line-side device is off-hook, it is powered from the line itself. This bit will read 1
when loop current is not flowing. For example, if this line-derived power supply collapses,
such as when the line is disconnected, this bit is set to 1. Additionally, when on-hook, and the
line-side device is enabled, this bit is set to 1.
0 = Normal operation.
1 = Line supply dropout detected when off-hook.
0
OPD
Overload Protect Detect.
This bit is used to indicate that the DAA has detected a loop current overload. The detector firing threshold depends on the setting of the ILIM bit (Register 26).
OPD
ILIM
Overcurrent Threshold
Overcurrent Status
0
0
160 mA
No overcurrent condition exists
0
1
60 mA
No overcurrent condition exists
1
0
160 mA
Overcurrent condition has been detected
1
1
60 mA
Overcurrent condition has been detected
74
Function
Read returns zero.
Rev. 1.31
Si3050 + Si3018/19
Register 20. Call Progress RX Attenuation
Bit
D7
D6
D5
D4
D3
Name
ARM[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
Name
Function
7:0
ARM[7:0]
AOUT Receive Path Attenuation.
When decremented from the default setting, these bits linearly attenuate the AOUT
receive path signal used for call progress monitoring. Setting the bits to all 0s mutes the
AOUT receive path.
Attenuation = 20 log(ARM[7:0]/64)
1111_1111 = +12 dB (gain)
0111_1111 = +6 dB (gain)
0100_0000 = 0 dB
0010_0000 = –6 dB (attenuation)
0001_0000 = –12 dB
...
0000_0000 = Mute
Register 21. Call Progress TX Attenuation
Bit
D7
D6
D5
D4
D3
Name
ATM[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
Name
7:0
ATM[7:0]
Function
AOUT Transmit Path Attenuation.
When decremented from the default settings, these bits linearly attenuate the AOUT transmit path signal used for call progress monitoring. Setting the bits to all 0s mutes the AOUT
transmit path.
Attenuation = 20 log(ATM[7:0]/64)
1111_1111 = +12 dB (gain)
0111_1111 = +6 dB (gain)
0100_0000 = 0 dB
0010_0000 = –6 dB (attenuation)
0001_0000 = –12 dB
...
0000_0000 = Mute
Rev. 1.31
75
Si3050 + Si3018/19
Register 22. Ring Validation Control 1
Bit
D7
D6
D5
D4
D3
D2
Name
RDLY[1:0]
RMX[5:0]
Type
R/W
R/W
D1
D0
Reset settings = 1001_0110
Bit
Name
Function
7:6
RDLY[1:0]
Ring Delay Bits 1 and 0.
These bits, in combination with the RDLY[2] bit (Register 23), set the amount of time
between when a ring signal is validated and when a valid ring signal is indicated.
RDLY[2]
RDLY[1:0]
Delay
0
00
0 ms
0
01
256 ms
0
10
512 ms
...
1
11
1792 ms
5:0
RMX[5:0]
Ring Assertion Maximum Count.
These bits set the maximum ring frequency for a valid ring signal within a 10% margin of
error. During ring qualification, a timer is loaded with the RAS[5:0] field upon a TIP/RING
event and decrements at a regular rate. When a subsequent TIP/RING event occurs, the
timer value is compared to the RMX[5:0] field and if it exceeds the value in RMX[5:0] then
the frequency of the ring is too high and the ring is invalidated. The difference between
RAS[5:0] and RMX[5:0] identifies the minimum duration between TIP/RING events to qualify as a ring, in binary-coded increments of 2.0 ms (nominal). A TIP/RING event typically
occurs twice per ring tone period. At 20 Hz, TIP/RING events would occur every 1/
(2 x 20 Hz) = 25 ms. To calculate the correct RMX[5:0] value for a frequency range [f_min,
f_max], the following equation should be used:
1
RMX  5:0   RAS  5:0  – --------------------------------------------- RMX  RAS
2  f_max  2 ms
To compensate for error margin and ensure a sufficient ring detection window, it is recommended that the calculated value of RMX[5:0] be incremented by 1.
76
Rev. 1.31
Si3050 + Si3018/19
Register 23. Ring Validation Control 2
Bit
D7
D6
D5
D4
D3
D2
D1
Name
RDLY[2]
RTO[3:0]
RCC[2:0]
Type
R/W
R/W
R/W
D0
Reset settings = 0010_1101
Bit
Name
Function
7
RDLY[2]
Ring Delay Bit 2.
This bit, in combination with the RDLY[1:0] bits (Register 22), sets the amount of time
between when a ring signal is validated and when a valid ring signal is indicated.
RDLY[2]
RDLY[1:0]
Delay
0
00
0 ms
0
01
256 ms
0
10
512 ms
...
1
11
1792 ms
6:3
RTO[3:0]
Ring Timeout.
These bits set when a ring signal is determined to be over after the most recent ring
threshold crossing.
RTO[3:0]
Ring Timeout
0000
DO NOT USE THIS SETTING
0001
128 ms
0010
256 ms
...
1111
1920 ms
2:0
RCC[2:0]
Ring Confirmation Count.
These bits set the amount of time that the ring frequency must be within the tolerances set
by the RAS[5:0] bits and the RMX[5:0] bits to be classified as a valid ring signal.
RCC[2:0]
Ring Confirmation Count Time
000
100 ms
001
150 ms
010
200 ms
011
256 ms
100
384 ms
101
512 ms
110
640 ms
111
1024 ms
Rev. 1.31
77
Si3050 + Si3018/19
Register 24. Ring Validation Control 3
Bit
D7
D6
D5
D4
D3
D2
Name
RNGV
RAS[5:0]
Type
R/W
R/W
D1
D0
Reset settings = 0001_1001
Bit
Name
7
RNGV
Function
Ring Validation Enable.
0 = Ring validation feature is disabled.
1 = Ring validation feature is enabled in both normal operating mode and low-power
mode.
6
Reserved
This bit must always be written to 0.
5:0
RAS[5:0]
Ring Assertion Time.
These bits set the minimum ring frequency for a valid ring signal. During ring qualification,
a timer is loaded with the RAS[5:0] field upon a TIP/RING event and decrements at a regular rate. If a second or subsequent TIP/RING event occurs after the timer has timed out
then the frequency of the ring is too low and the ring is invalidated. The difference between
RAS[5:0] and RMX[5:0] identifies the minimum duration between TIP/RING events to qualify as a ring, in binary-coded increments of 2.0 ms (nominal). A TIP/RING event typically
occurs twice per ring tone period. At 20 Hz, TIP/RING events would occur every
1/(2 x 20 Hz) = 25 ms. To calculate the correct RAS[5:0] value for a frequency range
[f_min, f_max], the following equation should be used:
1
RAS  5:0   ------------------------------------------2  f_min  2 ms
Register 25. Resistor Calibration
Bit
D7
D6
D5
Name
RCALS
RCALM
RCALD
Type
R
R/W
R/W
D4
D3
D2
D1
D0
RCAL[3:0]
R/W
Reset settings = xx0x_xxxx
Bit
Name
7
RCALS
Resistor Auto Calibration.
0 = Resistor calibration is not in progress.
1 = Resistor calibration is in progress.
6
RCALM
Manual Resistor Calibration.
0 = No calibration.
1 = Initiate manual resistor calibration. (After a manual calibration has been initiated, this bit
must be cleared within 1 ms.)
5
RCALD
Resistor Calibration Disable.
0 = Internal resistor calibration enabled.
1 = Internal resistor calibration disabled.
4
Reserved
3:0
78
Function
This bit can be written to a 0 or 1.
RCAL[3:0] Always write back the value read.
Rev. 1.31
Si3050 + Si3018/19
Register 26. DC Termination Control
Bit
D7
D6
D5
D4
Name
DCV[1:0]
MINI[1:0]
Type
R/W
R/W
D3
D2
D1
D0
0
0
ILIM
DCR
R/W
R/W
Reset settings = 0000_0000
Bit
Name
Function
7:6
DCV[1:0]
TIP/RING Voltage Adjust.
These bits adjust the voltage on the DCT pin of the line-side device, which affects the TIP/
RING voltage on the line. Low-voltage countries should use a lower TIP/RING voltage. Raising the TIP/RING voltage can improve signal headroom.
DCV[1:0] DCT Pin Voltage
00
3.1 V
01
3.2 V
10
3.35 V
11
3.5 V
5:4
MINI[1:0]
Minimum Operational Loop Current.
Adjusts the minimum loop current at which the DAA can operate. Increasing the minimum
operational loop current can improve signal headroom at a lower TIP/RING voltage.
MINI[1:0] Min Loop Current
00
10 mA
01
12 mA
10
14 mA
11
16 mA
3:2
Reserved
These bits must always be written to 0.
1
ILIM
Current Limiting Enable.
0 = Current limiting mode disabled.
1 = Current limiting mode enabled. This mode limits loop current to a maximum of 60 mA per
the TBR21 standard.
0
DCR
DC Impedance Selection.
0 = 50  dc termination is selected. This mode should be used for all standard applications.
1 = 800  dc termination is selected.
Rev. 1.31
79
Si3050 + Si3018/19
Register 27. Reserved
Bit
D7
D6
D5
D4
D3
D2
D1
D0
D2
D1
D0
Name
Type
Reset settings = xxxx_xxxx
Bit
Name
7:0
Reserved
Function
Do not write to these register bits.
Register 28. Loop Current Status
Bit
D7
D6
D5
D4
D3
Name
LCS2[7:0]
Type
R
Reset settings = 0000_0000
Bit
7:0
Name
Function
LCS2[7:0] Loop Current Status.
Eight-bit value returning the loop current. Each bit represents 1.1 mA of loop current.
0000_0000 = Loop current is less than required for normal operation.
Register 29. Line Voltage Status
Bit
D7
D6
D5
D4
D3
Name
LVS[7:0]
Type
R
D2
D1
D0
Reset settings = 0000_0000
Bit
Name
7:0
LVS[7:0]
Function
Line Voltage Status.
Eight-bit value returning the loop voltage. Each bit represents 1 V of loop voltage. This register operates in on- and off-hook modes. Bit seven of this register indicates the polarity of the
TIP/RING voltage. When this bit changes state, it indicates that a polarity reversal has
occurred. The value returned is represented in 2s complement format.
0000_0000 = No line is connected.
80
Rev. 1.31
Si3050 + Si3018/19
Register 30. AC Termination Control
Bit
D7
D6
D5
D4
D3
D2
D1
Name
FULL2
ACIM[3:0]
Type
R/W
R/W
D0
Reset settings = 0000_0000
Bit
Name
7:6
Reserved
Read returns zero.
5
Reserved
This bit may be written to a zero or one.
4
FULL2
3:0
Function
Enhanced Full Scale (2x) Transmit and Receive Mode.
0 = Default
1 = Transmit/Receive 2x Full Scale
This bit changes the full scale of the ADC and DAC from 0 min to +6 dBm into 600  load (or
1.5 dBV into all reference impedances). When this bit is set, the DCV[1:0] bits (Register 26)
should be set to all 1s to avoid distortion at low loop currents.
ACIM[3:0] AC Impedance Selection.
The off-hook ac termination is selected from the following:
0000 = 600 
0001 = 900 
0010 = 270  + (750 || 150 nF) and 275 + (780 || 150 nF)
0011 = 220  + (820 || 120 nF) and 220  + (820 || 115 nF)
0100 = 370  + (620  || 310 nF)
0101 = 320  + (1050  || 230 nF)
0110 = 370  + (820  || 110 nF)
0111 = 275  + (780  || 115 nF)
1000= 120  + (820  || 110 nF)
1001 = 350  + (1000  || 210 nF)
1010 = 200  + (680  || 100 nF)
1011 = 600  + 2.16 µF
1100 = 900  + 1 µF
1101 = 900  + 2.16 µF
1110 = 600  + 1 µF
1111 = Global impedance
Rev. 1.31
81
Si3050 + Si3018/19
Register 31. DAA Control 5
Bit
D7
D6
D5
Name
FULL
FOH[1:0]
Type
R/W
RW
D4
D3
D2
D1
D0
0
OHS2
0
FILT
LVFD
R/W
R/W
R/W
Reset settings = 0010_0000
Bit
Name
Function
7
FULL
Full Scale Transmit and Receive Mode.
0 = Default.
1 = Transmit/receive full scale.
This bit changes the full scale of the ADC and DAC from 0 dBm min to +3.2 dBm into a 600 
load (or 1 dBV into all reference impedances). When this bit is set, the DCV[1:0] bits
(Register 26) should be set to all 1s. The MINI[1:0] bits also should be set to all 0s. This ensures
correct operation of the full scale mode.
6:5
4
3
2
FOH[1:0] Fast Off-Hook Selection.
These bits determine the length of the off-hook counter. The default setting is 128 ms.
00 = 512 ms
01 = 128 ms
10 = 64 ms
11 = 8 ms
Reserved Always write these bits to zero.
OHS2
On-Hook Speed 2.
This bit, in combination with the OHS bit (Register 16) and the SQ[1:0] bits on-hook speeds
specified are measured from the time the OH bit is cleared until loop current equals zero.
OHS
OHS2
SQ[1:0]
Mean On-Hook Speed
0
0
00
Less than 0.5 ms
0
1
00
3 ms ±10% (meets ETSI standard)
1
X
11
26 ms ±10% (meets Australia spark quenching spec)
Reserved Always write these bits to zero.
1
FILT
Filter Pole Selection.
0 = The receive path has a low –3 dBFS corner at 5 Hz.
1 = The receive path has a low –3 dBFS corner at 200 Hz.
0
LVFD
Line Voltage Force Disable (Si3019 line-side only).
0 = Normal operation.
1 = The circuitry that forces the LVS register (Register 29) to all 0s at 3 V or less is disabled. The
LVS register may display unpredictable values at voltages between 0 to 2 V. All 0s are displayed
if the line voltage is 0 V.
82
Rev. 1.31
Si3050 + Si3018/19
Register 32. Ground Start Control
Bit
D7
D6
D5
D4
D3
D2
D1
D0
Name
TGD
TGDE
RG
Type
R
W
W
Reset settings = 0000_0x11
Bit
Name
Function
7:3
Reserved
Read returns zero.
2
TGD
TIP Ground Detect.
0 = The CO has grounded TIP, causing current to flow. When current ceases to flow, this bit
returns to a one.
1 = The CO has not grounded TIP causing current to flow.
1
TGDE
TIP Ground Detect Enable.
0 = The external relay connecting TIP to an isolated supply is closed, enabling current to flow
in TIP if the CO grounds TIP.
1 = The external relay connecting TIP to an isolated supply is open. In this state, the DAA is
unable to determine if the CO has grounded TIP.
0
RG
Ring Ground.
0 = The external relay connecting RING to ground is closed, causing current to flow in RING.
1 = The external relay connecting RING to ground is open, not allowing current to flow in
RING.
Rev. 1.31
83
Si3050 + Si3018/19
Register 33. PCM/SPI Mode Select
Bit
D7
Name
PCML
Type
R/W
D6
R/W
D5
D4
D3
D2
D1
D0
PCME
PCMF[1:0]
0
PHCF
TRI
R/W
R/W
R/W
R/W
R/W
Reset settings = 0000_0000
Bit
Name
7
PCML
PCM Analog Loopback.
0 = Normal operation.
1 = Enables analog data to be received from the line, converted to digital data and transmitted across the ISOcap link. The data passes through the RX filter and is looped back
through the TX filter and is transmitted back out to the line.
5
PCME
PCM Enable (Registers 34–37 should be set before PCM transfers are enabled).
0 = Disable PCM transfers.
1 = Enable PCM transfers.
4:3
PCMF[1:0]
PCM Data Format.
00 = A-Law. Signed magnitude data format (refer to Table 22 on page 44).
01 = µ-Law. Signed magnitude data format (refer to Table 21 on page 43).
10 = 8-bit linear. The top 8-bits of the 16-bit linear signal are transferred, and the bottom
8-bits are discarded (2s complement data format).
11 = 16-bit linear (2s complement data format).
2
Reserved
Always write this bit to zero.
1
PHCF
0
TRI
84
Function
PCM Highway Clock Format.
0 = 1 PCLK per data bit.
1 = 2 PCLKs per data bit.
Tri-state Bit 0.
0 = Tri-state bit 0 on positive edge of PCLK.
1 = Tri-state bit 0 on negative edge of PCLK.
Rev. 1.31
Si3050 + Si3018/19
Register 34. PCM Transmit Start Count—Low Byte
Bit
D7
D6
D5
D4
D3
Name
TXS[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
Name
Function
7:0
TXS[7:0]
PCM Transmit Start Count.
PCM Transmit Start Count equals the number of PCLKs following FSYNC before data
transmission begins.
Register 35. PCM Transmit Start Count—High Byte
Bit
D7
D6
D5
D4
D3
D2
D1
D0
Name
TXS[1:0]
Type
R/W
Reset settings = 0000_0000
Bit
Name
Function
7:2
Reserved
Read returns zero.
1:0
TXS[1:0]
PCM Transmit Start Count.
PCM Transmit Start Count equals the number of PCLKs following FSYNC before data
transmission begins.
Register 36. PCM Receive Start Count—Low Byte
Bit
D7
D6
D5
D4
D3
Name
RXS[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
Name
7:0
RXS[7:0]
Function
PCM Receive Start Count.
PCM Receive Start Count equals the number of PCLKs following FSYNC before data
reception begins.
Rev. 1.31
85
Si3050 + Si3018/19
Register 37. PCM Receive Start Count—High Byte
Bit
D7
D6
D5
D4
D3
D2
D1
D0
Name
RXS[1:0]
Type
R/W
Reset settings = 0000_0000
Bit
Name
Function
7:2
Reserved
Read returns zero.
1:0
RXS[1:0]
PCM Receive Start Count.
PCM Receive Start Count equals the number of PCLKs following FSYNC before data
reception begins.
Register 38. TX Gain Control 2
Bit
D7
D6
D5
D4
D3
D2
D1
Name
TGA2
TXG2[3:0]
Type
R/W
R/W
D0
Reset settings = 0000_0000
Bit
Name
7:5
Reserved
4
TGA2
3:0
86
Function
Read returns zero.
Transmit Gain or Attenuation 2.
0 = Incrementing the TXG2[3:0] bits results in gaining up the transmit path.
1 = Incrementing the TXG2[3:0] bits results in attenuating the transmit path.
TXG2[3:0] Transmit Gain 2.
Each bit increment represents 1 dB of gain or attenuation, up to a maximum of +12 dB and
–15 dB respectively.
For example:
TGA2
TXG2[3:0]
Result
X
0000
0 dB gain or attenuation is applied to the transmit path.
0
0001
1 dB gain is applied to the transmit path.
0
:
0
11xx
12 dB gain is applied to the transmit path.
1
0001
1 dB attenuation is applied to the transmit path.
1
:
1
1111
15 dB attenuation is applied to the transmit path.
Rev. 1.31
Si3050 + Si3018/19
Register 39. RX Gain Control 2
Bit
D7
D6
D5
D4
D3
D2
D1
Name
RGA2
RXG2[3:0]
Type
R/W
R/W
D0
Reset settings = 0000_0000
Bit
Name
7:5
Reserved
4
RGA2
3:0
Function
Read returns zero.
Receive Gain or Attenuation 2.
0 = Incrementing the RXG2[3:0] bits results in gaining up the receive path.
1 = Incrementing the RXG2[3:0] bits results in attenuating the receive path.
RXG2[3:0] Receive Gain 2.
Each bit increment represents 1 dB of gain or attenuation, up to a maximum of +12 dB and
–15 dB respectively.
For example:
RGA2
RXG2[3:0]
Result
X
0000
0 dB gain or attenuation is applied to the receive path.
0
0001
1 dB gain is applied to the receive path.
0
:
0
11xx
12 dB gain is applied to the receive path.
1
0001
1 dB attenuation is applied to the receive path.
1
:
1
1111
15 dB attenuation is applied to the receive path.
Rev. 1.31
87
Si3050 + Si3018/19
Register 40. TX Gain Control 3
Bit
D7
D6
D5
D4
D3
D2
D1
Name
TGA3
TXG3[3:0]
Type
R/W
R/W
D0
Reset settings = 0000_0000
Bit
Name
7:5
Reserved
4
TGA3
3:0
88
Function
Read returns zero.
Transmit Gain or Attenuation 3.
0 = Incrementing the TGA3[3:0] bits results in gaining up the transmit path.
1 = Incrementing the TGA3[3:0] bits results in attenuating the transmit path.
TXG3[3:0] Transmit Gain 3.
Each bit increment represents 0.1 dB of gain or attenuation, up to a maximum of 1.5 dB.
For example:
TGA3
TXG3[3:0]
Result
X
0000
0 dB gain or attenuation is applied to the transmit path.
0
0001
0.1 dB gain is applied to the transmit path.
0
:
0
1111
1.5 dB gain is applied to the transmit path.
1
0001
0.1 dB attenuation is applied to the transmit path.
1
:
1
1111
1.5 dB attenuation is applied to the transmit path.
Rev. 1.31
Si3050 + Si3018/19
Register 41. RX Gain Control 3
Bit
D7
D6
D5
D4
D3
D2
D1
Name
RGA3
RXG3[3:0]
Type
R/W
R/W
D0
Reset settings = 0000_0000
Bit
Name
7:5
Reserved
4
RGA3
3:0
Function
Read returns zero.
Receive Gain or Attenuation 2.
0 = Incrementing the RXG3[3:0] bits results in gaining up the receive path.
1 = Incrementing the RXG3[3:0] bits results in attenuating the receive path.
RXG3[3:0] Receive Gain 3.
Each bit increment represents 0.1 dB of gain or attenuation, up to a maximum of 1.5 dB.
For example:
RGA3
RXG3[3:0]
Result
X
0000
0 dB gain or attenuation is applied to the receive path.
0
0001
0.1 dB gain is applied to the receive path.
0
:
0
1111
1.5 dB gain is applied to the receive path.
1
0001
0.1 dB attenuation is applied to the receive path.
1
:
1
1111
1.5 dB attenuation is applied to the receive path.
Rev. 1.31
89
Si3050 + Si3018/19
Register 42. GCI Control
Bit
D7
D6
D5
D4
D3
D2
D1
D0
Name
GCIF[1:0]
B2D
B1D
Type
R/W
R/W
R/W
Reset settings = 0000_0000
Bit
Name
7:4
Reserved
Read returns zero.
3:2
GCIF[1:0]
GCI Data Format.
00 = A-Law.
01 = µ-Law.
10 = 8-bit linear. The top 8-bits of the 16-bit linear signal are transferred, and the bottom 8-bits
are discarded.
11 = 16-bit linear. B1 and B2 channels are used for the 16-bits of data. Regardless of whether
the DAA is set to transmit and receive in the B1 or B2 channel, both channels are used to
send and receive the 16-bit linear data.
1
B2D
Channel B2 Enable.
0 = Channel B2 transfers are disabled.
1 = Channel B2 transfers are enabled. If 16-bit linear data format is chosen, disabling the B2
channel results in only the top 8 bits of line data being sent and received in the B1 channel.
0
B1D
Channel B1 Enable.
0 = Channel B1 transfers are disabled.
1 = Channel B1 transfers are enabled. If 16-bit linear data format is chosen, disabling the B1
channel results in only the bottom 8 bits of line data being sent and received in the B2 channel.
90
Function
Rev. 1.31
Si3050 + Si3018/19
Register 43. Line Current/Voltage Threshold Interrupt (Si3019 line-side only)
Bit
D7
D6
D5
D4
D3
Name
CVT[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
Name
7:0
CVT[7:0]
Function
Current/Voltage Threshold.
These bits determine the threshold at which an interrupt is generated from either the LCS or
LVS register. This interrupt can be generated to occur when the line current or line voltage
rises above or drops below the value in the CVT[7:0] register.
Register 44. Line Current/Voltage Threshold Interrupt Control (Si3019 line-side only)
Bit
D7
D6
D5
D4
D3
D2
D1
D0
Name
CVI
CVS
CVM
CVP
Type
R/W
R/W
R/W
R/W
Reset settings = 0000_0000
Bit
Name
7:4
Reserved
3
CVI
Function
Read returns zero.
Current/Voltage Interrupt.
0 = The current/voltage threshold has not been crossed.
1 = The current/voltage threshold is crossed. If the CVM and INTE bits are set, a hardware
interrupt occurs on the AOUT/INT pin. Once set, this bit must be written to 0 to be cleared.
2
CVS
Current/Voltage Select.
0 = The line current shown in the LCS2 register is used to generate an interrupt.
1 = The line voltage shown in the LVS register is used to generate an interrupt.
1
CVM
Current/Voltage Interrupt Mask.
0 = The current/voltage threshold being triggered does not cause a hardware interrupt on the
AOUT/INT pin.
1 = The current/voltage threshold being triggered causes a hardware interrupt on the
AOUT/INT pin.
0
CVP
Current/Voltage Interrupt Polarity.
0 = The current/voltage threshold is triggered by the absolute value of the number in either
the LCS2 or LVS register falling below the value in the CVT[7:0] register.
1 = The current/voltage threshold is triggered by the absolute value of the number in either
the LCS2 or LVS register rising above the value in the CVT[7:0] register.
Rev. 1.31
91
Si3050 + Si3018/19
Register 45. Programmable Hybrid Register 1
Bit
D7
D6
D5
D4
D3
Name
HYB1[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
7:0
Name
Function
HYB1[7:0] Programmable Hybrid Register 1.
These bits can be programmed with a coefficient value to adjust the hybrid response to
reduce near-end echo. This register represents the first tap in the eight-tap filter. When this
register is set to all 0s, this filter stage does not have an effect on the hybrid response. See
the section entitled "5.28. Transhybrid Balance" on page 37 for more information on selecting
coefficients for the programmable hybrid.
Register 46. Programmable Hybrid Register 2
Bit
D7
D6
D5
D4
D3
Name
HYB2[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
7:0
Name
Function
HYB2[7:0] Programmable Hybrid Register 2.
These bits can be programmed with a coefficient value to adjust the hybrid response to
reduce near-end echo. This register represents the second tap in the eight-tap filter. When
this register is set to all 0s, this filter stage does not have an effect on the hybrid response.
See the section entitled "5.28. Transhybrid Balance" on page 37 for more information on
selecting coefficients for the programmable hybrid.
92
Rev. 1.31
Si3050 + Si3018/19
Register 47. Programmable Hybrid Register 3
Bit
D7
D6
D5
D4
D3
Name
HYB3[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
7:0
Name
Function
HYB3[7:0] Programmable Hybrid Register 3.
These bits can be programmed with a coefficient value to adjust the hybrid response to
reduce near-end echo. This register represents the third tap in the eight-tap filter. When this
register is set to all 0s, this filter stage does not have an effect on the hybrid response. See
the section entitled "5.28. Transhybrid Balance" on page 37 for more information on selecting
coefficients for the programmable hybrid.
Register 48. Programmable Hybrid Register 4
Bit
D7
D6
D5
D4
D3
Name
HYB4[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
7:0
Name
Function
HYB4[7:0] Programmable Hybrid Register 4.
These bits can be programmed with a coefficient value to adjust the hybrid response to
reduce near-end echo. This register represents the fourth tap in the eight-tap filter. When this
register is set to all 0s, this filter stage does not have an effect on the hybrid response. See
the section entitled "5.28. Transhybrid Balance" on page 37 for more information on selecting
coefficients for the programmable hybrid.
Rev. 1.31
93
Si3050 + Si3018/19
Register 49. Programmable Hybrid Register 5
Bit
D7
D6
D5
D4
D3
Name
HYB5[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
7:0
Name
Function
HYB5[7:0] Programmable Hybrid Register 5.
These bits can be programmed with a coefficient value to adjust the hybrid response to
reduce near-end echo. This register represents the fifth tap in the eight-tap filter. When this
register is set to all 0s, this filter stage does not have an effect on the hybrid response. See
the section entitled "5.28. Transhybrid Balance" on page 37 for more information on selecting
coefficients for the programmable hybrid.
Register 50. Programmable Hybrid Register 6
Bit
D7
D6
D5
D4
D3
Name
HYB6[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
7:0
Name
Function
HYB6[7:0] Programmable Hybrid Register 6.
These bits can be programmed with a coefficient value to adjust the hybrid response to
reduce near-end echo. This register represents the sixth tap in the eight-tap filter. When this
register is set to all 0s, this filter stage does not have an effect on the hybrid response. See
the section entitled "5.28. Transhybrid Balance" on page 37 for more information on selecting
coefficients for the programmable hybrid.
94
Rev. 1.31
Si3050 + Si3018/19
Register 51. Programmable Hybrid Register 7
Bit
D7
D6
D5
D4
D3
Name
HYB7[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
Name
7:0
Function
HYB7[7:0] Programmable Hybrid Register 7.
These bits can be programmed with a coefficient value to adjust the hybrid response to
reduce near-end echo. This register represents the seventh tap in the eight-tap filter. When
this register is set to all 0s, this filter stage does not have an effect on the hybrid response.
See the section entitled "5.28. Transhybrid Balance" on page 37 for more information on
selecting coefficients for the programmable hybrid.
Register 52. Programmable Hybrid Register 8
Bit
D7
D6
D5
D4
D3
Name
HYB8[7:0]
Type
R/W
D2
D1
D0
Reset settings = 0000_0000
Bit
Name
7:0
Function
HYB8[7:0] Programmable Hybrid Register 8.
These bits can be programmed with a coefficient value to adjust the hybrid response to
reduce near-end echo. This register represents the eighth tap in the eight-tap filter. When this
register is set to all 0s, this filter stage does not have an effect on the hybrid response. See
the section entitled "5.28. Transhybrid Balance" on page 37 for more information on selecting
coefficients for the programmable hybrid.
Register 53-58. Reserved
Bit
D7
D6
D5
D4
D3
D2
D1
D0
Name
Type
Reset settings = xxxx_xxxx
Bit
7:0
Name
Function
Reserved Do not write to these register bits.
Rev. 1.31
95
Si3050 + Si3018/19
Register 59. Spark Quenching Control
Bit
D7
D6
D5
D4
D3
D2
D1
Name
SQ1
SQ0
RG1
GCE
Type
R/W
R/W
R/W
R/W
D0
Reset settings = xxxx_xxxx
Bit
7
6
5
4
3
Function
Reserved Always write this bit to zero.
SQ1
Spark Quenching.
This bit, in combination with the OHS bit (Register 16), and the OHS2 bit (Register 31), sets
the amount of time for the line-side device to go on-hook. The on-hook speeds specified are
measured from the time the OH bit is cleared until loop current equals zero.
OHS
OHS2
SQ[1:0]
Mean On-Hook Speed
0
0
00
Less than 0.5 ms
0
1
00
3 ms±10% (meets ETSI standard)
1
X
11
26 ms ±10% (meets Australia spark quenching spec)
Reserved Always write this bit to zero.
SQ0
Spark Quenching.
This bit, in combination with the OHS bit (Register 16), and the OHS2 bit (Register 31), sets
the amount of time for the line-side device to go on-hook. The on-hook speeds specified are
measured from the time the OH bit is cleared until loop current equals zero.
OHS
OHS2
SQ[1:0]
Mean On-Hook Speed
0
0
00
Less than 0.5 ms
0
1
00
3 ms±10% (meets ETSI standard)
1
X
11
26 ms ±10% (meets Australia spark quenching spec)
Reserved Always write this bit to zero.
2
RG1
Receive Gain 1 (Line-side Revision E or later).
This bit enables receive path gain adjustment.
0 = No gain applied to hybrid, full scale RX on line = 0 dBm.
1 = 1 dB of gain applied to hybrid, full scale RX on line = –1 dBm.
1
GCE
Guarded Clear Enable (Line-side Revision E or later).
This bit (in conjunction with the R2 bit set to 1) enables the Si3050 to meet BT’s Guarded
Clear Spec (B5 6450, Part 1: 1993, Section 15.4.3.3). With these bits set, the DAA will draw
approximately 2.5 mA of current from the line while on-hook.
0 = Default, DAA does not draw loop current.
1 = Guarded Clear enabled, DAA draws 2.5 mA while on-hook to meet Guarded Clear
requirement.
0
96
Name
Reserved Always write this bit to zero.
Rev. 1.31
Si3050 + Si3018/19
APPENDIX—UL1950 3RD EDITION
Introduction
Although designs using the Si3018 and Si3019 comply
with UL1950 3rd Edition and pass all overcurrent and
overvoltage tests, there are still several issues to
consider.
The bottom schematic of Figure 48 shows the
configuration in which the ferrite beads (FB1, FB2) are
on the protected side of the sidactor (RV1). For this
design, the ferrite beads can be rated at 200 mA.
Figure 48 shows two designs that can pass the UL1950
overvoltage tests and electromagnetic emissions. The
top schematic shows the configuration in which the
ferrite beads (FB1, FB2) are on the unprotected side of
the sidactor (RV1). For this configuration, the current
rating of the ferrite beads needs to be 6 A. However, the
higher current ferrite beads are less effective in
reducing electromagnetic emissions.
In a cost optimized design, it is important to remember
that compliance to UL1950 does not always require
overvoltage tests. It is best to plan ahead and know
which overvoltage tests applies to your system.
System-level elements in the construction, such as fire
enclosure and spacing requirements, must be
considered during the design stages. Consult with your
professional testing agency during the design of the
product to determine which tests apply to your system.
C8
75  @ 100 MHz, 6 A
FB1
1.25 A
TIP
RV1
75  @ 100 MHz, 6 A
FB2
RING
C9
C8
600  at 100 MHz, 200 mA
FB1
1.25 A
TIP
RV1
600  at 100 MHz, 200 mA
FB2
RING
C9
Figure 48. Circuits that Pass all UL1950 Overvoltage Tests
Rev. 1.31
97
Si3050 + Si3018/19
7. Pin Descriptions: Si3050
SDO
SDI
CS
FSYNC
PCLK
DTX
DRX
RGDT
AOUT/INT
RG
1
2
3
4
5
6
7
8
9
20
19
18
17
16
15
14
13
12
11
10
SDITHRU
SCLK
GND
VDD
VA
C1A
C2A
RESET
TGDE
TGD
Table 25. Si3050 Pin Descriptions
Pin #
Pin Name
1
SDO
Serial Port Data Output.
Serial port control data output.
2
SDI
Serial Port Data Input.
Serial port control data input.
3
CS
Chip Select Input.
An active low input control signal that enables the SPI Serial port. When inactive,
SCLK and SDI are ignored and SDO is high impedance.
4
FSYNC
5
PCLK
6
DTX
Transmit PCM or GCI Highway Data Output.
Outputs data from either the PCM or GCI highway bus.
7
DRX
Receive PCM or GCI Highway Data Input.
Receives data from either the PCM or GCI highway bus.
8
RGDT
Ring Detect Output.
Produces an active low rectified version of the ring signal.
9
AOUT/INT
10
RG
98
Description
Frame Sync Input.
Data framing signal that is used to indicate the start and stop of a communication/data
frame.
Master Clock Input.
Master clock input.
Analog Speaker Output/Interrupt Output.
Provides an analog output signal for driving a call progress speaker in AOUT mode.
Alternatively, this pin can be set to provide a hardware interrupt signal.
Ring Ground Output.
Control signal for ring ground relay. Used to support ground start applications.
Rev. 1.31
Si3050 + Si3018/19
Table 25. Si3050 Pin Descriptions (Continued)
Pin #
Pin Name
Description
11
TGD
12
TGDE
TIP Ground Detect Enable Output.
Control signal for the ground detect relay. Used to support ground start applications.
13
RESET
Reset Input.
An active low input that is used to reset all control registers to a defined, initialized
state. Also used to bring the Si3050 out of sleep mode.
14
C2A
Isolation Capacitor 2A.
Connects to one side of the isolation capacitor C2. Used to communicate with the
line-side device.
15
C1A
Isolation Capacitor 1A.
Connects to one side of the isolation capacitor C1. Used to communicate with the
line-side device.
16
VA
17
VDD
Digital Supply Voltage.
Provides the 3.3 V digital supply voltage to the Si3050.
18
GND
Ground.
Connects to the system digital ground.
19
SCLK
Serial Port Bit Clock Input.
Controls the serial data on SDO and latches the data on SDI.
20
SDITHRU
TIP Ground Detect Input.
Used to detect current flowing in TIP for supporting ground start applications.
Regulator Voltage Reference.
This pin connects to an external capacitor and serves as the reference for the internal
voltage regulator.
SDI Passthrough Output.
Cascaded SDI output signal to daisy-chain the SPI interface with additional devices.
Rev. 1.31
99
Si3050 + Si3018/19
8. Pin Descriptions: Si3018/19
QE
DCT
RX
IB
C1B
C2B
VREG
RNG1
1
2
3
4
5
6
7
8
16
15
14
13
12
11
10
9
DCT2
IGND
DCT3
QB
QE2
SC
VREG2
RNG2
Pin #
Pin Name
1
QE
2
DCT
3
RX
Receive Input.
Serves as the receive side input from the telephone network.
4
IB
Internal Bias.
Provides a bias voltage to the device.
5
C1B
Isolation Capacitor 1B.
Connects to one side of isolation capacitor C1. Used to communicate with the
system-side device.
6
C2B
Isolation Capacitor 2B.
Connects to one side of isolation capacitor C2. Used to communicate with the
system-side device.
7
VREG
Voltage Regulator.
Connects to an external capacitor to provide bypassing for an internal power supply.
8
RNG1
Ring 1.
Connects through a resistor to the TIP lead of the telephone line. Provides the ring and caller
ID signals to the DAA.
9
RNG2
Ring 2.
Connects through a resistor to the RING lead of the telephone line. Provides the ring and
caller ID signals to the DAA.
10
VREG2
11
SC
12
QE2
Transistor Emitter 2.
Connects to the emitter of Q4.
13
QB
Transistor Base.
Connects to the base of transistor Q4.
100
Description
Transistor Emitter.
Connects to the emitter of Q3.
DC Termination.
Provides dc termination to the telephone network.
Voltage Regulator 2.
Connects to an external capacitor to provide bypassing for an internal power supply.
SC Connection.
Enables external transistor network. Should be tied through a 0  resistor to IGND.
Rev. 1.31
Si3050 + Si3018/19
Pin #
Pin Name
Description
14
DCT3
DC Termination 3.
Provides dc termination to the telephone network.
15
IGND
Isolated Ground.
Connects to ground on the line-side interface.
16
DCT2
DC Termination 2.
Provides dc termination to the telephone network.
Rev. 1.31
101
Si3050 + Si3018/19
9. Ordering Guide
Chipset
Region
System-side
(TSSOP)
Line-side
(SOIC)
Line-side
(TSSOP)
Pb-Free and
RoHS-Compliant
Temperature
Si3050 +
Si3019
Enhanced Global Si3050-E-FT Si3019-F-FS
Si3019-F-FT
Yes
0 to +70 C
Si3050 +
Si3019
Enhanced Global Si3050-E-GT Si3019-F-GS Si3019-F-GT
Yes
–40 to +85 C
Yes
0 to +70 C
Si3050 +
Si3018
Global
Si3050-E-FT Si3018-F-FS
Si3018-F-FT
Note: Refer to "10. Product Identification" on page 102 for more information on part naming conventions.
10. Product Identification
The product identification number is a finished goods part number or is specified by a finished goods part number,
such as a special customer part number.
Example:
Si3050-E-FSR
Product Designator
Product Revision
Shipping Option
Blank = Tubes
R = Tape and Reel
Package Type
S = SOIC
T = TSSOP
Part Type/Lead Finish
F = Commercial/Lead-Free
G = Industrial Temp/Lead-Free
102
Rev. 1.31
Si3050 + Si3018/19
11. Package Outline: 20-Pin TSSOP
Figure 49 illustrates the package details for the Si3050. Table 26 lists the values for the dimensions shown in the
illustration.
Figure 49. 20-Pin Thin Shrink Small Outline Package (TSSOP)
Rev. 1.31
103
Si3050 + Si3018/19
Table 26. 20-Pin TSSOP Package Diagram Dimensions
Dimension
Min
Nom
Max
A
—
—
1.20
A1
0.05
—
0.15
A2
0.80
1.00
1.05
b
0.19
—
0.30
c
0.09
—
0.20
D
6.40
6.50
6.60
E
E1
6.40 BSC
4.40
4.40
e
L
0.65 BSC
0.45
0.60
L2
θ
4.50
0.75
0.25 BSC
0°
—
aaa
0.10
bbb
0.10
ccc
0.20
8°
Notes:
1. All dimensions shown are in millimeters (mm) unless otherwise noted.
2. Dimensioning and Tolerancing per ANSI Y14.5M-1994.
3. This drawing conforms to the JEDEC Solid State Outline MO-153, Variation AC.
4. Recommended card reflow profile is per the JEDEC/IPC J-STD-020C
specification for Small Body Components.
104
Rev. 1.31
Si3050 + Si3018/19
12. Package Outline: 16-Pin SOIC
Figure 50 illustrates the package details for the Si3018/19. Table 27 lists the values for the dimensions shown in
the illustration.
Figure 50. 16-Pin Small Outline Integrated Circuit (SOIC) Package
Rev. 1.31
105
Si3050 + Si3018/19
Table 27. 16-Pin SOIC Package Diagram Dimensions
Dimension
Min
Max
A
—
1.75
A1
0.10
0.25
A2
1.25
—
b
0.31
0.51
c
0.17
0.25
D
9.90 BSC
E
6.00 BSC
E1
3.90 BSC
e
1.27 BSC
L
0.40
L2
1.27
0.25 BSC
h
0.25
0.50
θ
0°
8°
aaa
0.10
bbb
0.20
ccc
0.10
ddd
0.25
Notes:
1. All dimensions shown are in millimeters (mm) unless otherwise noted.
2. Dimensioning and Tolerancing per ANSI Y14.5M-1994.
3. This drawing conforms to the JEDEC Solid State Outline MS-012, Variation AC.
4. Recommended card reflow profile is per the JEDEC/IPC J-STD-020C
specification for Small Body Components.
106
Rev. 1.31
Si3050 + Si3018/19
13. Package Outline: 16-Pin TSSOP
Figure 51 illustrates the package details for the Si3018/19. Table 28 lists the values for the dimensions shown in
the illustration.
Figure 51. 16-Pin Thin Shrink Small Outline Package (TSSOP)
Rev. 1.31
107
Si3050 + Si3018/19
Table 28. 16-Pin TSSOP Package Diagram Dimensions
Dimension
Min
Nom
Max
A
—
—
1.20
A1
0.05
—
0.15
A2
0.80
1.00
1.05
b
0.19
—
0.30
c
0.09
—
0.20
D
4.90
5.00
5.10
E
E1
6.40 BSC
4.40
4.40
e
L
0.65 BSC
0.45
0.60
L2
θ
4.50
0.75
0.25 BSC
0°
—
aaa
0.10
bbb
0.10
ccc
0.20
8°
Notes:
1. All dimensions shown are in millimeters (mm) unless otherwise noted.
2. Dimensioning and Tolerancing per ANSI Y14.5M-1994.
3. This drawing conforms to the JEDEC Solid State Outline MO-153, Variation AB.
4. Recommended card reflow profile is per the JEDEC/IPC J-STD-020C
specification for Small Body Components.
108
Rev. 1.31
Si3050 + Si3018/19
SILICON LABS Si3050 SUPPORT DOCUMENTATION









AN16: Multiple Device Support
AN17: Designing for International Safety Compliance
AN30: Ground Start Implementation with Silicon Laboratories’ DAAs
AN67: Layout Guidelines
AN72: Ring Detection/Validation with the Si305x DAAs
AN77: Silicon DAA Software Guidelines (Si3050)
AN81: Emissions Design Considerations
AN84: Digital Hybrid with the Si305x DAAs
Si3050PPT-EVB Data Sheet
Note: Refer to www.silabs.com for a current list of support documents for this chipset.
Rev. 1.31
109
Si3050 + Si3018/19
DOCUMENT CHANGE LIST
present in any Si3050 device).
 Removed SnPb package options
 Minor typo corrections
Revision 1.01 to Revision 1.1













Added package thermal information in Table 1,
“Recommended Operating Conditions and Thermal
Information,” on page 5.
Added Note 10 to the transhybrid balance parameter
in Table 4 on page 8.
Updated Table 7, “Switching Characteristics—Serial
Peripheral Interface,” on page 11.
Removed R54 and R55 from "3. Bill of Materials" on
page 18.
Changed recommended DCV setting for Japan from
01 to 10 in Table 13 on page 21.
Updated initialization procedure in "5.3. Initialization"
on page 24.
Removed incorrect description of FDT bit in "5.8.
Exception Handling" on page 26.
Updated Billing Tone and Receive Overload section.
Changed to "5.22. Receive Overload Detection" on
page 33.
Updated text and added description of hybrid
coefficient format in "5.28. Transhybrid Balance" on
page 37.
Removed references to line-side revisions C and E.
Updated "9. Ordering Guide" on page 102.
Updated package information for 20-Pin TSSOP and
16-Pin SOIC on pages 103 and 104.
Added "13. Package Outline: 16-Pin TSSOP" on
page 107.
Revision 1.1 to Revision 1.31

Revision 1.1 to Revision 1.2

Updated Table 7, “Switching Characteristics—Serial
Peripheral Interface,” on page 11.
Updated

delay time between chip selects.
Updated Table 13, “Country Specific Register
Settings,” on page 21.
Corrected

Revised

Step 6 with standard hexadecimal notation.
Updated Figure 25, “Si3018/19 Signal Flow
Diagram,” on page 37.
Corrected

ACIM settings for Brazil.
Updated "5.3. Initialization" on page 24.
HPF pole.
Updated "9. Ordering Guide" on page 102.
Revision 1.2 to Revision 1.3
Updated Deep Sleep Total Supply Current from 1.0
to 1.3 mA typical
 Updated package pictures
 Removed all SPIM references (SPIM bit is never

110
Rev. 1.31
The internal System-Side Revision value (REVA[3:0]
in Register 11) has been incremented by one for
Si3050 revision E.
Si3050 + Si3018/19
NOTES:
Rev. 1.31
111
Si3050 + Si3018/19
CONTACT INFORMATION
Silicon Laboratories Inc.
400 West Cesar Chavez
Austin, TX 78701
Tel: 1+(512) 416-8500
Fax: 1+(512) 416-9669
Toll Free: 1+(877) 444-3032
Email: [email protected]
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 and Silicon Labs are trademarks of Silicon Laboratories Inc.
Other products or brandnames mentioned herein are trademarks or registered trademarks of their respective holders.
112
Rev. 1.31