AD ADPD103BCPZRL Temperature and power specification Datasheet

Photometric Front End
ADPD103
Data Sheet
FEATURES
GENERAL DESCRIPTION
Multifunction photometric front end
Fully integrated AFE, ADC, LED drivers, and timing core
Usable in a broad range of optical measurement
applications, including photoplethysmography
Enables best-in-class ambient light rejection capability
without the need for photodiode optical filters
Three 8 mA to 250 mA LED drivers
Separate data registers for each LED/photodiode
combination
1 to 8 optical inputs
Flexible, multiple, short LED pulses per optical sample
20-bit burst accumulator enabling 20 bits per sample period
On-board sample to sample accumulator, enabling up to
27 bits per data read
Low power operation
I2C interface and 1.8 V analog/digital core
Flexible sampling frequency ranging from 0.122 Hz to
3.820 kHz
FIFO data operation
The ADPD103 is a highly efficient photometric front end with
an integrated 14-bit analog-to-digital converter (ADC) and a
20-bit burst accumulator that works in concert with flexible
light emitting diode (LED) drivers. It is designed to stimulate an
LED and measure the corresponding optical return signal. The
data output and functional configuration occur over a 1.8 V I2C
interface. The control circuitry includes flexible LED signaling and
synchronous detection.
The analog front end (AFE) features best-in-class rejection of signal
offset and corruption due to modulated interference commonly
caused by ambient light.
Couple the ADPD103 with a low capacitance photodiode of
<100 pF for optimal performance. The ADPD103 can be used with
any LED.
APPLICATIONS
Body worn health and fitness monitors, for example, heart
rate monitoring
Clinical measurements, for example, SpO2
Industrial monitoring
Background light measurements
Rev. B
Document Feedback
Information furnished by Analog Devices is believed to be accurate and reliable. However, no
responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other
rights of third parties that may result from its use. Specifications subject to change without notice. No
license is granted by implication or otherwise under any patent or patent rights of Analog Devices.
Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700 ©2015–2016 Analog Devices, Inc. All rights reserved.
Technical Support
www.analog.com
ADPD103* Product Page Quick Links
Last Content Update: 11/01/2016
Comparable Parts
Design Resources
View a parametric search of comparable parts
• ADPD103 Evaluation Board
•
•
•
•
Documentation
Discussions
Data Sheet
• ADPD103: Photometric Front End Data Sheet
User Guides
• UG-947: Evaluating the ADPD103 Photometric Front End
View all ADPD103 EngineerZone Discussions
Evaluation Kits
ADPD103 Material Declaration
PCN-PDN Information
Quality And Reliability
Symbols and Footprints
Sample and Buy
Visit the product page to see pricing options
Technical Support
Submit a technical question or find your regional support
number
* This page was dynamically generated by Analog Devices, Inc. and inserted into this data sheet. Note: Dynamic changes to
the content on this page does not constitute a change to the revision number of the product data sheet. This content may be
frequently modified.
ADPD103
Data Sheet
TABLE OF CONTENTS
Features .............................................................................................. 1
AFE Integration Offset Adjustment ......................................... 18
Applications ....................................................................................... 1
I2C Serial Interface ..................................................................... 20
General Description ......................................................................... 1
Typical Connection Diagram ................................................... 21
Revision History ............................................................................... 2
LED Driver Pins and LED Supply Voltage .............................. 23
Functional Block Diagram .............................................................. 3
LED Driver Operation ............................................................... 23
Specifications..................................................................................... 4
Determining the Average Current ........................................... 23
Temperature and Power Specifications ..................................... 4
Determining CVLED ..................................................................... 23
Performance Specifications ......................................................... 5
LED Inductance Considerations .............................................. 24
Analog Specifications ................................................................... 6
Recommended Start-Up Sequence .......................................... 24
Digital Specifications ................................................................... 7
Reading Data ............................................................................... 24
Timing Specifications .................................................................. 8
Clocks and Timing Calibration ................................................ 26
Absolute Maximum Ratings............................................................ 9
Calculating Current Consumption .......................................... 27
Thermal Resistance ...................................................................... 9
Optimizing SNR per Watt ......................................................... 27
Recommended Soldering Profile ............................................... 9
Single AFE channel mode ......................................................... 28
ESD Caution .................................................................................. 9
TIA_ADC Mode ......................................................................... 28
Pin Configurations and Function Descriptions ......................... 10
Digital Integrate Mode............................................................... 30
Typical Performance Characteristics ........................................... 12
Register Listing ............................................................................... 34
Theory of Operation ...................................................................... 13
LED Control Registers ............................................................... 38
Introduction ................................................................................ 13
AFE Configuration Registers .................................................... 41
Dual Time Slot Operation ......................................................... 13
System Registers ......................................................................... 46
Time Slot Switch ......................................................................... 14
ADC Registers ............................................................................ 50
Adjustable Sampling Frequency ............................................... 15
Data Registers ............................................................................. 51
State Machine Operation ........................................................... 16
Outline Dimensions ....................................................................... 52
Normal Mode Operation and Data Flow ................................ 16
Ordering Guide .......................................................................... 52
AFE Operation ............................................................................ 18
REVISION HISTORY
2/16—Revision B: Initial Version
Rev. B | Page 2 of 52
Data Sheet
ADPD103
FUNCTIONAL BLOCK DIAGRAM
AVDD
TIME SLOT
SWITCH
DVDD
ANALOG BLOCK
PDC
AFE:
SIGNAL CONDITIONING
PD1
TIA
PD5
BPF
ADPD103
±1 INTEGRATOR
VBIAS
VREF
AFE:
SIGNAL CONDITIONING
PD2
TIA
BPF
PD6
14-BIT
ADC
VBIAS
AFE:
SIGNAL CONDITIONING
PD3
TIA
BPF
PD7
AFE
CONFIGURATION
±1 INTEGRATOR
VBIAS
AFE:
SIGNAL CONDITIONING
PD4
TIA
PD8
1µF
TIME SLOT A
DATA
±1 INTEGRATOR
BPF
TIME SLOT B
DATA
SDA
SCL
INT
A
PDSO
B
DGND
SLOT
SELECT
DIGITAL
DATAPATH
AND
INTERFACE
CONTROL
±1 INTEGRATOR
AGND
VBIAS
LEDX3
LED3
LEDX2
LED2
LED3 DRIVER
LED3 LEVEL AND TIMING CONTROL
LED2 DRIVER
LED2 LEVEL AND TIMING CONTROL
LED1 DRIVER
LED1 LEVEL AND TIMING CONTROL
LEDX1
LED1
VLED
12722-001
LGND
Figure 1. Typical Functional Block Diagram
Rev. B | Page 3 of 52
ADPD103
Data Sheet
SPECIFICATIONS
TEMPERATURE AND POWER SPECIFICATIONS
Table 1. Operating Conditions
Parameter
TEMPERATURE RANGE
Operating Range
Storage Range
POWER SUPPLY VOLTAGES
VDD
Test Conditions/Comments
Min
Typ
−40
−65
Applied at the AVDD and DVDD pins
1.7
1.8
Max
Unit
+85
+150
°C
°C
1.9
V
AVDD = DVDD = 1.8 V, ambient temperature, unless otherwise noted.
Table 2. Current Consumption 1, 2
Parameter
POWER SUPPLY (VDD) CURRENT
VDD Supply Current
Symbol
LED_OFFSET = 25 µs; LED_PERIOD =19 µs; LED peak current =
25 mA, 4 channels active
100 Hz data rate; Time Slot A only
100 Hz data rate; Time Slot B only
100 Hz data rate; both Time Slot A and Time Slot B
100 Hz data rate; Time Slot A only
100 Hz data rate; Time Slot B only
100 Hz data rate; both Time Slot A and Time Slot B
1 Pulse
10 Pulses
Peak VDD Supply Current
(1.8 V)
4-Channel Operation
1-Channel Operation
Standby Mode Current
VLEDA AND VLEDB SUPPLY
CURRENT
Average Supply Current
VLEDA or VLEDB
1 Pulse
10 Pulses
1
2
Test Conditions/Comments
Min
Typ
Max
Unit
106
94
151
258
246
455
µA
µA
µA
µA
µA
µA
9.3
2.3
3.5
mA
mA
µA
15
30
60
150
300
600
µA
µA
µA
µA
µA
µA
IVDD_PEAK
IVDD_STANDBY
Peak LED current = 100 mA; LED_PULSE width = 3 µs
50 Hz data rate
100 Hz data rate
200 Hz data rate
50 Hz data rate
100 Hz data rate
200 Hz data rate
LEDA or LEDB is one of LED1, LED2, or LED3. VLEDA or VLEDB is one of VLED1, VLED2, or VLED3.
VDD is the voltage applied at the AVDD and DVDD pins.
Rev. B | Page 4 of 52
Data Sheet
ADPD103
PERFORMANCE SPECIFICATIONS
AVDD = DVDD = 1.8 V, TA = full operating temperature range, unless otherwise noted.
Table 3.
Parameter
DATA AQUISITION
Resolution
Resolution/Sample
Resolution/Data Read
LED DRIVER
LED Current Slew Rate 1
Rise
Fall
LED Peak Current
Driver Compliance Voltage
LED PERIOD
Sampling Frequency 2
CATHODE PIN (PDC) VOLTAGE
During All Sampling Periods
During Slot A Sampling
During Slot B Sampling
During Sleep Periods
Test Conditions/Comments
Min
Typ
Max
Unit
Single pulse
64 to 255 pulses
64 to 255 pulses and sample average = 128
14
20
27
Bits
Bits
Bits
Slew rate control setting = 0; TA = 25°C; ILED = 70 mA
Slew rate control setting = 7; TA = 25°C; ILED = 70 mA
Slew rate control setting = 0, 1, 2; TA = 25°C; ILED = 70 mA
Slew rate control setting = 6, 7; TA = 25°C; ILED = 70 mA
LED pulse enabled
Voltage above ground required for LED driver operation
AFE width = 4 µs
AFE width = 3 µs
Time Slot A only; normal mode; 1 pulse; OFFSET_LEDA = 23 µs; PERIOD_LEDA = 19 µs
Time Slot B only; normal mode; 1 pulse; OFFSET_LEDA = 23 µs; PERIOD_LEDA = 19 µs
Both time slots; normal mode; 1 pulse; OFFSET_LEDA = 23 µs; PERIOD_LEDA = 19 µs
Time Slot A only; normal mode; 8 pulses; OFFSET_LEDA = 23 µs; PERIOD_LEDA = 19 µs
Time Slot B only; normal mode; 8 pulses; OFFSET_LEDA = 23 µs; PERIOD_LEDA = 19 µs
Both time slots; normal mode; 8 pulses; OFFSET_LEDA = 23 µs; PERIOD_LEDA = 19 µs
240
1400
3200
4500
mA/µs
mA/µs
mA/µs
mA/µs
mA
V
µs
µs
Hz
Hz
Hz
Hz
Hz
Hz
Register 0x54, Bit 7 = 0x0; Register 0x3C, Bit 9 = 1 3
Register 0x54, Bit 7 = 0x0; Register 0x3C, Bit 9 = 0
Register 0x54, Bit 7 = 0x1; Register 0x54, Bits[9:8] = 0x03
Register 0x54, Bit 7 = 0x1; Register 0x54, Bits[9:8] = 0x1
Register 0x54, Bit 7 = 0x1; Register 0x54, Bits[9:8] = 0x2
Register 0x54, Bit 7 = 0x1; Register 0x54, Bits[9:8] = 0x3 4
Register 0x54, Bit 7 = 0x1; Register 0x54, Bits[11:10] = 0x03
Register 0x54, Bit 7 = 0x1; Register 0x54, Bits[11:10] = 0x1
Register 0x54, Bit 7 = 0x1; Register 0x54, Bits[11:10] = 0x2
Register 0x54, Bit 7 = 0x1; Register 0x54, Bits[11:10] = 0x34
Register 0x54, Bit 7 = 0x0; Register 0x3C, Bit 9 = 1
Register 0x54, Bit 7 = 0x0; Register 0x3C, Bit 9 = 0
Register 0x54, Bit 7 = 0x1; Register 0x54, Bits[13:12] = 0x0
Register 0x54, Bit 7 = 0x1; Register 0x54[13:12] = 0x1
Register 0x54, Bit 7 = 0x1; Register 0x54[13:12] = 0x2
Register 0x54, Bit 7 = 0x1; Register 0x54[13:12] = 0x3
PHOTODIODE INPUT PINS/ ANODE
VOLTAGE
During All Sampling Periods
During Sleep Periods
8
0.2
19
17
0.122
0.122
0.122
0.122
0.122
0.122
250
3230
3820
1750
2257
2531
1193
1.8
1.3
1.8
1.3
1.55
0
1.8
1.3
1.55
0
1.8
1.3
1.8
1.3
1.55
0
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
1.3
Cathode voltage
V
V
LED inductance is negligible for these values. The effective slew rate slows with increased inductance.
The maximum values in this specification are the internal ADC sampling rates in normal mode. The I2C read rates in some configurations may limit the actual output
data rate of the device
3
This mode may induce additional noise and is not recommended unless absolutely necessary. The 1.8 V setting uses VDD, which contains greater amounts of
differential voltage noise with respect to the anode voltage. A differential voltage between the anode and cathode injects a differential current across the capacitance
of the photodiode of the magnitude C × dV/dt.
4
This setting is not recommended for photodiodes because it causes a 1.3 V forward bias of the photodiode.
1
2
Rev. B | Page 5 of 52
ADPD103
Data Sheet
ANALOG SPECIFICATIONS
AVDD = DVDD = 1.8 V, TA = full operating temperature range, unless otherwise noted. Compensation of the AFE offset is explained in
the AFE Operation section.
Table 4.
Parameter
INPUT CAPACITANCE
PULSED SIGNAL CONVERSIONS, 3 μs WIDE LED PULSE 1
ADC Resolution 2
ADC Saturation Level
Ambient Signal Headroom on Pulsed Signal
PULSED SIGNAL CONVERSIONS, 2 μs WIDE LED PULSE1
ADC Resolution2
ADC Saturation Level
Ambient Signal Headroom on Pulsed Signal
FULL SIGNAL CONVERSIONS 3
TIA Saturation Level of Pulsed Signal and Ambient Level
Test Conditions/Comments
4 μs wide AFE integration; normal operation, Register 0x43
(Time Slot A) and Register 0x45 (Time Slot B) = 0xADA5
Transimpedance amplifier (TIA) feedback resistor
25 kΩ
50 kΩ
100 kΩ
200 kΩ
TIA feedback resistor
25 kΩ
50 kΩ
100 kΩ
200 kΩ
TIA feedback resistor
25 kΩ
50 kΩ
100 kΩ
200 kΩ
3 μs wide AFE integration; normal operation, Register 0x43
(Time Slot A) and Register 0x45 (Time Slot B) = 0xADA5
TIA feedback resistor
25 kΩ
50 kΩ
100 kΩ
200 kΩ
TIA feedback resistor
25 kΩ
50 kΩ
100 kΩ
200 kΩ
TIA feedback resistor
25 kΩ
50 kΩ
100 kΩ
200 kΩ
TIA feedback resistor
25 kΩ
50 kΩ
100 kΩ
200 kΩ
Rev. B | Page 6 of 52
Min
Typ
Max
100
Unit
pF
1.64
0.82
0.41
0.2
nA/LSB
nA/LSB
nA/LSB
nA/LSB
13.4
6.7
3.35
1.67
μA
μA
μA
μA
37
18.5
9.25
4.63
μA
μA
μA
μA
2.31
1.15
0.58
0.29
nA/LSB
nA/LSB
nA/LSB
nA/LSB
18.9
9.46
4.73
2.37
μA
μA
μA
μA
31.5
15.7
7.87
3.93
μA
μA
μA
μA
50.4
25.2
12.6
6.3
μA
μA
μA
μA
Data Sheet
ADPD103
Parameter
SYSTEM PERFORMANCE
Total Output Noise Floor
Test Conditions/Comments
Min
Normal mode; per pulse; per channel; no LED; CPD = 70 pF
25 kΩ; referred to ADC input
25 kΩ; referred to peak input signal for 2 µs LED pulse
25 kΩ; referred to peak input signal for 3 µs LED pulse
25 kΩ; saturation signal-to-noise ratio (SNR) per pulse
per channel 4
50 kΩ; referred to ADC input
50 kΩ; referred to peak input signal for 2 µs LED pulse
50 kΩ; referred to peak input signal for 3 µs LED pulse
50 kΩ; saturation SNR per pulse per channel4
100 kΩ; referred to ADC input
100 kΩ; referred to peak input signal for 2 µs LED pulse
100 kΩ; referred to peak input signal for 3 µs LED pulse
100 kΩ; saturation SNR per pulse per channel4
200 kΩ; referred to ADC input
200 kΩ; referred to peak input signal for 2 µs LED pulse
200 kΩ; referred to peak input signal for 3 µs LED pulse
200 kΩ; saturation SNR per pulse per channel4
DC Power Supply Rejection Ratio (DC PSRR)
Typ
Max
Unit
2.0
4.6
3.3
72.3
LSB rms
nA rms
nA rms
dB
2.4
2.8
2.0
70.6
3.4
1.9
1.4
67.6
5.5
1.6
1.1
63.5
−37
LSB rms
nA rms
nA rms
dB
LSB rms
nA rms
nA rms
dB
LSB rms
nA rms
nA rms
dB
dB
This saturation level applies to the ADC only and, therefore, includes only the pulsed signal. Any nonpulsatile signal is removed prior to the ADC stage.
ADC resolution is listed per pulse when the AFE offset is correctly compensated per the AFE Operation section. If using multiple pulses, divide by the number of pulses.
3
This saturation level applies to the full signal path and, therefore, includes both the ambient signal and the pulsed signal.
4
The noise term of the saturation SNR value refers to the receive noise only and does not include photon shot noise or any noise on the LED signal itself.
1
2
DIGITAL SPECIFICATIONS
DVDD = 1.7 V to 1.9 V, unless otherwise noted.
Table 5.
Parameter
LOGIC INPUTS (SCL, SDA)
Input Voltage Level
High
Low
Input Current Level
High
Low
Input Capacitance
LOGIC OUTPUTS
INT Output Voltage Level
High
Low
PDSO Output Voltage Level
High
Low
SDA Output Voltage Level
Low
SDA Output Current Level
Low
Symbol
Test Conditions/Comments
Min
Typ
Max
Unit
VIH
VIL
0.7 × DVDD
3.6
0.3 × DVDD
V
V
IIH
IIL
CIN
−10
−10
+10
+10
µA
µA
pF
10
VOH
VOL
2 mA high level output current
2 mA low level output current
DVDD − 0.5
VOH
VOL
2 mA high level output current
2 mA low level output current
DVDD − 0.5
VOL1
2 mA low level output current
IOL
VOL1 = 0.6 V
Rev. B | Page 7 of 52
6
0.5
V
V
0.5
V
V
0.2 × DVDD
V
mA
ADPD103
Data Sheet
TIMING SPECIFICATIONS
Table 6. I2C Timing Specifications
Parameter
I2C PORT 1
SCL
Frequency
Minimum Pulse Width
High
Low
Start Condition
Hold Time
Setup Time
SDA Setup Time
SCL and SDA
Rise Time
Fall Time
Stop Condition
Setup Time
Test Conditions/Comments
See Figure 2
Min
Typ
Max
400
t1
t2
600
1300
ns
ns
t3
t4
t5
600
600
100
ns
ns
ns
t6
t7
1000
300
t8
600
t5
t3
t3
SDA
t6
t1
SCL
t7
t4
Figure 2. I2C Timing
Rev. B | Page 8 of 52
t8
ns
ns
ns
Guaranteed by design.
t2
Unit
kHz
12722-002
1
Symbol
Data Sheet
ADPD103
ABSOLUTE MAXIMUM RATINGS
RECOMMENDED SOLDERING PROFILE
Table 7.
Figure 3 and Table 9 provide details about the recommended
soldering profile.
Rating
−0.3 V to +2.2 V
−0.3 V to +2.2 V
−0.3 V to +2.2 V
−0.3 V to +2.2 V
−0.3 V to +3.6 V
−0.3 V to +3.9 V
−0.3 V to +3.9 V
150°C
RAMP-UP
tL
TSMAX
TSMIN
tS
RAMP-DOWN
1500 V
1250 V
100 V
t25°C TO PEAK
TIME
Figure 3. Recommended Soldering Profile
Table 9. Recommended Soldering Profile
1500 V
500 V
100 V
THERMAL RESISTANCE
Table 8. Thermal Resistance
θJA
54.9
60
12722-003
TL
PREHEAT
Stresses at or above those listed under Absolute Maximum
Ratings may cause permanent damage to the product. This is a
stress rating only; functional operation of the product at these
or any other conditions above those indicated in the operational
section of this specification is not implied. Operation beyond
the maximum operating conditions for extended periods may
affect product reliability.
Package Type
28-Lead LFCSP_WQ
16-Ball WLCSP
CRITICAL ZONE
TL TO TP
tP
TP
TEMPERATURE
Parameter
AVDD to AGND
DVDD to DGND
INT to DGND
PDSO to DGND
LEDXx to LGND
SCL to DGND
SDA to DGND
Junction Temperature
ESD
28-Lead LFCSP
Human Body Model (HBM)
Charge Device Model (CDM)
Machine Model (MM)
16-Ball WLCSP
Human Body Model (HBM)
Charge Device Model (CDM)
Machine Model (MM)
Unit
°C/W
°C/W
Profile Feature
Average Ramp Rate (TL to TP)
Preheat
Minimum Temperature (TSMIN)
Maximum Temperature (TSMAX)
Time (TSMIN to TSMAX) (tS)
TSMAX to TL Ramp-Up Rate
Time Maintained Above Liquidous
Temperature
Liquidous Temperature (TL)
Time (tL)
Peak Temperature (TP)
Time Within 5°C of Actual Peak
Temperature (tP)
Ramp-Down Rate
Time from 25°C to Peak Temperature
ESD CAUTION
Rev. B | Page 9 of 52
Condition (Pb-Free)
3°C/sec max
150°C
200°C
60 sec to 180 sec
3°C/sec maximum
217°C
60 sec to 150 sec
+260 (+0/−5)°C
<30 sec
6°C/sec maximum
8 minutes maximum
ADPD103
Data Sheet
22 NIC
25 LEDX2
24 LEDX3
23 LEDX1
28 SDA
27 SCL
26 LGND
PIN CONFIGURATIONS AND FUNCTION DESCRIPTIONS
21 NIC
20 NIC
INT 1
PDSO 2
DVDD 3
ADPD103
19 NIC
AGND 4
TOP VIEW
(Not to Scale)
18 NIC
PD7 14
PD6 13
PD3
PD2
PD5 12
15 PD8
PD4 10
PD1 7
PDC 11
17 NIC
16 NIC
8
9
AVDD 6
NOTES
1. NIC = NONBONDED PAD, CAN BE GROUNDED.
2. EXPOSED PAD (DIGITAL GROUND). CONNECT THE
EXPOSED PAD TO GROUND.
12722-004
VREF 5
Figure 4. 28-Lead LFCSP Pin Configuration
Table 10. 28-Lead LFCSP Pin Function Descriptions
Pin No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16 to 22
23
24
25
26
27
28
1
Mnemonic
INT
PDSO
DVDD
AGND
VREF
AVDD
PD1
PD2
PD3
PD4
PDC
PD5
PD6
PD7
PD8
NIC
LEDX1
LEDX3
LEDX2
LGND
SCL
SDA
EPAD (DGND)
Type 1
DO
DO
S
S
REF
S
AI
AI
AI
AI
AO
AI
AI
AI
AI
R
AO
AO
AO
S
DI
DIO
S
Description
Interrupt Output.
Power-Down Status Output.
1.8 V Digital Supply.
Analog Ground.
Internally Generated ADC Voltage Reference. Buffer this pin with a 1 µF capacitor to AGND.
1.8 V Analog Supply.
Photodiode Current Input (Anode). If not in use, leave this pin floating.
Photodiode Current Input (Anode). If not in use, leave this pin floating.
Photodiode Current Input (Anode). If not in use, leave this pin floating.
Photodiode Current Input (Anode). If not in use, leave this pin floating.
Photodiode Common Cathode Bias.
Photodiode Current Input (Anode). If not in use, leave this pin floating.
Photodiode Current Input (Anode). If not in use, leave this pin floating.
Photodiode Current Input (Anode). If not in use, leave this pin floating.
Photodiode Current Input (Anode). If not in use, leave this pin floating.
Not Internally Connected (Nonbonded Pad). This pin can be grounded.
LED Driver 1 Current Sink. If not in use, leave this pin floating.
LED Driver 3 Current Sink. If not in use, leave this pin floating.
LED Driver 2 Current Sink. If not in use, leave this pin floating.
LED Driver Ground.
I2C Clock Input.
I2C Data Input/Output.
Exposed Pad (Digital Ground). Connect the exposed pad to ground.
DO means digital output, S means supply, REF means voltage reference, AI means analog input, AO means analog output, R means reserved, DI means digital input,
and DIO means digital input/output.
Rev. B | Page 10 of 52
Data Sheet
ADPD103
ADPD103
TOP VIEW, BALL SIDE DOWN
(Not to Scale)
2
A
LGND
LEDX2
B
LEDX3
LEDX1
SDA
C
SCL
INT
DVDD
DGND
AGND
D
3
E
PDSO
VREF
AVDD
F
PD5-8
PDC
PD1-4
12722-005
1
Figure 5. 16-Ball WLCSP Pin Configuration
Table 11. 16-Ball WLCSP Pin Function Descriptions
Pin No.
A1
A2
B1
B2
B3
C1
C2
C3
D2
D3
E1
E2
E3
F1
F2
F3
1
Mnemonic
LGND
LEDX2
LEDX3
LEDX1
SDA
SCL
INT
DVDD
DGND
AGND
PDSO
VREF
AVDD
PD5-8
PDC
PD1-4
Type 1
S
AO
AO
AO
DIO
S
DO
S
S
S
DO
REF
S
AI
AO
AI
Description
LED Driver Ground.
LED Driver 2 Current Sink. If not in use, leave this pin floating.
LED Driver 3 Current Sink. If not in use, leave this pin floating.
LED Driver 1 Current Sink. If not in use, leave this pin floating.
I2C Data Input/Output.
I2C Clock Input.
Interrupt Output.
1.8 V Digital Supply.
Digital Ground.
Analog Ground.
Power-Down Status Output.
Internally Generated ADC Voltage Reference. Buffer this pin with a 1 µF capacitor to AGND.
1.8 V Analog Supply.
Photodiode Combined Current Input of PD5 to PD8. If not in use, leave this pin floating.
Photodiode Common Cathode Bias.
Photodiode Combined Current Input of PD1 to PD4. If not in use, leave this pin floating.
S means supply, AO means analog output, DIO means digital input/output, DO means digital output, REF means voltage reference, and AI means analog input.
Rev. B | Page 11 of 52
ADPD103
Data Sheet
TYPICAL PERFORMANCE CHARACTERISTICS
4.0
3.5
25
25k
3.0
20
NOISE (nA rms)
PERCENT OF POPULATION (%)
30
15
10
2.5
50k
2.0
100k
1.5
200k
1.0
5
–25
–20
–15
–10
–5
0
5
10
SAMPLE FREQUENCY DEVIATION FROM NOMINAL (%)
15
0
12722-006
0
0
20
40
60
80
100
120
140
160
180
PHOTODIODE CAPACITANCE (pF)
Figure 6. 32 kHz Clock Frequency Distribution
(Default Settings, Before User Calibration: Register 0x4B = 0x2612)
12722-200
0.5
Figure 8. Input Referred Noise vs. Photodiode Capacitance,
LED Pulse Width = 3 µs
5
20
NOISE (nA rms)
25k
15
10
4
50k
3
100k
2
200k
5
0
0
27.5 28.0 28.5 29.0 29.5 30.0 30.5 31.0 31.5 32.0
FREQUENCY (MHz)
0
20
40
60
80
100
120
140
160
180
PHOTODIODE CAPACITANCE (pF)
Figure 9. Input Referred Noise vs. Photodiode Capacitance,
LED Pulse Width = 2 µs
Figure 7. 32 MHz Clock Frequency Distribution
(Default Settings, Before User Calibration: Register 0x4D = 0x425E)
Rev. B | Page 12 of 52
12722-201
1
12722-007
PERCENT OF POPULATION (%)
6
Data Sheet
ADPD103
THEORY OF OPERATION
INTRODUCTION
DUAL TIME SLOT OPERATION
The ADPD103 operates as a complete optical transceiver
stimulating up to three LEDs and measuring the return signal
on up to eight separate current inputs. The core consists of a
photometric front end coupled with an ADC, digital block, and
three independent LED drivers. The core circuitry stimulates
the LEDs and measures the return in the analog block through
one to eight photodiode inputs, storing the results in discrete data
locations. The eight inputs are broken into two blocks of four
simultaneous input channels. Data can be read directly by a
register, or through a FIFO. This highly integrated system includes
an analog signal processing block, digital signal processing block,
I2C communication interface, and programmable pulsed LED
current sources.
The ADPD103 operates in two independent time slots, Time Slot A
and Time Slot B, which are carried out sequentially. The entire
signal path from LED stimulation to data capture and processing is
executed during each time slot. Each time slot has a separate
datapath that uses independent settings for the LED driver, AFE
setup, and the resulting data. Time Slot A and Time Slot B operate
in sequence for every sampling period, as shown in Figure 10.
The timing parameters are defined as follows:
tA (µs) = SLOTA_LED_OFFSET + nA × SLOTA_LED_PERIOD
where nA is the number of pulses for Time Slot A (Register 0x31,
Bits[15:8]).
tB (µs) = SLOTB_LED_OFFSET + nB × SLOTB_LED_PERIOD
The LED driver is a current sink and is agnostic to LED supply
voltage and LED type. The photodiode (PDx) inputs can
accommodate any photodiode with an input capacitance of less
than 100 pF. The ADPD103 is purposefully designed to produce
a high SNR for relatively low LED power while greatly reducing
the effect of ambient light on the measured signal.
where nB is the number of pulses for Time Slot B (Register 0x36,
Bits[15:8]).
Calculate the LED period using the following equation:
LED_PERIOD, minimum = 2 × AFE_WIDTH + 11
t1 and t2 are fixed and based on the computation time for each
slot. If a slot is not in use, these times do not add to the total
active time. Table 12 defines the values for these LED and
sampling time parameters.
ACTIVE
ACTIVE
tA
nA PULSES
t1
tB
t2
nB PULSES
TIME SLOT A
12722-008
SLEEP
TIME SLOT B
1/fSAMPLE
Figure 10. Time Slot Timing Diagram
Table 12. LED Timing and Sample Timing Parameters
Parameter
SLOTA_LED_OFFSET 1
SLOTB_LED_OFFSET1
SLOTA_LED_PERIOD 2
SLOTB_LED_PERIOD2
t1
t2
tSLEEP
1
2
Register
0x30
0x35
0x31
0x36
Bits
[7:0]
[7:0]
[7:0]
[7:0]
Test Conditions/Comments
Delay from power-up to LEDA rising edge
Delay from power-up to LEDB rising edge
Time between LED pulses in Time Slot A; SLOTx_AFE_WIDTH = 4 μs
Time between LED pulses in Time Slot B; SLOTx_AFE_WIDTH = 4 μs
Compute time for Time Slot A
Compute time for Time Slot B
Sleep time between sample periods
Min
23
23
19
19
68
20
222
Setting the SLOTx_LED_OFFSET below the specified minimum value may cause failure of ambient light rejection for large photodiodes.
Setting the SLOTx_LED_PERIOD below the specified minimum value can cause invalid data captures.
Rev. B | Page 13 of 52
Typ
Max
63
63
63
63
Unit
µs
µs
µs
µs
µs
µs
µs
ADPD103
Data Sheet
TIME SLOT SWITCH
PD1
Up to eight photodiodes (PD1 to PD8) can be connected to the
ADPD103. The photodiode anodes are connected to the PD1 to
PD8 input pins; the photodiode cathodes are connected to the
cathode pin, PDC. The anodes are assigned in three different
configurations depending on the settings of Register 0x14 (see
Figure 11, Figure 12, and Figure 13).
CH1
PD2
PD3
CH2
A switch sets which photodiode group is connected during Time
Slot A and Time Slot B. See Table 13 for the time slot switch
registers. When using less than eight photodiodes, it is important
to leave the unused inputs floating for proper operation of the
device. The photodiode inputs are current inputs and as such, these
pins are also considered to be voltage outputs. Tying these inputs to
a voltage may saturate the analog block.
PD4
Register 0x14, PD1 to PD8 Input Configurations
PD7
PD5
CH3
PD6
CH4
PD1
PD8
INPUT CONFIGURATION FOR
REGISTER 0x14[11:8] = 4
REGISTER 0x14[7:4] = 4
12722-110
CH1
PD2
Figure 12. PD5 to PD8 Connection
PD3
PD1
CH2
PD4
CH1
PD2
PD5
PD3
CH3
PD6
CH2
PD4
PD7
PD5
CH4
PD8
CH3
PD7
Figure 11. PD1 to PD4 Connection
CH4
PD8
INPUT CONFIGURATION FOR
REGISTER 0x14[11:8] = 1
REGISTER 0x14[7:4] = 1
12722-111
INPUT CONFIGURATION FOR
REGISTER 0x14[11:8] = 5
REGISTER 0x14[7:4] = 5
12722-109
PD6
Figure 13. 2-to-1 PD Current Summation
Rev. B | Page 14 of 52
Data Sheet
ADPD103
Table 13. Time Slot Switch (Register 0x14)
Address
0x14
Bits
[11:8]
Name
SLOTB_PD_SEL
[7:4]
SLOTA_PD_SEL
Description
Selects connection of photodiode for Time Slot B as shown in Figure 11, Figure 12,
and Figure 13.
0x0: inputs are floating in Time Slot B.
0x1: all PDx pins (PD1 to PD8) are connected during Time Slot B.
0x4: PD5 to PD8 are connected during Time Slot B.
0x5: PD1 to PD4 are connected during Time Slot B.
Other: reserved.
Selects connection of photodiode for Time Slot A as shown in Figure 11, Figure 12,
and Figure 13.
0x0: inputs are floating in Time Slot A.
0x1: All PDx pins (PD1 to PD8) are connected during Time Slot A.
0x4: PD5 to PD8 are connected during Time Slot A.
0x5: PD1 to PD4 are connected during Time Slot A.
Other: reserved.
ADJUSTABLE SAMPLING FREQUENCY
Register 0x12 controls the sampling frequency setting of the
ADPD103 and Register 0x4B, Bits[5:0] further tunes this clock
for greater accuracy. The sampling frequency is governed by an
internal 32 kHz sample rate clock that also drives the transition
of the internal state machine. The maximum sampling frequencies
for some sample conditions are listed in Table 3. The maximum
sample frequency for all conditions is determined by the
following equation:
fSAMPLE, MAX = 1/(tA + t1 + tB + t2 + tSLEEP, MIN)
If a given time slot is not in use, elements from that time slot do
not factor into the calculation. For example, if Time Slot A is not
in use, tA and t1 do not add to the sampling period and the new
maximum sampling frequency is calculated as follows:
fSAMPLE, MAX = 1/( tB + t2 + tSLEEP, MIN)
where tSLEEP, MIN is the minimum sleep time required between
samples. See the Dual Time Slot Operation section for the
definitions of tA, t1, tB, and t2.
4.
5.
The maximum frequency constraints also apply in this case.
Providing an External 32kHz Clock
The ADPD103 has an option for the user to provide an external
32 kHz clock to the device for system synchronization or for
situations where a clock with better accuracy than the internal
32 kHz clock is required. The external 32 kHz clock is provided
on the PDSO pin. To enable the 32 kHz external clock, use the
following procedure at startup:
1.
External Sync for Sampling
The ADPD103 provides an option to use an external sync signal
to trigger the sampling periods. This external sample sync signal
can be provided either on the INT pin or the PDSO pin. This
functionality is controlled by Register 0x4F, Bits[3:2]. When
enabled, a rising edge on the selected input specifies when the
next sample cycle occurs. When triggered, there is a delay of one to
two internal sampling clock (32 kHz) cycles, and then the normal
start-up sequence occurs. This sequence is the same as if the
normal sample timer provided the trigger. To enable the external
sync signal feature, use the following procedure:
1.
2.
3.
Write 0x1 to Register 0x10 to enter program mode.
Write the appropriate value to Register 0x4F, Bits[3:2] to
select whether the INT pin or the PDSO pin specifies when
the next sample cycle occurs. Also, enable the appropriate
input buffer using Register 0x4F, Bit 1, for the INT pin, or
Register 0x4F, Bit 5, for the PDSO pin.
Write b1 to EXT_SYNC_ENA, Register 0x38, Bit 14 to
enable the external sampling trigger.
Write 0x2 to Register 0x10 to start the sampling operations.
Apply the external sync signal on the selected pin at the
desired rate; sampling occurs at that rate. As with normal
sampling operations, read the data using the FIFO or the
data registers.
2.
3.
4.
5.
6.
Rev. B | Page 15 of 52
Drive the PDSO pin to a valid logic level or with the
desired 32 kHz clock prior to enabling the PDSO pin as an
input. Do not leave the pin floating prior to enabling it.
Write b1 to Register 0x4F, Bit 5 to enable the PDSO pin as
an input.
Write b11 to register 0x4B, Bit 7 and Bit 8 (CLK32K_EN
and CLK32K_BYP, respectively) to configure the device to
use an external 32 kHz clock.
Write 0x1 to Register 0x10 to enter program mode.
Write additional control registers in any order while the
device is in program mode to configure the device as
required.
Write 0x2 to Register 0x10 to start the normal sampling
operation.
ADPD103
Data Sheet
STATE MACHINE OPERATION
Program mode is used for programming registers. Always cycle
the ADPD103 through program mode when writing registers or
changing modes. Because no power cycling occurs in this mode,
the device may consume higher current in program mode than
in normal operation. To place the device in program mode,
write 0x1 to Register 0x10, Bits[1:0].
During each time slot, the ADPD103 operates according to a
state machine. The state machine operates in the following
sequence, shown in Figure 14.
STANDBY
REGISTER 0x10 = 0x0000
ULTRALOW POWER MODE
In normal operation, the ADPD103 pulses light and collects
data. Power consumption in this mode depends on the pulse
count and data rate. To place the device in normal sampling
mode, write 0x2 to Register 0x10, Bits[1:0].
NO DATA COLLECTION
ALL REGISTER VALUES ARE RETAINED.
NORMAL MODE OPERATION AND DATA FLOW
PROGRAM
In normal mode, the ADPD103 follows a specific pattern set up by
the state machine. This pattern is shown in the corresponding data
flow in Figure 15. The pattern is as follows:
REGISTER 0x10 = 0x0001
SAFE MODE FOR PROGRAMING REGISTERS
NO DATA COLLECTION
DEVICE IS FULLY POWERED IN THIS MODE.
1.
NORMAL OPERATION
12722-016
REGISTER 0x10 = 0x0002
LEDs ARE PULSED AND PHOTODIODES ARE SAMPLED
STANDARD DATA COLLECTION
DEVICE POWER IS CYCLED BY INTERNAL STATE MACHINE.
2.
Figure 14. State Machine Operation Flowchart
The ADPD103 operates in one of three modes: standby,
program, and normal sampling mode.
3.
Standby mode is a power saving mode in which no data collection
occurs. All register values are retained in this mode. To place
the device in standby mode, write 0x0 to Register 0x10,
Bits[1:0]. The device powers up in standby mode.
4.
LED pulse and sample.The ADPD103 pulses external LEDs.
The response of a photodiode or photodiodes to the reflected
light is measured by the ADPD103. Each data sample is
constructed from the sum of n individual pulses, where n
is user configurable between 1 and 255.
Intersample averaging. If desired, the logic can average n
samples, from 2 to 128 in powers of 2, to produce output
data. New output data is saved to the output registers every
N samples.
Data read. The host processor reads the converted results
from the data register or the FIFO.
Repeat. The sequence has a few different loops that enable
different types of averaging while keeping both time slots
close in time relative to each other.
[14 + LOG2(nA × NA)] BITS
UP TO 27 BITS
[14 + LOG2(nA)] BITS
UP TO 20 BITS
NA
NA
14 BITS
1
14 BITS
nA
nA
nA
14-BIT
ADC
1
ADC OFFSET
20-BIT CLIP
IF VAL ≤ (220 – 1)
VAL = VAL
ELSE VAL = 2 20 – 1
0
16-BIT CLIP
16 BITS
IF VAL ≤ (216 – 1)
VAL = VAL
ELSE VAL = 216 – 1
1
REGISTER
0x11[13]
[14 + LOG2(nA)] BITS
UP TO 22 BITS
32-BIT DATA
REGISTERS
FIFO
16-BIT
DATA
REGISTERS
SAMPLE 1: TIME SLOT A
SAMPLE 1: TIME SLOT B
0
1
SAMPLE N A: TIME SLOT A
SAMPLE N B: TIME SLOT B
NB
NOTES
1. nA AND nB = NUMBER OF LED PULSES FOR TIME SLOT A AND TIME SLOT B.
2. NA AND NB = NUMBER OF AVERAGES FOR TIME SLOT A AND TIME SLOT B.
NB
1
[14 + LOG2(nB)] BITS
UP TO 20 BITS
[14 + LOG2(nB × NB)] BITS
UP TO 27 BITS
Figure 15. ADPD103 Datapath
Rev. B | Page 16 of 52
16-BIT CLIP
IF VAL ≤ (216 – 1)
VAL = VAL
ELSE VAL = 2 16 – 1 16 BITS
12722-009
TIME SLOT A
TIME SLOT B
Data Sheet
ADPD103
select from 2, 4, 8 … up to 128 samples to be averaged. Pulse
data is still acquired by the AFE at the sampling frequency,
fSAMPLE (Register 0x12), but new data is written to the registers at
the rate of fSAMPLE/N every Nth sample. This new data consists of
the sum of the previous N samples. The full 32-bit sum is stored
in the 32-bit registers. However, before sending this data to the
FIFO, a divide by N operation occurs. This divide operation
maintains bit depth to prevent clipping on the FIFO.
LED Pulse and Sample
At each sampling period, the selected LED driver drives a series
of LED pulses, as shown in Figure 16. The magnitude, duration,
and number of pulses are programmable over the I2C interface.
Each LED pulse coincides with a sensing period so that the sensed
value represents the total charge acquired on the photodiode in
response to only the corresponding LED pulse. Charge, such as
ambient light, that does not correspond to the LED pulse is
rejected.
After each LED pulse, the photodiode output relating the pulsed
LED signal is sampled and converted to a digital value by the
14-bit ADC. Each subsequent conversion within a sampling
period is summed with the previous result. Up to 255 pulse values
from the ADC can be summed in an individual sampling period.
There is a 20-bit maximum range for each sampling period.
Use this between sample averaging to lower the noise while
maintaining 16-bit resolution. If the pulse count registers are
kept to 8 or less, the 16-bit width is never exceeded. Therefore,
when using Register 0x15 to average subsequent pulses, many
pulses can be accumulated without exceeding the 16-bit word
width. This can reduce the number of FIFO reads required by
the host processor.
Averaging
Data Read
The ADPD103 offers sample accumulation and averaging
functionality to increase signal resolution.
The host processor reads output data from the ADPD103, via
the I2C protocol, from the data registers or from the FIFO. New
output data is made available every N samples, where N is the user
configured averaging factor. The averaging factors for Time Slot A
and Time Slot B are configurable independently of each other. If
they are the same, both time slots can be configured to save data to
the FIFO. If the two averaging factors are different, only one time
slot can save data to the FIFO; data from the other time slot can
be read from the output registers.
Within a sampling period, the AFE can sum up to 256 sequential
pulses. As shown in Figure 15, samples acquired by the AFE are
clipped to 20 bits at the output of the AFE. Additional resolution,
up to 27 bits, can be achieved by averaging between sampling
periods. This accumulated data of N samples is stored as 27-bit
values and can be read out directly by using the 32-bit output
registers or the 32-bit FIFO configuration.
The data read operations are described in more detail in the
Reading Data section.
When using the averaging feature set up by the register,
subsequent pulses can be averaged by powers of 2. The user can
SHOWN WITH fSAMPLE = 10 Hz
OPTICAL SAMPLING
LOCATIONS
0.5
1.0
1.5
2.0
2.5
3.0
TIME (s)
LED
CURRENT
(ILED)
NUMBER OF LED PULSES (nA OR nB)
Figure 16. Example of a Photoplethysmography (PPG) Signal
Sampled at a Data Rate of 10 Hz Using Five Pulses per Sample
Rev. B | Page 17 of 52
12722-010
0
ADPD103
Data Sheet
AFE OPERATION
The timing within each pulse burst is important for optimizing
the operation of the ADPD103. Figure 17 shows the timing waveforms for a single time slot as an LED pulse response propagates
through the analog block of the AFE. The first graph, shown in
green, shows the ideal LED pulsed output. The filtered LED
response, shown in blue, shows the output of the analog integrator.
The third graph, shown in orange, illustrates an optimally placed
integration window. When programmed to the optimized value,
the full signal of the filtered LED response can be integrated. The
AFE integration window is then applied to the output of the bandpass filter (BPF) and the result is sent to the ADC and summed
for N pulses. If the AFE window is not correctly sized or located, all
of the receive signal is not properly reported and system perfor-
mance is not optimal; therefore, it is important to verify proper
AFE position for every new hardware design or the LED width.
AFE INTEGRATION OFFSET ADJUSTMENT
The AFE integration width must be equal or larger than the LED
width. As AFE width increases, the output noise increases and
the ability to suppress high frequency content from the environment decreases. It is therefore desirable to keep the AFE integration
width small. However, if the AFE width is too small, the LED
signal is attenuated. With most hardware selections, the AFE
width produces the optimal SNR at 1 μs more than the LED
width. After setting LED width, LED offset, and AFE width, the
ADC offset can then be optimized. The AFE offset must be
manually set such that the falling edge of the first segment of the
integration window matches the zero crossing of the filtered LED
response.
Figure 17. AFE Operation Diagram
Rev. B | Page 18 of 52
Data Sheet
ADPD103
LED_FALLING_EDGE = LED_OFFSET + LED_WIDTH
and,
AFE_INTEGRATION_FALLING_EDGE = 9 +
AFE_OFFSET + AFE_WIDTH
If both falling edges are set equal to each other, solve for
AFE_OFFSET to obtain the following equation:
AFE_OFFSET_STARTING_POINT = LED_OFFSET +
LED_WIDTH − 9 − AFE_WIDTH
Setting the AFE offset to any point in time earlier than the
starting point is equivalent to setting the integration in the
future; the AFE cannot integrate the result from an LED pulse
that has not yet occurred. As a result, an AFE_OFFSET value less
than the AFE_OFFSET_STARTING_POINT is an erroneous
setting. Such a result may indicate that current in the TIA is
operating in the reverse direction from the intended schematic,
where the LED pulse is causing the current to leave the TIA
rather than enter it.
Because, for most setups, the AFE_WIDTH is 1 µs wider than
the LED_WIDTH, the AFE_OFFSET_STARTING_POINT
value is typically 10 µs less than the LED_OFFSET value. Any
value less than LED_OFFSET – 10 is erroneous. The optimal
AFE offset is some time after the AFE_OFFSET_STARTING_
POINT. The band-pass filter response, LED response, and
photodiode response each add some delay. In general, the component choice, board layout, LED_OFFSET, and LED_WIDTH are
the variables that can change the AFE_OFFSET. After a specific
design is set, the AFE_OFFSET can be locked down and does
not need to be optimized further.
Sweeping the AFE Position
The AFE offsets for Time Slot A and Time Slot B are controlled
by Bits[10:0] of Register 0x39 and Register 0x3B, respectively.
Each LSB represents one cycle of the 32 MHz clock, or 31.25 ns.
100
0.687
95
90
85
80
75
0
0.15
0.30
0.45
0.60
0.75
0.90
1.05
1.20
AFE OFFSET FROM STARTING POINT (µs)
1.35
1.50
12722-113
The starting point of this offset, as expressed in microseconds, is set
such that the falling edge of the integration window aligns with the
falling edge of the LED.
The register can be thought of as 211 − 1 of these 31.25 ns steps, or
it can be broken into an AFE_COARSE setting using Bits[10:5]
to represent 1 µs steps and Bits[4:0] to represent 31.25 ns steps.
Sweeping the AFE position from the starting point to find a
local maximum is the recommended way to optimize the AFE
offset. The setup for this test is to allow the LED light to fall on
the photodiode in a static way. This is typically done with a
reflecting surface at a fixed distance. The AFE position can then
be swept to look for changes in the output level. When adjusting
the AFE position, it is important to sweep the position using the
31.25 ns steps. Typically, a local maximum is found within 2 µs
of the starting point for most systems. Figure 18 shows an example
of an AFE sweep, where 0 on the x-axis represents the AFE starting
point defined previously. Each data point in the plot corresponds to
one 31.25 ns step of the AFE_OFFSET. The optimal location for
AFE_OFFSET in this example is 0.687 µs from the AFE starting
point.
RELATIVE OUTPUT VALUE (%)
AFE Integration Offset Starting Point
Figure 18. AFE Sweep Example
Table 14 lists some typical LED and AFE values after optimization.
In general, it is not recommended to use the AFE_OFFSET
numbers in Table 14 without first verifying them against the AFE
sweep method. Repeat this method for every new LED width and
with every new set of hardware made with the ADPD103. For
maximum accuracy, it is recommended that the 32 MHz clock be
calibrated prior to sweeping the AFE.
Table 14. AFE Window Settings
LED Register 0x30 or Register 0x35
0x0219
0x0319
AFE Register 0x39 or Register 0x3B
0x19FB
0x21F4
Rev. B | Page 19 of 52
Comment
2 µs LED pulse, 3 µs AFE width, 25 µs LED delay
3 µs LED pulse, 4 µs AFE width, 25 µs LED delay
ADPD103
Data Sheet
I2C SERIAL INTERFACE
4.
The ADPD103 supports an I C serial interface via the SDA
(data) and SCL (clock) pins. All internal registers are accessed
through the I2C interface.
2
The ADPD103 conforms to the UM10204 I2C-Bus Specification
and User Manual, Rev. 05—9 October 2012, available from NXP
Semiconductors. It supports a fast mode (400 kbps) data transfer.
Register read and write are supported, as shown in Figure 19.
Figure 2 shows the timing diagram for the I2C interface.
Slave Address
The default 7-bit I2C slave address for the device is 0x64,
followed by the R/W bit. For a write, the default I2C slave
address is 0xC8; for a read, the default I2C address is 0xC9. The
slave address is configurable by writing to Register 0x09, Bits[7:1].
When multiple ADPD103 devices are on the same bus lines, the
INT and PDSO pins can be used to select specific devices for
the address change. Register 0x0D can be used to select a key to
enable address changes in specific devices. Use the following
procedure to change the slave address when multiple ADPD103
devices are connected to the same I2C bus lines:
1.
2.
3.
Using Register 0x4F, enable the input buffer of the PDSO pin,
the INT pin, or both, depending on the key being used.
For the device identified as requiring an address change,
set the INT and/or PDSO pins high or low to match the
key being used.
Write the SLAVE_ADDRESS_KEY using Register 0x0D,
Bits[15:0] to match the desired function. The allowed keys
are shown in Table 24.
5.
6.
Write the desired SLAVE_ADDRESS using Register 0x09,
Bits[7:1]. While writing to Register 0x09, Bits[7:1], write
0xAD to Register 0x09, Bit[15:8]. Register 0x09 must be
written to immediately after writing to Register 0x0D.
Repeat Step 1 to Step 4 for all the devices that need the
SLAVE_ADDRESS changed.
Set the INT and PDSO pins as desired for normal operation
using the new SLAVE_ADDRESS for each device.
I2C Write and Read Operations
Figure 19 illustrates the ADPD103 I2C write and read operations.
Single word and multiword read operations are supported. For a
single register read, the host sends a no acknowledge after the
second data byte is read and a new register address is needed for
each access.
For multiword operations, each pair of data bytes is followed by
an acknowledge from the host until the last byte of the last word is
read. The host indicates the last read word by sending a no
acknowledge. When reading from the FIFO (Register 0x60), the
data is automatically advanced to the next word in the FIFO
and the space is freed. When reading from other registers, the
register address is automatically advanced to the next register,
except at Register 0x5F or Register 0x7F, where the address does
not increment. This allows lower overhead reading of sequential
registers.
All register writes are single word only and require 16 bits (one
word) of data.
The software reset (Register 0x0F, Bit 0) is the only command
that does not return an acknowledge because the command is
instantaneous.
Table 15. Definition of I2C Terminology
Term
SCL
SDA
Master
Slave
Start (S)
Start (Sr)
Stop (P)
ACK
NACK
Slave Address
Read (R)
Write (W)
Description
Serial clock.
Serial address and data.
The master is the device that initiates a transfer, generates clock signals, and terminates a transfer.
The slave is the device addressed by a master. The ADPD103 operates as a slave device.
A high to low transition on the SDA line while SCL is high; all transactions begin with a start condition.
Repeated start condition.
A low to high transition on the SDA line while SCL is high. A stop condition terminates all transactions.
During the acknowledge or no acknowledge clock pulse, the SDA line is pulled low and remains low.
During the acknowledge or no acknowledge clock pulse, the SDA line remains high.
After a start (S), a 7-bit slave address is sent, which is followed by a data direction bit (read or write).
A 1 indicates a request for data.
A 0 indicates a transmission.
Rev. B | Page 20 of 52
Data Sheet
ADPD103
I2C WRITE
REGISTER WRITE
MASTER START
SLAVE ADDRESS + WRITE
SLAVE
DATA[15:8]
REGISTER ADDRESS
ACK
ACK
DATA[7:0]
STOP
ACK
ACK
I2C SINGLE WORD READ MODE
REGISTER READ
MASTER START
SLAVE ADDRESS + WRITE
SLAVE
REGISTER ADDRESS
ACK
ACK
SLAVE ADDRESS + READ
Sr
ACK
ACK
DATA[15:8]
ACK
DATA[15:8]
NACK
STOP
DATA[7:0]
I2C MULTIWORD READ MODE
REGISTER READ
SLAVE
SLAVE ADDRESS + WRITE
REGISTER ADDRESS
ACK
ACK/NACK
ACK
SLAVE ADDRESS + READ
Sr
ACK
STOP
DATA[7:0]
12722-012
MASTER START
DATA TRANSFERRED
n (DATA[15:8]+ACK+DATA[7:0] + ACK/NACK)
NOTES
1. THE SHADED AREAS REPRESENT WHEN THE DEVICE IS LISTENING.
Figure 19. I2C Write and Read Operations
A
B
Provide the 1.8 V supply, VDD, to AVDD and DVDD. Use single
(VLED) or multiple (VLED1, VLED2, and VLED3) sources for the LED
supply using standard regulator circuits according to the peak
current requirements specified in Table 3 and calculated in the
Calculating Current Consumption section.
C
LGND
LEDX2
LEDX3
LEDX1
SDA
SCL
INT
DVDD
3
AGND
DGND
PDSO
VREF
AVDD
F
PD5-8
PDC
PD1-4
12722-014
E
Figure 20. WLCSP Package Connection and PCB Layout Diagram (Top View)
VLED3
VLED2 VLED1
4.7µF
4.7µF
NIC
LEDX1
LEDX3
LEDX2
LGND
SCL
22
28
POWER-DOWN
CONTROL
The current input pins (PD1 to PD8) have a typical voltage of
1.3 V during the sampling period. During the sleep period,
Rev. B | Page 21 of 52
21
1
PDSO
NIC
DVDD
0.1µF
NIC
ADPD103
AGND
NIC
TOP VIEW
(Not to Scale)
0.1µF VREF
NIC
AVDD
0.1µF
PD1
NIC
NIC
7
15
PD8
14
PD7
PD6
PD5
8
12722-017
Figure 20 shows the recommended connection diagram and
printed circuit board (PCB) layout for the ADPD103 WLCSP
package. See Figure 21 or Figure 22 for connection details.
INT
PDC
1.8V
4.7µF
I2C BUS
SDA
TO/FROM
HOST PROCESSOR
PD2
With large photodiodes, the dynamic range can be increased by
splitting the current between multiple inputs. As a result, if only
one large photodiode is used and the receive signal is expected
to be large, the diode can be branched across all four inputs in a
given time slot. This type of configuration is shown in Figure 21.
For situations where the photodiode is small or the signal is greatly
attenuated, the photodiode can be connected directly to a single
channel such as PD1 or PD5. This connection, shown in Figure 22,
maximizes SNR for low signals. Do not connect the same
photodiode to all eight input channels. It is important to leave the
unused input channels floating for proper device operation. The
WLCSP package is internally wired for high dynamic range mode.
2
D
For best noise performance, connect AGND, DGND (exposed
pad), and LGND together at a large conductive surface such as a
ground plane, a ground pour, or a large ground trace.
The number of photodiodes or LEDs used varies. There are
multiple ways to connect photodiodes to the input channels,
as shown in Table 16 and Figure 23. The photodiode anodes are
connected to the PD1 to PD8 input pins, and the photodiode
cathodes are connected to the cathode pin.
1
PD4
Figure 21 and Figure 22 show two possible photodiode input
connections for the ADPD103. The 1.8 V I2C communication
lines, SCL and SDA, along with the INT line, connect to a
system microprocessor or sensor hub. The I2C signals can have
pull-up resistors connected to a 1.8 V or a 3.3 V power supply.
The INT and PDSO signals are only compatible with a 1.8 V
supply and may need a level translator.
these pins are connected to the cathode pin. The cathode and
anode voltages are listed in Table 3.
PD3
TYPICAL CONNECTION DIAGRAM
Figure 21. Connection Diagram for Increased Dynamic Range
ADPD103
Data Sheet
VLED2 VLED1
VLED3
4.7µF
NIC
LEDX1
LEDX3
LEDX2
LGND
SCL
22
28
POWER-DOWN
CONTROL
INT
0.1µF
NIC
21
1
PDSO
NIC
DVDD
1.8V
NIC
ADPD103
AGND
NIC
TOP VIEW
(Not to Scale)
0.1µF VREF
NIC
AVDD
0.1µF
4.7µF
4.7µF
I2C BUS
SDA
TO/FROM
HOST PROCESSOR
PD1
NIC
7
14
12722-013
PD7
PD6
PD5
PDC
PD4
PD3
8
PD2
PD8
15
Figure 22. Connection Options for Individual Single Channel Diodes
14
PD8
14
PD7
PD6
PD4
PDC
PD5
PD2
15
PD8
14
8
12722-020
PDC
PD4
PD3
15
8
PDC
PD5
7
15
8
PD2
14
8
7
PD7
7
PD1
15
PD6
PD4
PDC
7
PD3
14
8
PD1
PD1
15
PD5
14
8
7
PDC
PD1
PDC
15
PD3
7
PD2
PD1
Figure 23. Typical Photodiode Connection Diagram
Table 16. Typical Photodiode Anode to Input Channel Connections
Input Channel
Photodiode Anode Configuration
Single Photodiode (D1)
Two Photodiodes (D1, D2)
Four Photodiodes (D1 to D4)
Eight Photodiodes (D1 to D8)
1
PD1
D1
NC 1
D1
NC1
D1
D1
D1
NC1
D1
PD2
NC1
NC1
D1
NC1
NC1
D1
D2
NC1
D2
PD3
NC1
NC1
D1
NC1
NC1
D1
D3
NC1
D3
NC means do not connect under the conditions provided in Table 16. Leave all unused inputs floating.
Rev. B | Page 22 of 52
PD4
NC1
NC1
D1
NC1
NC1
D1
D4
NC1
D4
PD5
NC1
D1
NC1
D1
D2
D2
NC1
D1
D5
PD6
NC1
NC1
NC1
D1
NC1
D2
NC1
D2
D6
PD7
NC1
NC1
NC1
D1
NC1
D2
NC1
D3
D7
PD8
NC1
NC1
NC1
D1
NC1
D2
NC1
D4
D8
Data Sheet
ADPD103
LED DRIVER PINS AND LED SUPPLY VOLTAGE
The LEDx1, LEDx2, and LEDx3 pins have an absolute maximum
voltage rating of 3.6 V. Any voltage exposure over this rating affects
the reliability of the device operation and, in certain circumstances,
causes the device to cease proper operation. The voltage of the
LEDx pins must not be confused with the supply voltages for the
LED themselves (VLEDx). VLEDx is the voltage applied to the anode of
the external LED, whereas the LEDx pin is the input of the internal
current driver, and the pins are connected to the cathode of the
external LED.
LED DRIVER OPERATION
The LED driver for the ADPD103 is a current sink requiring
0.2 V of compliance above ground to maintain the programmed
current level. Figure 24 shows the basic schematic of how the
ADPD103 connects to an LED through the LED driver. The
Determining the Average Current and the Determining CVLED
sections define the requirements for the bypass capacitor
(CVLED) and the supply voltages of the LEDs (VLEDx).
DETERMINING CVLED
To determine the CVLED capacitor value, determine the maximum
forward-biased voltage, VFB_LED_MAX, of the LED in operation.
The LED current, IFB_LED_MAX, converts to VFB_LED_MAX as shown in
Figure 26. In this example, 250 mA of current through two green
LEDs in parallel yields VFB_LED_MAX = 3.95 V. Any series resistance
in the LED path must also be included in this voltage. When
designing the LED path, keep in mind that small resistances can
add up to large voltage drops due to the LED peak current being
very large. In addition, these resistances can be unnecessary
constraints on the VLEDx supply.
4.5
Figure 24. VLEDx Supply Schematic
DETERMINING THE AVERAGE CURRENT
The ADPD103 drives an LED in a series of short pulses. Figure 25
shows the typical ADPD103 configuration of a pulse burst
sequence.
19µs
3µs
TWO 528nm LEDs
ONE 850nm LED
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0
50
100
150
200
LED DRIVER CURRENT SETTING (mA)
ILED_MAX
12722-119
Figure 26. Example of the Average LED Forward-Biased Voltage Drop as a
Function of the Driver Current
Figure 25. Typical LED Pulse Burst Sequence Configuration
In this example, the LED pulse width, tLED_PULSE, is 3 µs, and the LED
pulse period, tLED_PERIOD, is 19 µs. The LED being driven is a pair
of green LEDs driven to a 250 mA peak. The goal of CVLED is to
buffer the LED between individual pulses. In the worst case
scenario, where the pulse train shown in Figure 25 is a continuous
sequence of short pulses, the VLEDx supply must supply the
average current. Therefore, calculate ILED_AVERAGE as follows:
ILED_AVERAGE = (tLED_PULSE/tLED_PERIOD) × ILED_PEAK
250
12722-120
LEDx
VLEDx
SUPPLY
For the numbers shown in Equation 1, ILED_AVERAGE = 3/19 ×
ILED_PEAK. For typical LED timing, the average VLEDx supply
current is 3/19 × 250 mA = 39.4 mA, indicating that the VLEDx
supply must support a dc current of 40 mA.
LED FORWARD-BIAS VOLTAGE DROP (V)
LGND
CVLED
12722-118
ADPD103
where:
ILED_AVERAGE is the average current needed from the VLEDx supply
during the pulse period, and it is also the VLEDx supply current
rating.
ILED_PEAK is peak current setting of the LED.
(1)
To correctly size the CVLED capacitor, do not deplete it during the
pulse of the LED to the point where the voltage on the capacitor is
less than the forward bias on the LED.
To calculate the minimum value for the VLEDx bypass capacitor, use
the following equation:
CVLED =
t LED_PULSE × I FB _ LED _ MAX
VLED _ MIN − (VFB _ LED _ MAX + 0.2)
(2)
where:
tLED_PULSE is the LED pulse width.
IFB_LED_MAX is the maximum forward-biased current on the LED
used in operating the device.
VLED_MIN is the lowest voltage from the VLEDx supply with no load.
VFB_LED_MAX is the maximum forward-biased voltage required on
the LED to achieve ILED_PEAK.
Rev. B | Page 23 of 52
ADPD103
Data Sheet
The numerator of the CVLED equation sets up the total discharge
amount in coulombs from the bypass capacitor to satisfy a
single programmed LED pulse of the maximum current. The
denominator represents the difference between the lowest voltage
from the VLEDx supply and the LED required voltage. The LED
required voltage is the voltage of the anode of the LED such that
the 0.2 V compliance of the LED driver and the forward-biased
voltage of the LED operating at the maximum current is satisfied.
For a typical ADPD103 example, assume that the lowest value
for the VLEDx supply is 4.4 V, and that the peak current is 250 mA
for two 528 nm LEDs in parallel. The minimum value for CVLED
is then equal to 3 µF.
CVLED = (3 × 10−6 × 0.250)/(4.4 – (3.95 + 0.2)) = 3 µF
(3)
As shown in the Equation 3, as the minimum supply voltage drops
close to the maximum anode voltage, the demands on CVLED
become more stringent, forcing the capacitor value higher. It is
important to insert the correct values into these equations. For
example, using an average value for VLED_MIN instead of the
worst case value for VLED_MIN can cause a serious design deficiency,
resulting in a CVLED value that is too small and that causes
insufficient optical power in the application. Therefore, adding a
sufficient margin on CVLED is strongly recommended. Add
additional margin to CVLED to account for derating of the capacitor
value over voltage, bias, temperature and other factors over the life
of the component.
LED INDUCTANCE CONSIDERATIONS
The LED drivers (LEDXx) on the ADPD103 have configurable
slew rate settings (Register 0x22, Bits[6:4], Register 0x23, Bits[6:4],
and Register 0x24, Bits[6:4]). These slew rates are defined in
Table 3. Even at the lowest setting, careful consideration must be
taken in board design and layout. If a large series inductor, such as a
long PCB trace, is placed between the LED cathode and one of
the LEDXx pins, voltage spikes from the switched inductor can
cause violations of absolute maximum and minimum voltages
on the LEDXx pins during the slew portion of the LED pulse.
To verify that there are no voltage spikes on the LEDXx pins due
to parasitic inductance, use an oscilloscope on the LEDXx pins to
monitor the voltage during normal operation. Any positive spike
>3.6 V may damage the device.
In addition, a negative spike <−0.3 V may also damage the device.
RECOMMENDED START-UP SEQUENCE
At power-up, the device is in standby mode (Register 0x10 =
0x0), as shown in Figure 14. The ADPD103 does not require a
particular power-up sequence.
From standby mode, to begin measurement, initiate the
ADPD103 as follows:
1.
2.
3.
4.
Set the CLK32K_EN bit (Register 0x4B, Bit 7) to start the
sample clock (32 kHz clock). This clock controls the state
machine. If this clock is off, the state machine is not able to
transition as defined by Register 0x10.
Write 0x1 to Register 0x10 to force the device into program
mode. Step 1 and Step 2 can be swapped, but the actual
state transition does not occur until both steps occur.
Write additional control registers in any order while the
device is in program mode to configure the device as
required.
Write 0x2 to Register 0x10 to start normal sampling
operation.
To terminate normal operation, follow this sequence to place
the ADPD103 in standby mode:
1.
2.
3.
4.
5.
Write 0x1 to Register 0x10 to force the device into program
mode.
Write to the registers in any order while the device is in
program mode.
Write 0x00FF to Register 0x00 to clear all interrupts. If
desired, clear the FIFO as well by setting the DIGITAL_
CLOCK_ENA bit (Register 0x5F, Bit 0) and writing 0x80FF
to Register 0x00.
Write 0x0 to Register 0x10 to force the device into standby
mode.
Optionally, stop the 32 kHz clock by resetting the CLK32K_
EN bit (Register 0x4B, Bit 7). Register 0x4B, Bit 7 = 0 is the
only write that must be written when the device is in standby
mode (Register 0x10 = 0x0). If 0 is written to this bit while in
program mode or normal mode, the device becomes
unable to transition into any other mode, including standby
mode, even if it is subsequently written to do so. As a result,
the power consumption in what appears to be standby mode
is greatly elevated. For this reason, and due to the very low
current draw of the 32 kHz clock while in operation, it is
recommended from an ease of use perspective to keep the
32 kHz clock running after it is turned on.
READING DATA
The ADPD103 provides multiple methods for accessing the sample
data. Each time slot can be independently configured to provide
data access using the FIFO or the data registers. Interrupt signaling
is also available to simplify timely data access. The FIFO is available
to loosen the system timing requirements for data accesses.
Reading Data Using the FIFO
The ADPD103 includes a 128-byte FIFO memory buffer that
can be configured to store data from either or both time slots.
Register 0x11 selects the kind of data from each time slot to be
Rev. B | Page 24 of 52
Data Sheet
ADPD103
c. Read a complete packet using one or more multiword
accesses using Register 0x60. Reading the FIFO
automatically frees the space for new samples.
d. Write 0 to the FIFO_ACCESS_ENA bit (Register 0x5F,
Bit 0).
written to the FIFO. Note that both time slots can be enabled to
use the FIFO, but only if their output data rate is the same.
Output data rate = fSAMPLE/N
where:
fSAMPLE is the sampling frequency.
N is the averaging factor for each time slot (NA for Time Slot A
and NB for Time Slot B). In other words, NA = NB must be true
to store data from both time slots in the FIFO.
Data packets are written to the FIFO at the output data rate. A
data packet for the FIFO consists of a complete sample for each
enabled time slot. Data for each photodiode channel can be stored
as either 16 or 32 bits. Each time slot can store 2, 4, 8, or 16
bytes of data per sample, depending on the mode and data
format. To ensure that data packets are intact, new data is only
written to the FIFO if there is sufficient space for a complete
packet. Any new data that arrives when there is not enough
space is lost. The FIFO continues to store data when sufficient
space exists. Always read FIFO data in complete packets to
ensure that data packets remain intact.
The number of bytes currently stored in the FIFO is available in
Register 0x00, Bits[15:8]. A dedicated FIFO interrupt is also
available and automatically generates when a specified amount
of data is written to the FIFO.
Interrupt-Based Method
To read data from the FIFO using an interrupt-based method,
use the following procedure:
1.
2.
3.
4.
5.
6.
In program mode, set the configuration of the time slots as
desired for operation.
Write Register 0x11 with the desired data format for each
time slot.
Set FIFO_THRESH in Register 0x06, Bits[13:8] to the
interrupt threshold. A good value for this is the number of
16-bit words in a data packet, minus 1. This causes an
interrupt to generate when there is at least one complete
packet in the FIFO.
Enable the FIFO interrupt by writing a 0 to the FIFO_
INT_MASK in Register 0x01, Bit 8. Also, configure the
interrupt pin (INT) by writing the appropriate value to the
bits in Register 0x02.
Enter normal operation mode by setting Register 0x10 to
0x2.
When an interrupt occurs
a. There is no requirement to read the FIFO_SAMPLES
register, because the interrupt is generated only if there
is one or more full packets. Optionally, the interrupt
routine can check for the presence of more than one
available packet by reading this register.
b. Write 1 to the FIFO_ACCESS_ENA bit (Register 0x5F,
Bit 0) twice in two consecutive write operations.
The interrupt automatically clears when enough data is read
from the FIFO to bring the data level below the threshold.
Polling Method
To read data from the FIFO in a polling method, use the
following procedure:
1.
2.
3.
In program mode, set the configuration of the time slots as
desired for operation.
Write Register 0x11 with the desired data format for each
time slot.
Enter normal operation mode by setting Register 0x10 to 2.
Next, begin the polling operations.
1. Wait for the polling interval to expire.
2. Read the FIFO_SAMPLES bits (Register 0x00, Bits[15:8]).
3. If FIFO_SAMPLES ≥ the packet size, read a packet using
the following steps:
a. Write 1 to the FIFO_ACCESS_ENA bit (Register 0x5F,
Bit 0) twice in two consecutive write operations.
b. Read a complete packet using one or more multiword
accesses using Register 0x60. Reading the FIFO
automatically frees the space for new samples.
c. Write 0 to the FIFO_ACCESS_ENA bit (Register 0x5F,
Bit 0).
d. Repeat Step 1.
When a mode change is required, or any other disruption to
normal sampling is necessary, the FIFO must be cleared. Use
the following procedure to clear the state and empty the FIFO:
1.
2.
3.
4.
Enter program mode by setting Register 0x10 to 0x1.
Write 1 to the FIFO_ACCESS_ENA bit (Register 0x5F,
Bit 0) twice in two consecutive write operations.
Write 1 to Register 0x00, Bit 15.
Write 0 to the FIFO_ACCESS_ENA bit (Register 0x5F, Bit 0).
Reading Data from Registers Using Interrupts
The latest sample data is always available in the data registers and is
updated simultaneously at the end of each time slot. The data value
for each photodiode channel is available as a 16-bit value in Register 0x64 through Register 0x67 for Time Slot A, and Register 0x68
through Register 0x6B for Time Slot B. If allowed to reach their
maximum value, Register 0x64 through Register 0x6B clip. If
Register 0x64 through Register 0x6B saturate, the unsaturated (up
to 27 bits) values for each channel are available in Register 0x70
through Register 0x77 for Time Slot A and Register 0x78 through
Register 0x7F for Time Slot B. Sample interrupts are available to
Rev. B | Page 25 of 52
ADPD103
Data Sheet
indicate when the registers are updated and can be read. To use
the interrupt for a given time slot, use the following procedure:
1.
2.
3.
4.
Enable the sample interrupt by writing a 0 to the appropriate
bit in Register 0x01. To enable the interrupt for Time Slot A,
write 0 to Bit 5. To enable the interrupt for Time Slot B,
write 0 to Bit 6. Either or both interrupts can be set.
Configure the interrupt pin by writing the appropriate
value to the bits in Register 0x02.
An interrupt generates when the data registers are updated.
The interrupt handler must perform the following:
a. Read Register 0x00 and observe Bit 5 or Bit 6 to confirm
which interrupt has occurred. This step is not required if
only one interrupt is in use.
b. Read the data registers before the next sample can be
written. The system must have interrupt latency and
service time short enough to respond before the next
data update, based on the output data rate.
c. Write a 1 to Bit 5 or Bit 6 in Register 0x00 to clear the
interrupt.
If both time slots are in use, it is possible to use only the
Time Slot B interrupt to signal when all registers can be read. It
is recommended to use the multiword read to transfer the data
from the data registers.
Reading Data from Registers Without Interrupts
Heart rate monitoring applications require an accurate time
base to achieve an accurate count of beats per minute. The
ADPD103 provides a simple calibration procedure for both clocks.
1.
2.
If the system interrupt response is not fast or predictable
enough to use the interrupt method, or if the interrupt pin is
not used, it is possible to get reliable data access by using the
data hold mechanism. To guarantee that the data read from the
registers is from the same sample time, it is necessary to prevent
the update of samples while reading the current values. The
method for doing register reads without interrupt timing is as
follows:
1.
2.
3.
Write a 1 to SLOTA_DATA_HOLD or SLOTB_DATA_
HOLD (Register 0x5F, Bit 1 and Bit 2, respectively) for the
time slot requiring access (both time slots can be accessed).
This prevents sample updates.
Read the registers as desired.
Write a 0 to the SLOTA_DATA_HOLD or SLOTB_DATA_
HOLD bits (Register 0x5F, Bit 1 and Bit 2, respectively)
previously set. Sample updates are allowed again.
Because a new sample may arrive while the reads are occurring,
this method prevents the new sample from partially overwriting
the data being read.
CLOCKS AND TIMING CALIBRATION
The ADPD103 operates using two internal time bases: a 32 kHz
clock sets the sample timing, and a 32 MHz clock controls the
timing of the internal functions such as LED pulsing and data
capture. Both clocks are internally generated and exhibit deviceto-device variation of approximately 10% (typical).
Calibrating the 32 kHz clock. This calibrates items
associated with the output data rate. Calibration of this
clock is important for applications where an accurate data
rate is important, such as heart rate measurements.
a. Set the sampling frequency to the highest the system
can handle, such as 2000 Hz. Because the 32 kHz clock
controls sample timing, its frequency is readily accessible
via the INT pin. Configure the interrupt by writing the
appropriate value to the bits in Register 0x02 and set the
interrupt to occur at the sampling frequency by writing 0
to Register 0x01, Bit 5 or Bit 6. Monitor the INT pin.
The interrupt frequency must match the set sample
frequency.
b. If the monitored interrupt frequency is less than the set
sampling frequency, increase the CLK32K_ADJUST bit
(Register 0x4B, Bits[5:0]). If the monitored interrupt
frequency is larger than the set sampling frequency,
decrease the CLK32K_ADJUST bits.
c. Repeat Step b until the monitored interrupt signal
frequency is close enough to the set sampling frequency.
Calibrate the 32 MHz clock. This calibrates items associated
with the fine timing within a sample period, such as LED
pulse width and spacing, assuming that the 32 kHz clock
has been calibrated.
a. Write 0x1 to Register 0x5F, Bit 0.
b. Enable the CLK_RATIO calculation by writing 0x1 to
Register 0x50, Bit 5. This function counts the number of
32 MHz clock cycles in two cycles of the 32 kHz clock.
With this function enabled, this cycle value is stored in
Register 0xA, Bits[11:0] and nominally this ratio is 2000
(0x7D0).
c. Calculate the 32 MHz clock error as follows:
Clock Error = 32 MHz × (1 − CLK_RATIO/2000)
d. Adjust the frequency by setting Bits[7:0] in Register
0x4D per the following equation:
CLK32M_ADJUST = Clock Error/109 kHz
e. Write 0x0 to Register 0x50, Bit 5 to reset the
CLK_RATIO function.
Repeat Step 2b through Step 2e until the desired accuracy is
achieved.
Write 0x0 to Register 0x5F, Bit 0. Also, set the INT pin back to
the mode desired for normal operation.
Rev. B | Page 26 of 52
Data Sheet
ADPD103
CALCULATING CURRENT CONSUMPTION
The current consumption of the ADPD103 depends on the user
selected operating configuration, as described in the following
equations.
where LEDAPEAK is LED1PEAK, LED2PEAK, or LED3PEAK, expressed
in mA, for whichever LED is selected for Time Slot A.
Average VLEDB Supply Current
To calculate the average VLEDB supply current, use Equation 9.
Total Power Consumption
ILED_AVG_B = (SLOTB_LED_WIDTH/1 × 106) × LEDBPEAK ×
DR × PULSE_COUNT
(9)
To calculate the total power consumption, use Equation 4.
Total Power = IVDD _ AVG × VDD + I LEDA _ AVG × VLEDA +
(4)
I LEDB _ AVG × VLEDB
where LEDBPEAK is LED1PEAK, LED2PEAK, or LED3PEAK, expressed
in mA, for whichever LED is selected for Time Slot B.
Average VDD Supply Current
OPTIMIZING SNR PER WATT
To calculate the average VDD supply current, use Equation 5.
The ADPD103 offers a variety of parameters that the user can
adjust to achieve the best signal. One of the key goals of system
performance is to obtain the best system SNR for the lowest
total power. This is often referred to as optimizing SNR/watt. Even
in systems where only the SNR matters and power is a secondary
concern, there may be a lower power or a high power means of
achieving the same SNR.
IVDD _ AVG = DR × (( I AFE _ A × t SLOTA ) + ( I AFE _ B × t SLOTB )
(5)
+ QPROC ) + IVDD _ STANDBY
where:
DR = the data rate in Hz.
IVDD_STANDBY = 3.5 × 10−3 mA.
QPROC is an average charge associated with a processing time, as
follows:
•
•
•
Only Time Slot A enabled: QPROC = 0.64 × 10 mC
Only Time Slot B enabled: QPROC = 0.51 × 10−3 mC
Time Slot A and Time Slot B enabled: QPROC = 0.69 ×
10−3 mC
−3
I AFE _ x (mA) = 2.9 + (1.5 × NUM _ CHANNELS) +
(6)
(LEDx PEAK − 25) / 225
t SLOTx (sec) = LEDx _ OFFSET + LEDx _ PERIOD ×
(7)
PULSE _ COUNT
where:
NUM_CHANNELS is the number of active channels.
LEDxPEAK is the peak LED current expressed in mA.
LEDx_OFFSET is the pulse start time offset expressed in
seconds.
LEDx_PERIOD is the pulse period expressed in seconds.
PULSE_COUNT is the number of pulses.
Note that if either Time Slot A or Time Slot B are disabled, IAFE_x
= 0 for that respective time slot. Additionally, if operating in
digital integrate mode, power savings can be realized by setting
Register 0x3C, Bits[8:3] = b010010. This setting disables the
band-pass filters that are bypassed in digital integrate mode,
changing the AFE power contribution calculation to:
I AFE _ x ( mA) = 2.9 + (1.0 × NUM _ CHANNELS ) +
( LEDx PEAK − 25) / 225
Average VLEDA Supply Current
To calculate the average VLEDA supply current, use Equation 8.
ILED_AVG_A = (SLOTA_LED_WIDTH/1 × 106) × LEDAPEAK ×
DR × PULSE_COUNT
(8)
Optimizing for Peak SNR
The first step in optimizing for peak SNR is to find a TIA gain
and LED level that gives the best performance where the
number of LED pulses remains constant. If peak SNR is the
goal, the noise section of Table 3 can be used as a guide. It is
important to note that the SNR improves as a square root of the
number of pulses averaged together, whereas the increase in the
LED power consumed is directly proportional to the number of
LED pulses. In other words, for every doubling of the LED pulse
count, there is a doubling of the LED power consumed and a 3 dB
SNR improvement. As a result, avoid any change in the gain
configuration that provides less than 3 dB of improvement for a 2×
power penalty; any TIA gain configuration that provides more
than 3 dB of improvement for a 2× power penalty is a good
choice. If peak SNR is the goal and there is no issue saturating
the photodiode with LED current at any gain, the 50k TIA gain
setting is an optimal choice. After the SNR per pulse per
channel is optimized, the user can then increase the number of
pulses to achieve the desired system SNR.
Optimizing SNR per Watt in a Signal Limited System
In practice, optimizing for peak SNR is not always practical.
One scenario in which the photoplethysmography (PPG) signal
has a poor SNR is the signal limited regime. In this scenario, the
LED current reaches an upper limit before the desired dc return
level is achieved.
Tuning in this case starts where the peak SNR tuning stops. The
starting point is nominally a 50k gain, as long as the lowest LED
current setting of 8 mA does not saturate the photodiode and
the 50k gain provides enough protection against intense background light. In these cases, use a 25k gain as the starting point.
The goal of the tuning process is to bring the dc return signal to
a specific ADC range, such as 50% or 60%. The ADC range
Rev. B | Page 27 of 52
ADPD103
Data Sheet
choice is a function of the margin of headroom needed to
prevent saturation as the dc level fluctuates over time. The SNR
of the PPG waveform is always some percentage of the dc level.
If the target level cannot be achieved at the base gain, increase
the gain and repeat the procedure. The tuning system may need
to place an upper limit on the gain to prevent saturation from
ambient signals.
Tuning the Pulse Count
After the LED peak current and TIA gain are optimized,
increasing the number of pulses per sample increases the SNR
by the square root of the number of pulses. There are two ways to
increase the pulse count. The pulse count registers (Register 0x31,
Bits[15:8], and Register 0x36, Bits[15:8]) change the number of
pulses per internal sample. Register 0x15, Bits[6:4] and Bits[10:8],
controls the number of internal samples that are averaged together
before the data is sent to the output. Therefore, the number of
pulses per sample is the pulse count register multiplied by the
number of subsequent samples being averaged. In general, the
internal sampling rate increases as the number of internal
sample averages increase to maintain the desired output data
rate. The SNR/watt is most optimal with pulse count values of
16 or less. Above pulse count values of 16, the square root
relationship does not hold in the pulse count register. However,
this relationship continues to hold when averaged between
samples using Register 0x15.
Note that increasing LED peak current increases SNR almost
directly proportional to LED power, whereas increasing the
number of pulses by a factor of n results in only a nominal√(n)
increase in SNR.
When using the sample sum/average function (Register 0x15),
the output data rate decreases by the number of summed
samples. To maintain a static output data rate, increase the
sample frequency (Register 0x12) by the same factor as that
selected in Register 0x15. For example, for a 100 Hz output data
rate and a sample sum/average of four samples, set the sample
frequency to 400 Hz.
SINGLE AFE CHANNEL MODE
When using a single photodiode in an application, and that
photodiode is connected to a single AFE channel (see Table 16),
the ADPD103 has an option to power down Channel 2, Channel 3,
and Channel 4, which places the device in single AFE channel
mode. Because three of the four AFE channels are turned off in this
mode, the power consumption is considerably reduced.
It is important to leave the unused input channels floating for
proper device operation. To run the device in single AFE channel
mode, write 0x38 to Register 0x3C, Bits[8:3]. If it is not required
to run the device in single AFE channel mode, leave Register 0x3C,
Bits[8:3] at 0x00.
TIA_ADC MODE
There is a way to put the device into a mode that effectively runs
the TIA directly in the ADC without using the analog band-pass
filter and integrator. This mode is referred to as TIA_ADC mode.
There are two basic applications of TIA_ADC mode. In normal
operation, all of the background light is blocked from the signal
chain, and therefore cannot be measured. TIA_ADC mode can be
used to measure the amount of background/ambient light. This
mode can also be used to measure other dc input currents, such as
leakage resistance.
When the device is in TIA_ADC mode, the band-pass filter and
the integrator stage are bypassed. This effectively wires the TIA
directly into the ADC. At the set sampling frequency, the ADC
samples Channel 1 through Channel 4 (or Channel 5 through
Channel 8) in sequential order, and each sample is taken at 1 µs
intervals. The TIA is in an inverting configuration; therefore, the
signal drops as more light hits the photodiode. Zero light or dark
conditions result in approximately 13,000 LSBs from the ADC.
To put the ADPD103 in TIA_ADC mode during Time Slot A,
write 0xB065 to Register 0x43 to bypass the band-pass filter and
integrator. Similarly, to place the ADPD103 in TIA_ADC mode
during Time Slot B, write 0xB065 to Register 0x45. One way to
monitor dc and pulsed signal at the same time is to operate
TIA_ADC mode in one time slot and pulse mode in the other
time slot. In TIA_ADC mode, increasing light level causes a
decrease in ADC codes because the TIA stage is inverting.
Protecting Against TIA Saturation in Normal Operation
One of the reasons to monitor TIA_ADC mode is to protect
against environments that may cause saturation. One concern
when operating in high light conditions, especially with larger
photodiodes, is that the TIA stage may become saturated and
the ADPD103 continues to communicate data. The resulting
saturation is not typical. The TIA, based on its settings, can only
handle a certain level of photodiode current. Based on the way
the ADPD103 is configured, if there is a current level from the
photodiode that is larger than the TIA can handle, the TIA output
during the LED pulse effectively extends the current pulse,
making it wider. The AFE timing is then violated because the
positive portion of the band-pass filter output extends into the
negative section of the integration window. Thus, the photosignal
is subtracted from itself, causing the output signal to decrease
when the effective light signal increases.
To measure the response from the TIA and verify that this stage
is not saturating, place the device in TIA_ADC mode and slightly
modify the timing. Specifically, sweep SLOTx_AFE_OFFSET
until two or three of the four channels reach a minimum value
(note that TIA is in an inverting configuration). All four
channels do not reach this minimum value because, typically, 3 µs
LED pulse widths are used and the ADC samples the four channels
sequentially at 1 µs intervals. This procedure aligns the ADC
Rev. B | Page 28 of 52
Data Sheet
ADPD103
sampling time with the LED pulse to measure the total amount of
light falling on the photodetector (for example, background light +
LED pulse).
If this minimum value is above 0 LSB, the TIA is not saturated.
However, take care, because even if the result is not 0 LSB,
operating the device near saturation can quickly result in
saturation if light conditions change. A safe operating region is
typically at ¾ full scale and lower. Use Table 17 to determine
how the input codes map to ADC levels on a per channel per
pulse basis. These codes are not the same as in normal mode
because the band-pass filter and integrator are not unity-gain
elements.
Coarse Ambient Light Measurement
Using the typical values in Table 17, TIA_ADC mode can be used
to measure or quantify the amount of background or ambient light
present on the photodetector. The settings are the same in the
method described in the Protecting Against TIA Saturation in
Normal Operation section, except the timing used in the
normal operating mode is sufficient for this mode. There is no
need to sweep AFE_OFFSET. If AFE_OFFSET is in the same place
as the normal mode operation, the TIA_ADC mode does not
return the same value, regardless of whether the LED is on or off.
In TIA_MODE, the dark level is a high level near 13,000 LSBs
per channel per pulse (see Table 17). To measure this value,
select no PD by writing a 0x0 to Register 0x14, Bits[11:8] for
Time Slot B or Register 0x14, Bits[7:4] for Time Slot A. This
setting internally opens the photodiode connection. This gives a
baseline LSB value that coincides with a zero signal input.
After Register 0x14 is restored to its normal value, while
connecting the photodiode to the TIA, this TIA_ADC result
can be subtracted from the open photodiode case to yield a
background light measurement. Use Table 17 to translate this
measurement into an input photocurrent. Use this result for
coarse absolute measurements only, because it is typically only
accurate to within 10%.
Measuring PCB Parasitic Input Resistance
During the process of mounting the ADPD103, undesired
resistance can develop on the inputs through assembly errors or
debris on the PCB. These resistances can form between the
anode and cathode, or between the anode and some other
supply or ground. In normal operation, the ambient rejection
feature of the ADPD103 masks the primary effects of these
resistances, making it very difficult to detect them. However,
even at 1 MΩ to 10 MΩ, such resistance can impact performance
significantly through added noise or decreased dynamic range.
TIA_ADC mode can be used to screen for these assembly issues.
Measuring Shunt Resistance on the Photodiode
A shunt resistor across the photodiode does not generally affect
the output level of the device in operation because the effective
impedance of the TIA is very low. This is especially true if the
photodiode is held to 0 V in operation. However, such resistance
can add noise to the system, degrading performance. The best
way to detect photodiode leakage, also called photodiode shunt
resistance, is to place the device in TIA_ADC mode in the dark
and vary the operation mode cathode voltage. When the cathode is
at 1.3 V, this places 0 V across the photodiode because the anode is
always at 1.3 V while in operation. When the cathode is at 1.8 V,
this places 0.5 V across the photodiode. Using the register settings
in Table 3 to control the cathode voltage, measure the TIA_ADC
value at both voltages. Next, divide the voltage difference of 0.5 V
by the difference of the ADC result after converting it to a current.
This result is the approximate shunt resistance. Values greater
than 10 MΩ may be difficult to measure, but this method is useful
in identifying gross failures.
Table 17. Analog Specifications for TIA_ADC and Digital Integrate Modes
Parameter
TIA_ADC/Digital Integration Saturation Levels
TIA_ADC Resolution
Output with No Input Current
Test Conditions/Comments
Values expressed per channel, per sample
TIA Feedback Resistor:
25 kΩ
50 kΩ
100 kΩ
200 kΩ
Values expressed per channel, per sample
TIA Feedback Resistor:
25 kΩ
50 kΩ
100 kΩ
200 kΩ
ADC offset (Register 0x18 to Register 0x21) = 0x0
Rev. B | Page 29 of 52
Typ
Unit
38.32
19.16
9.58
4.79
μA
μA
μA
μA
2.92
1.5
0.73
0.37
13000
nA/LSB
nA/LSB
nA/LSB
nA/LSB
LSB
ADPD103
Data Sheet
Measuring TIA Input Shunt Resistance
Another problem that can occur is for a resistance to develop
between the TIA input and another supply or ground on the
PCB. These resistances can force the TIA into saturation
prematurely. This, in turn, takes away dynamic range from the
device in operation and adds a Johnson noise component to the
input. To measure these resistances, place the device in TIA_ADC
mode in the dark and start by measuring the TIA_ADC offset level
with the photodiode inputs disconnected (Register 0x14,
Bits[11:8] = 0 or Register 0x14, Bits[7:4] = 0). From this,
subtract the value of TIA_ADC mode with the darkened
photodiode connected and convert the difference into a current.
If the value is positive, and the ADC signal decreased, the
resistance is to a voltage higher than 1.3 V, such as VDD. Current
entering the TIA causes the output to drop. If the output
difference is negative due to an increase of codes at the ADC,
current is being pulled out of the TIA and there is a shunt
resistance to a lower potential than 1.3 V, such as ground.
DIGITAL INTEGRATE MODE
Digital integrate mode is built into the ADPD103 and allows the
device to accommodate longer LED/AFE pulse widths and
different types of sensors at the input. The analog integration
mode described in the AFE Operation section is ideally suited
for applications requiring a large LED duty cycle, or applications
that require customization of the sampling scheme. Digital
integrate mode allows the integration function to be performed
after the ADC in the digital domain. This mode enables the
device to handle a much wider range of sensors at the input.
In digital integrate mode, the ADC performs a conversion every
1 µs during the integration window. During the integration
window, the digital engine either adds to or subtracts from the
previous sample. The band-pass filter is bypassed and the
integrator is converted to a voltage buffer, allowing the digital
engine to perform the integration function. In this mode, after
the timing is optimized, the output of the ADC increases as the
light level on the photodiode increases.
The integration window is a combination of negative and positive
windows where the duration of these windows is set by SLOTx_
AFE_WIDTH. At the end of the digital integration window, the
resulting sum is sent to the decimate unit as the sample for that
LED pulse. There is one sample per time slot for every sample
cycle. Table 18 lists the registers required for placing the device
in digital integrate mode.
There may also be changes needed in the SLOTx_AFE_OFFSET
registers and FIFO configuration register (0x11). To read the
final value through the FIFO, set the appropriate values in Register 0x11, Bits[4:2] for Time Slot A, and Register 0x11, Bits[8:6]
for Time Slot B. Alternatively, the final output is also available
through the data registers; Register 0x64, Register 0x70, and
Register 0x74 for Time Slot A, and Register 0x68, Register 0x78,
and Register 0x7C for Time Slot B.
To put the ADPD103 into digital integration mode during Time
Slot A, write 0x1 to Register 0x58, Bit 12. To put the ADPD103
into digital integration mode in Time Slot B, write 0x1 to
Register 0x58, Bit 13. The other writes required to switch to
digital integration mode are listed in Table 18.
When using digital integrate mode, up to two photodiodes can
be connected to the ADPD103 inputs; one photodiode per PDx
input group (PD1/PD2/PD3/PD4 or PD5/PD6/PD7/PD8).
Never connect the same photodiode across the two PDx groups.
In digital integrate mode, there are options to connect the
photodiode to all four AFE channels (PD1/PD2/PD3/PD4 or
PD5/PD6/PD7/PD8), or just a single AFE channel (PD1 or
PD5). When connecting to a single AFE channel, write 0x1 to
Register 0x54, Bit 14 for Time Slot A, or, for Time Slot B, write
0x1 to Register 0x54, Bit 15.
When connecting to a single AFE channel, there is also an
option to turn off Channel 2, Channel 3, and Channel 4 (and to
save power) by writing 0x7 to Register 0x55, Bits[15:13]. When
connecting to all four channels (PD1/PD2/PD3/PD4 or
PD5/PD6/PD7/PD8), write 0x0 (default)to Register 0x54, Bit 14
for Time Slot A, or write 0x0 (default) to Register 0x54, Bit 15
for Time Slot B. Ensure that all AFE channels are powered up by
writing 0x0 to Register 0x55, Bits[15:13].
Connecting the single photodiode to a single AFE channel offers
the best SNR performance in cases where signal is limited, whereas
connecting the single photodiode to all four AFE channels offers
the best dynamic range in cases where signal is large.
Digital Integration Sampling Modes
There are two sampling modes that can be used while the
device is in digital integration mode. These modes are singlesample pair mode and double-sample pair mode.
In single-sample pair mode, there is a single negative sample
region and a single positive sample region, shown in Figure 29
and Figure 30. To use single-sample pair mode, write 0x1 to
Register 5A, Bit 5 for Time Slot A, or Register 5A, Bit 6 for
Time Slot B. The negative sample region starts at SLOTx_AFE_
OFFSET + 9 and its duration (the number of samples taken) is set
by SLOTx_AFE_WIDTH. The positive sample region starts at
SLOTx_AFE_OFFSET + 9 + SLOTx_AFE_WIDTH, and its
duration is also set by SLOTx_AFE_WIDTH. Set the timing
such that the negative sample region falls entirely in the flat
(dark) portion of the LED response, whereas the positive
sample region falls in the pulsed region of the LED response.
Placing the LED pulse offset, SLOTx_LED_OFFSET, at the
beginning of SLOTx_AFE_OFFSET + 9 + SLOTx_AFE_WIDTH
achieves this timing. The output is the difference of the signals
in the two regions.
Rev. B | Page 30 of 52
Data Sheet
ADPD103
Double-sample pair mode is another way to sample. In this
mode, there are two negative sample regions and one long
positive sample region (see Figure 27 and Figure 28). To use
double-sample pair mode, write 0x0 to Register 0x5A, Bit 5 for
Time Slot A, or Bit 6 for Time Slot B. The first negative sample
region starts at SLOTx_AFE_OFFSET + 9 and its duration is set
by SLOTx_AFE_WIDTH. The positive sample region starts at
SLOTx_AFE_OFFSET + 9+ SLOTx_AFE_WIDTH and its
duration is twice the SLOTx_AFE_WIDTH. After this, there is
another negative sample region that starts at SLOTx_AFE_OFFSET
+ 9+ 3 × SLOTx_AFE_WIDTH, and its duration is SLOTx_AFE_
WIDTH. Set the timing such that both of the negative sample
regions fall in the flat (dark) portion of the LED response and
the positive sample region falls in the pulsed portion of the LED
response. Placing the LED pulse offset, SLOTx_LED_OFFSET
at the beginning of SLOTx_AFE_OFFSET + 9 + SLOTx_AFE_
WIDTH achieves this timing. The output is calculated by
summing the response of all the regions in a negative/positive/
negative manner. The double-sample pair mode is useful for cases
when the background light is not constant because it has better
background rejection, but it also uses more power than singlesample pair mode.
Sample Timing Modes
There are two options for timing the sample regions: gapped
mode and continuous mode.
In gapped timing mode, there is a space between the negative and
positive sample regions. The width of this region is specified by
SLOTA_AFE_FOFFSET for Time Slot A and SLOTB_AFE_
FOFFSET for Time Slot B in microseconds. To enable this
feature, write 0x1 to Register 0x5A, Bit 7. This bit enables
gapped timing for the time slot (or time slots) that are in digital
integrate mode. This mode is helpful when there are unwanted
transients in the LED response that must be ignored for an
accurate output.
If there are no concerns about LED response transients, select
continuous timing mode. In this mode, there is no space
between the negative and positive sample regions. Write 0x0 to
Register 0x5A, Bit 7 for continuous timing of the sample regions.
Both gapped and continuous sample timing modes can be used
with single-sample pair or double-sample pair mode. Some
example timing diagrams are shown in Figure 27, Figure 28,
Figure 29, and Figure 30.
Background Values
In digital integrate mode, the digital integration background
value, DI_BACKGROUND, or dark values are also stored and
available as output data. This is in addition to the output value
during the LED pulse, DI_OUTPUT, which has the dark value
subtracted. DI_BACKGROUND is the sum of the negative
region samples.
To include these values in the FIFO, set Register 0x11, Bits[4:2]
for Time Slot A, and Register 0x11, Bits[8:6] for Time Slot B. For
16-bit data, set this value to 0x3; for 32-bit data, set this value to
0x04. These settings are also available through the data registers;
Register 0x65, Register 0x71, and Register 0x75 for Time Slot A,
and Register 0x69, Register 0x79, and Register 0x7D for Time
Slot B. It is recommended that the channel offsets (Register 0x18 to
Register 0x21) be set to 0x1F00 when including the background
values in the FIFO in digital integration mode. These channel
offsets do not affect the sample values, but do provide more
headroom for the background values.
Saturation Detection in Digital Integrate Mode
In normal operation, when using the band-pass filter and the
integrator, the ADC almost always saturates before the TIA.
Unlike in normal operation, saturation of the TIA or the ADC
cannot be detected solely by looking at the signal value where
the signal value is the positive sample region minus the
reference region in digital integrate mode. This is because the
integrated value does not by itself contain any information
indicating if one of the ADC conversions during the integration
period exceeded the ADC output range. As a result, the realtime output may have saturated only for a fraction of the ADC
conversions within a sample and the final accumulated sum may
not reflect this. To detect TIA saturation in digital integration
mode, both the background values, DI_BACKGROUND, and
the signal values, DI_OUTPUT, must be collected. Refer to the
Background Values section for the correct settings for
Register 0x11 that provide these values.
For single-sample pair mode, saturation has occurred when
(DI_OUTPUT/(min(LED_WIDTH, AFE_WIDTH)) +
DI_BACKGROUND/AFE_WIDTH)/NUM_PULSES >
0x3FFF
For double-sample pair mode, saturation has occurred when
Rev. B | Page 31 of 52
(DI_OUTPUT/(min(LED_WIDTH, 2 × AFE_WIDTH)) +
DI_BACKGROUND/(2 × AFE_WIDTH))/NUM_PULSES
> 0x3FFF
ADPD103
Data Sheet
LED
SLOTx_AFE_OFFSET + 9
SUB
ADD
SUB
12722-021
SAMPLE
Figure 27. Digital Integration Mode in Double-Sample Pair Mode with Continuous Sample Timing
LED
SAMPLE
SUB
ADD
AFE_FOFFSET
SUB
AFE_FOFFSET
12722-022
SLOTx_AFE_OFFSET + 9
Figure 28. Digital Integration Mode in Double-Sample Pair Mode with Gapped Sample Timing
LED
SAMPLE
SUB
ADD
AFE_FOFFSET
12722-023
SLOTx_AFE_OFFSET + 9
Figure 29. Digital Integration Mode in Single-Sample Pair Mode with Gapped Sample Timing
LED
SLOTx_AFE_OFFSET + 9
SUB
ADD
Figure 30. Digital Integration Mode in Single-Sample Pair Mode with Continuous Sample Timing
Rev. B | Page 32 of 52
12722-024
SAMPLE
Data Sheet
ADPD103
Table 18. Configuration Registers to Switch Between the Normal Sample Mode, TIA_ADC Mode, and Digital Integration Mode
Bit Name
SLOTA_AFE_MODE
Normal
Mode
Value
0x1C
TIA_ADC
Mode Value
Not applicable
Digital
Integration
Mode Value
0x1D
[15:0]
SLOTA_AFE_CFG
0xADA5
0xB065
0xAE65
0x44
[15:8]
SLOTB_AFE_MODE
0x1C
Not applicable
0x1D
0x45
[15:0]
SLOTB_AFE_CFG
0xADA5
0xB065
0xAE65
0x4E
[15:0]
ADC_TIMING
Not applicable
0x0040
0x58
13
SLOTB_DIGITAL_INT_EN
Not
applicable
0x0
0x0
0x1
12
SLOTA_DIGITAL_INT_EN
0x0
0x0
0x1
[15:0]
DIG_INT_CFG
Not
applicable
Not applicable
Variable
Address
0x42
Data
Bits
[15:8]
0x43
0x5A
Rev. B | Page 33 of 52
Description
In normal mode, this setting configures the
integrator block for optimal operation. In
digital integration mode, this setting
configures the integrator block as a buffer.
This setting is not important for TIA_ADC
mode.
Time Slot A AFE connection.
0xAE65 bypasses the band-pass filter.
0xB065 bypasses the band-pass filter and
the integrator.
In normal mode, this setting configures the
integrator block for optimal operation. In
digital integration mode, this setting
configures the integrator block as a buffer.
This setting is not important for TIA_ADC
mode.
Time Slot B AFE connection.
0xAE65 bypasses the band-pass filter.
0xB065 bypasses the BPF and the integrator.
Set ADC Clock to 1 MHz in TIA_ADC mode.
Digital integrate mode enable Time Slot B.
0: disable.
1: enable.
Digital integrate mode enable Time Slot A.
0: disable.
1: enable.
Configuration of digital integration depends
on the use case. This register is ignored for
other modes.
ADPD103
Data Sheet
REGISTER LISTING
Table 19. Numeric Register Listing 1
Hex
Addr
Name
Bits
0x00
Status
[15:8]
[7:0]
0x01
0x02
0x06
0x08
0x09
Bit 15
Bit 14
Bit 13
Bit 12
Bit 11
Bit 10
Bit 9
Bit 8
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reserved
SLOTB_
INT
SLOTA_INT
[15:8]
INT_IO_
CTL
[15:8]
FIFO_
THRESH
[15:8]
[7:0]
Reserved[7:0]
DEVID
[15:8]
REV_NUM[15:8]
[7:0]
DEV_ID[7:0]
I2CS_ID
Reserved[15:9]
Reserved
SLOTB_INT_ SLOTA_INT_
MASK
MASK
0x0D
0x0F
0x10
CLK_
RATIO
Reserved[7:3]
INT_ENA
Reserved[15:14]
0x12
FSAMPLE [15:8]
[7:0]
0x15
0x18
0x19
0x1A
0x1B
0x1E
0x1F
0x20
0x21
0x0416
R
0x00C8
R/W
0x0000
R
0x0000
R/W
0x0000
R/W
0x0000
R/W
0x1000
R/W
0x0028
R/W
0x0541
R/W
0x0600
R/W
0x2000
R/W
0x2000
R/W
0x2000
R/W
0x2000
R/W
0x2000
R/W
0x2000
R/W
0x2000
R/W
0x2000
RW
SLAVE_ADDRESS_KEY[15:8]
SLAVE_ADDRESS_KEY[7:0]
Reserved[15:8]
Reserved[7:1]
SW_RESET
Reserved[15:8]
Reserved[7:2]
Reserved[15:14]
RDOUT_M
ODE
SLOTB_FIFO_MODE[7:6]
SLOTB_EN
Mode[1:0]
FIFO_OVRN_
PREVENT
Reserved[11:9]
SLOTA_FIFO_MODE[4:2]
SLOTB_FIFO_
MODE
Reserved
SLOTA_EN
FSAMPLE[15:8]
FSAMPLE[7:0]
PD_LED_ [15:8]
SELECT [7:0]
Reserved[15:12]
SLOTA_PD_SEL[7:4]
SLOTB_PD_SEL[11:8]
SLOTB_LED_SEL[3:2]
NUM_
AVG
[15:8]
SLOTA_
CH1_
OFFSET
[15:8]
SLOTA_CH1_OFFSET[15:8]
[7:0]
SLOTA_CH1_OFFSET[7:0]
SLOTA_
CH2_
OFFSET
[15:8]
SLOTA_CH2_OFFSET[15:8]
[7:0]
SLOTA_CH2_OFFSET[7:0]
SLOTA_
CH3_
OFFSET
[15:8]
SLOTA_CH3_OFFSET[15:8]
[7:0]
SLOTA_CH3_OFFSET[7:0]
SLOTA_
CH4_
OFFSET
[15:8]
SLOTA_CH4_OFFSET[15:8]
[7:0]
SLOTA_CH4_OFFSET[7:0]
SLOTB_
CH1_
OFFSET
[15:8]
SLOTB_CH1_OFFSET[15:8]
[7:0]
SLOTB_CH1_OFFSET[7:0]
SLOTB_
CH2_
OFFSET
[15:8]
SLOTB_CH2_OFFSET[15:8]
[7:0]
SLOTB_CH2_OFFSET[7:0]
SLOTB_
CH3_
OFFSET
[15:8]
SLOTB_CH3_OFFSET[15:8]
[7:0]
SLOTB_CH3_OFFSET[7:0]
SLOTB_
CH4_
OFFSET
[15:8]
SLOTB_CH4_OFFSET[15:8]
[7:0]
SLOTB_CH4_OFFSET[7:0]
[7:0]
R/W
CLK_RATIO[11:8]
[7:0]
0x14
0x0000
CLK_RATIO[7:0]
[7:0]
SLOT_EN [15:8]
R/W
Reserved
Reserved[15:12]
[7:0]
0x11
0x0000
INT_POL
ADDRESS_WRITE_KEY[15:8]
SLAVE_ [15:8]
ADDRESS [7:0]
_KEY
[15:8]
INT_DRV
FIFO_THRESH[13:8]
[7:0]
Mode
R/W
SLAVE_ADDRESS[7:1]
[15:8]
[15:8]
FIFO_INT_MASK 0x00FF
Reserved[4:0]
[15:8]
SW_
RESET
R/W
Reserved[15:8]
[7:0]
[7:0]
0x0A
RW
0x0000
Reserved[4:0]
INT_
MASK
[7:0]
Reset
FIFO_SAMPLES[15:8]
Reserved
Reserved
SLOTA_LED_SEL[1:0]
SLOTB_NUM_AVG
SLOTA_NUM_AVG
Reserved
Rev. B | Page 34 of 52
Data Sheet
Hex
ADPD103
Bit 15
Bit 14
Bit 13
Bit 12
Bit 11
Bit 10
Bit 9
Bit 8
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Addr
Name
Bits
0x22
ILED3_
COARSE
[15:8]
0x23
ILED1_
COARSE
0x24
ILED2_
COARSE
[15:8]
0x25
ILED_
FINE
[15:8]
SLOTA_
LED_
PULSE
[15:8]
[7:0]
SLOTA_LED_OFFSET[7:0]
SLOTA_
NUMPULSES
[15:8]
SLOTA_LED_NUMBER[15:8]
[7:0]
SLOTA_LED_PERIOD[7:0]
[7:0]
[15:8]
[7:0]
[7:0]
0x30
0x31
[7:0]
Reserved[15:14]
ILED3_
SCALE
Reserved
Reserved[12:8]
ILED3_SLEW[6:4]
Reserved[15:14]
Reserved[12:8]
ILED1_SLEW[6:4]
Reserved[15:14]
Reserved[12:8]
ILED2_SLEW[6:4]
ILED2_FINE[10:8]
Reserved
Reserved[15:13]
SLOTA_LED_WIDTH[12:8]
LED_
[15:8]
DISABLE
0x35
SLOTB_
LED_
PULSE
Reserved[15:10]
[15:8]
[7:0]
SLOTB_LED_OFFSET[7:0]
SLOTB_
NUMPULSES
[15:8]
SLOTB_LED_NUMBER[15:8]
[7:0]
SLOTB_LED_PERIOD[7:0]
[7:0]
SLOTA_ [15:8]
AFE_
[7:0]
WINDOW
Reserved
EXT_SYNC_
ENA
Reserved[13:8]
SLOTA_AFE_OFFSET[10:5]
SLOTB_ [15:8]
AFE_
[7:0]
WINDOW
SLOTB_AFE_OFFSET[10:5]
[7:0]
0x42
SLOTA_ [15:8]
TIA_CFG [7:0]
0x44
0x45
0x4B
0x4D
0x4E
Reserved[15:14]
0x4F
SLOTA_AFE_OFFSET[10:8]
SLOTB_AFE_OFFSET[10:8]
Reserved
AFE_POWERDOWN[7:3]
V_
CATHODE
AFE_
POWER-DOWN
0x0320
R/W
0x0818
R/W
0x0000
R/W
0x0320
R/W
0x0818
R/W
0x000
R/W
0x22FC
R/W
0x22FC
R/W
0x3006
R/W
0x1C38
R/W
0xADA5
R/W
0x1C38
R/W
0xADA5
R/W
0x2612
R/W
0x425E
R/W
0x0060
R/W
0x2090
R/W
Reserved[2:0]
SLOTA_AFE_MODE[15:8]
Reserved
Reserved[5:2] (write 0xD)
SLOTA_
TIA_
IND_EN
SLOTA_TIA_GAIN[1:0]
SLOTA_AFE_CFG[15:8]
SLOTA_AFE_CFG[7:0]
SLOTB_AFE_MODE[15:8]
Reserved
Reserved[5:2] (write 0xD)
SLOTB_
TIA_
IND_EN
SLOTB_TIA_GAIN[1:0]
SLOTB_AFE_CFG[15:8]
SLOTB_AFE_CFG[7:0]
SAMPLE_ [15:8]
CLK
[7:0] CLK32K_EN
Reserved[15:8]
Reserved
CLK32K_ADJUST[5:0]
CLK32M_ [15:8]
ADJUST [7:0]
RESERVED[15:8]
CLK32M_ADJUST[7:0]
[15:8]
ADC_CLOCK[15:8]
[7:0]
ADC_CLOCK[7:0]
EXT_
[15:8]
SYNC_SEL [7:0]
R/W
SLOTB_AFE_FOFFSET[4:0]
Reserved[13:11]
SLOTB_ [15:8]
AFE_CFG [7:0]
ADC_
CLOCK
0x630C
SLOTA_AFE_FOFFSET[4:0]
SLOTB_AFE_WIDTH[15:11]
SLOTA_ [15:8]
AFE_CFG [7:0]
SLOTB_ [15:8]
TIA_CFG [7:0]
R/W
Reserved[7:0]
SLOTA_AFE_WIDTH[15:11]
[7:0]
0x43
SLOTA_
LED_DIS
SLOTB_LED_WIDTH[12:8]
0x39
AFE_PWR [15:8]
_CFG1
0x3000
Reserved[7:0]
TIMING_ [15:8]
CFG
0x3C
SLOTB_LED_
DIS
Reserved[15:13]
0x38
0x3B
R/W
ILED1_FINE[4:0]
0x34
0x36
0x3000
ILED2_COARSE[3:0]
ILED3_FINE[15:11]
ILED2_FINE[7:6]
R/W
ILED1_COARSE[3:0]
ILED2_
SCALE
Reserved
RW
0x3000
ILED3_COARSE[3:0]
ILED1_
SCALE
Reserved
Reset
Reserved[15:8]
Reserved
PDSO_
OE
PDSO_IE
Reserved
EXT_SYNC_SEL[3:2]
Rev. B | Page 35 of 52
INT_IE
Reserved
ADPD103
Data Sheet
Hex
Addr
Name
0x50
CLK32M_ [15:8]
CAL_EN [7:0]
0x54
Bits
AFE_PWR [15:8]
_CFG2
[7:0]
0x55
TIA_INDEP_
GAIN
[15:8]
[7:0]
Bit 15
Bit 14
Bit 13
Bit 12
Bit 11
Bit 10
Bit 9
Bit 8
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Reserved
SLOTB_SINGLE_CH_
DIG_INT
PDSO_
CTRL
SLOTA_
SINGLE_
CH_DIG_
INT
DIG_INT_ [15:8]
CFG
[7:0] DIG_INT_
GAPMODE
Reserved[15:14]
0x65
0x66
0x67
0x68
0x69
0x6A
0x6B
0x70
0x71
0x72
0x73
0x74
Reserved
SLOTA_TIA_GAIN_4[5:4]
SLOTB_
DIGITAL_
INT_EN
[7:0]
0x64
SLOTA_V_CATHODE[9:8]
0x0020
R/W
SLOTB_TIA_GAIN_4 [11:10]
SLOTB_TIA_GAIN_3[9:8]
0x0000
R/W
SLOTA_TIA_GAIN_3[3:2]
SLOTA_TIA_GAIN_2[1:0]
0x0000
R/W
0x0000
R/W
0x0000
R/W
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
Reserved[4:0]
SLEEP_V_CATHODE[13:12]
DIGINT_POWER[15:13]
0x5A
0x60
R/W
SLOTB_V_
CATHODE[11:10]
Reserved[6:0]
SLOTB_TIA_GAIN_2[7:6]
DIGITAL_ [15:8]
INT_EN
DATA_AC [15:8]
CESS_
[7:0]
CTL
RW
0x0000
CLK32M_
CAL_EN
REG54_VCAT_
ENABLE
0x58
0x5F
Reset
Reserved[15:8]
Reserved[11:8]
SLOTA_
DIGITAL_INT_
EN
Reserved[7:0]
Reserved[15:8]
SLOTB_
DIG_INT_
SAMPLEMODE
Reserved[4:0]
SLOTA_
DIG_INT_
SAMPLEMODE
Reserved[15:8]
Reserved[7:3]
SLOTB_
DATA_
HOLD
FIFO_
ACCESS
[15:8]
FIFO_DATA[15:8]
[7:0]
FIFO_DATA[7:0]
SLOTA_
PD1_
16BIT
[15:8]
SLOTA_CH1_16BIT[15:8]
[7:0]
SLOTA_CH1_16BIT[7:0]
SLOTA_
PD2_
16BIT
[15:8]
SLOTA_CH2_16BIT[15:8]
[7:0]
SLOTA_CH2_16BIT[7:0]
SLOTA_
PD3_
16BIT
[15:8]
SLOTA_CH3_16BIT[15:8]
[7:0]
SLOTA_CH3_16BIT[7:0]
SLOTA_
PD4_
16BIT
[15:8]
SLOTA_CH4_16BIT[15:8]
[7:0]
SLOTA_CH4_16BIT[7:0]
SLOTB_
PD1_
16BIT
[15:8]
SLOTB_CH1_16BIT[15:8]
[7:0]
SLOTB_CH1_16BIT[7:0]
SLOTB_
PD2_
16BIT
[15:8]
SLOTB_CH2_16BIT[15:8]
[7:0]
SLOTB_CH2_16BIT[7:0]
SLOTB_
PD3_
16BIT
[15:8]
SLOTB_CH3_16BIT[15:8]
[7:0]
SLOTB_CH3_16BIT[7:0]
SLOTB_
PD4_
16BIT
[15:8]
SLOTB_CH4_16BIT[15:8]
[7:0]
SLOTB_CH4_16BIT[7:0]
A_PD1_
LOW
[15:8]
SLOTA_CH1_LOW[15:8]
[7:0]
SLOTA_CH1_LOW[7:0]
A_PD2_
LOW
[15:8]
SLOTA_CH2_LOW[15:8]
[7:0]
SLOTA_CH2_LOW[7:0]
A_PD3_
LOW
[15:8]
SLOTA_CH3_LOW[15:8]
[7:0]
SLOTA_CH3_LOW[7:0]
A_PD4_
LOW
[15:8]
SLOTA_CH4_LOW[15:8]
[7:0]
SLOTA_CH4_LOW[7:0]
A_PD1_
HIGH
[15:8]
SLOTA_CH1_HIGH[15:8]
[7:0]
SLOTA_CH1_HIGH[7:0]
Rev. B | Page 36 of 52
SLOTA_DAT
A_HOLD
DIGITAL_
CLOCK_ENA
Data Sheet
Hex
ADPD103
Bit 15
Bit 14
Bit 13
Bit 12
Bit 11
Bit 10
Bit 9
Bit 8
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Bit 2
Bit 1
Bit 0
Addr
Name
Bits
0x75
A_PD2_
HIGH
[15:8]
SLOTA_CH2_HIGH[15:8]
[7:0]
SLOTA_CH2_HIGH[7:0]
A_PD3_
HIGH
[15:8]
SLOTA_CH3_HIGH[15:8]
[7:0]
SLOTA_CH3_HIGH[7:0]
A_PD4_
HIGH
[15:8]
SLOTA_CH4_HIGH[15:8]
[7:0]
SLOTA_CH4_HIGH[7:0]
B_PD1_
LOW
[15:8]
SLOTB_CH1_LOW[15:8]
[7:0]
SLOTB_CH1_LOW[7:0]
B_PD2_
LOW
[15:8]
SLOTB_CH2_LOW[15:8]
[7:0]
SLOTB_CH2_LOW[7:0]
B_PD3_
LOW
[15:8]
SLOTB_CH3_LOW[15:8]
[7:0]
SLOTB_CH3_LOW[7:0]
B_PD4_
LOW
[15:8]
SLOTB_CH4_LOW[15:8]
[7:0]
SLOTB_CH4_LOW[7:0]
B_PD1_
HIGH
[15:8]
SLOTB_CH1_HIGH[15:8]
[7:0]
SLOTB_CH1_HIGH[7:0]
B_PD2_
HIGH
[15:8]
SLOTB_CH2_HIGH[15:8]
[7:0]
SLOTB_CH2_HIGH[7:0]
B_PD3_
HIGH
[15:8]
SLOTB_CH3_HIGH[15:8]
[7:0]
SLOTB_CH3_HIGH[7:0]
B_PD4_
HIGH
[15:8]
SLOTB_CH4_HIGH[15:8]
[7:0]
SLOTB_CH4_HIGH[7:0]
0x76
0x77
0x78
0x79
0x7A
0x7B
0x7C
0x7D
0x7E
0x7F
1
Reset
RW
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
0x0000
R
Recommended values not shown. Only power-on reset values are in Table 19. The recommended values are largely dependent on use case. See Table 20 to Table 26
for the recommended values.
Rev. B | Page 37 of 52
ADPD103
Data Sheet
LED CONTROL REGISTERS
Table 20. LED Control Registers
Address
0x14
0x22
Data Bit
[15:12]
[11:8]
Default
Value
0x0
0x5
Access
R/W
R/W
Name
Reserved
SLOTB_PD_SEL
[7:4]
0x4
R/W
SLOTA_PD_SEL
[3:2]
0x0
R/W
SLOTB_LED_SEL
[1:0]
0x1
R/W
SLOTA_LED_SEL
[15:14]
13
0x0
0x1
R/W
R/W
Reserved
ILED3_SCALE
12
[11:7]
[6:4]
0x1
0x0
0x0
R/W
R/W
R/W
Reserved
Reserved
ILED3_SLEW
[3:0]
0x0
R/W
ILED3_COARSE
Description
Write 0x0 to these bits for proper operation.
PDx connection selection for Time Slot B. See Figure 13.
0x1: All photodiode inputs are connected during Time Slot B.
0x4: PD5/PD6/PD7/PD8 are connected during Time Slot B.
0x5: PD1/PD2/PD3/PD4 are connected during Time Slot B.
Other: reserved.
PDx connection selection for Time Slot A. See Figure 13.
0x1: All photodiode inputs are connected during Time Slot A.
0x4: PD5/PD6/PD7/PD8 are connected during Time Slot A.
0x5: PD1/PD2/PD3/PD4 are connected during Time Slot A.
Other: reserved.
Time Slot B LED configuration. These bits determine which LED is
associated with Time Slot B.
0x0: reserved.
0x1: LEDX1 pulses during Time Slot B.
0x2: LEDX2 pulses during Time Slot B.
0x3: LEDX3 pulses during Time Slot B.
Time Slot A LED configuration. These bits determine which LED is
associated with Time Slot A.
0x0: reserved.
0x1: LEDX1 pulses during Time Slot A.
0x2: LEDX2 pulses during Time Slot A.
0x3: LEDX3 pulses during Time Slot A.
Write 0x0.
LEDX3 current scale factor.
1: 100% strength.
0: 40% strength; sets the LEDX3 driver in low power mode.
LEDX3 Current Scale = 0.4 + 0.6 × (Register 0x22, Bit 13).
Write 0x1.
Write 0x0.
LEDX3 driver slew rate control. The slower the slew rate, the safer
the performance in terms of reducing the risk of overvoltage of the
LED driver.
0: the slowest slew rate.
…
7: the fastest slew rate.
LEDX3 coarse current setting. Coarse current sink target value of
LEDX3 in standard operation.
0: 25 mA.
1: 40 mA.
2: 55 mA.
…
15: 250 mA.
LED3PEAK = LED3COARSE × LED3FINE × LED3SCALE
where:
LED3PEAK is the LEDX3 peak target value (mA).
LED3COARSE = 28 + 15.46 × (Register 0x22, Bits[3:0]).
LED3FINE = 0.71 + 0.024 × (Register 0x25, Bits[15:11]).
LED3SCALE = 0.4 + 0.6 × (Register 0x22, Bit 13).
Rev. B | Page 38 of 52
Data Sheet
Address
0x23
0x24
ADPD103
Data Bit
[15:14]
13
Default
Value
0x0
0x1
Access
R/W
R/W
Name
Reserved
ILED1_SCALE
12
[11:7]
[6:4]
0x1
0x0
0x0
R/W
R/W
R/W
Reserved
Reserved
ILED1_SLEW
[3:0]
0x0
R/W
ILED1_COARSE
[15:14]
13
0x0
0x1
R/W
R/W
Reserved
ILED2_SCALE
12
[11:7]
[6:4]
0x1
0x0
0x0
R/W
R/W
R/W
Reserved
Reserved
ILED2_SLEW
[3:0]
0x0
R/W
ILED2_COARSE
Description
Write 0x0.
LEDX1 current scale factor.
1: 100% strength.
0: 40% strength; sets the LEDX1 driver in low power mode.
LEDX1 Current Scale = 0.4 + 0.6 × (Register 0x23, Bit 13).
Write 0x1.
Write 0x0.
LEDX1 driver slew rate control. The slower the slew rate, the safer
the performance in terms of reducing the risk of overvoltage of the
LED driver.
0: the slowest slew rate.
…
7: the fastest slew rate.
LEDX1 coarse current setting. Coarse current sink target value of
LEDX1 in standard operation.
0: 25 mA.
1: 40 mA.
2: 55 mA.
…
15: 250 mA.
LED1PEAK = LED1COARSE × LED1FINE × LED1SCALE
where:
LED1PEAK is the LEDX1 peak target value (mA).
LED1COARSE = 28 + 15.46 × (Register 0x23, Bits[3:0]).
LED1FINE = 0.71 + 0.024 × (Register 0x25, Bits[4:0]).
LED1SCALE = 0.4 + 0.6 × (Register 0x23, Bit 13).
Write 0x0.
LEDX2 current scale factor.
1: 100% strength.
0: 40% strength; sets the LEDX2 driver in low power mode.
LED2 Current Scale = 0.4 + 0.6 × (Register 0x24, Bit 13)
Write 0x1.
Write 0x0.
LEDX2 driver slew rate control. The slower the slew rate, the safer
the performance in terms of reducing the risk of overvoltage of the
LED driver.
0: the slowest slew rate.
…
7: the fastest slew rate.
LEDX2 coarse current setting. Coarse current sink target value of
LED2 in standard operation. See Register 0x23, Bits[3:0] for values.
LED2PEAK = LED2COARSE × LED2FINE × LED2SCALE
where:
LED2PEAK is the LEDX2 peak target value (mA).
LED2COARSE = 28 + 15.46 × (Register 0x24, Bits[3:0]).
LED2FINE = 0.71 + 0.024 × (Register 0x25, Bits[10:6]).
LED2SCALE = 0.4 + 0.6 × (Register 0x24, Bit 13).
Rev. B | Page 39 of 52
ADPD103
Address
0x25
0x30
0x31
0x34
0x35
0x36
0x3C
Data Sheet
Data Bit
[15:11]
Default
Value
0xC
Access
R/W
Name
ILED3_FINE
[10:6]
0xC
R/W
ILED2_FINE
5
[4:0]
0x0
0xC
R/W
R/W
Reserved
ILED1_FINE
[15:13]
[12:8]
[7:0]
[15:8]
0x0
0x3
0x20
0x08
R/W
R/W
R/W
R/W
Reserved
SLOTA_LED_WIDTH
SLOTA_LED_OFFSET
SLOTA_LED_NUMBER
[7:0]
[15:10]
9
0x18
0x00
0x0
R/W
R/W
R/W
SLOTA_LED_PERIOD
Reserved
SLOTB_LED_DIS
8
0x0
R/W
SLOTA_LED_DIS
[7:0]
[15:13]
[12:8]
[7:0]
[15:8]
0x00
0x0
0x3
0x20
0x08
R/W
R/W
R/W
Reserved
Reserved
SLOTB_LED_WIDTH
SLOTB_LED_OFFSET
SLOTB_LED_NUMBER
[7:0]
[15:14]
[13:11]
10
9
0x18
0x0
0x6
0x0
0x0
R/W
R/W
R/W
R/W
R/W
SLOTB_LED_PERIOD
RESERVED
RESERVED
Reserved
V_CATHODE
[8:3]
0x0
R/W
AFE_POWERDOWN
[2:0]
0x6
R/W
Reserved
Description
LEDX3 fine adjust. Current adjust multiplier for LED3.
LEDX3 fine adjust = 0.71 + 0.024 × (Register 0x25, Bits[15:11]).
See Register 0x22, Bits[3:0], for the full LED3 formula.
LEDX2 fine adjust. Current adjust multiplier for LED2.
LEDX2 fine adjust = 0.71 + 0.024 × (Register 0x25, Bits[10:6]).
See Register 0x24, Bits[3:0], for the full LED2 formula.
Write 0x0.
LEDX1 fine adjust. Current adjust multiplier for LED1.
LEDX1 fine adjust = 0.71 + 0.024 × (Register 0x25, Bits[4:0]).
See Register 0x23, Bits[3:0], for the full LED1 formula.
Write 0x0.
LED pulse width (in 1 μs step) for Time Slot A.
LED offset width (in 1 μs step) for Time Slot A.
LED Time Slot A pulse count. nA: number of LED pulses in Time
Slot A. This is typically LED1. Adjust in the application. A setting of
six pulses (0x06) is typical.
LED Time Slot A pulse period (in 1 μs step).
Write 0x0.
Time Slot B LED disable. 1: disables the LED assigned to Time Slot B.
Register 0x34 keeps the drivers active and prevents them from
pulsing current to the LEDs. Disabling both LEDs via this register is
often used to measure the dark level.
Use Register 0x11 instead to enable or disable the actual time slot
usage and not only the LED.
Time Slot A LED disable. 1: disables the LED assigned to Time Slot A.
Use Register 0x11 instead to enable or disable the actual time slot
usage and not only the LED.
Write 0x00.
Write 0x0.
LED pulse width (in 1 μs step) for Time Slot B.
LED offset width (in 1 μs step) for Time Slot B.
LED Time Slot B pulse count. nB: number of LED pulses in Time Slot B.
This is typically LED2. A setting of six pulses (0x06) is typical.
LED Time Slot B pulse period (in 1 μs step).
Write 0x0.
Write 0x6.
Write 0x0.
0x0: 1.3 V (identical to anode voltage); recommended setting.
0x1: 1.8 V (reverse bias photodiode by 550 mV; this setting may
add noise).
AFE channels power-down select.
0x38: powers down AFE Channel 2, Channel 3, and Channel 4.
0x0: keeps all channels on.
Write 0x6.
Rev. B | Page 40 of 52
Data Sheet
ADPD103
AFE CONFIGURATION REGISTERS
Table 21. AFE Global Configuration Registers
Address
0x3C
0x54
Data Bit
[15:14]
[13:11]
Default
Value
0x0
0x6
Access
R/W
R/W
Name
RESERVED
RESERVED
Description
Write 0x0.
Write 0x6.
10
0x0
R/W
Reserved
Write 0x0.
9
0x0
R/W
V_CATHODE
[8:3]
0x0
R/W
AFE_POWERDOWN
0x0: 1.3 V (identical to anode voltage); recommended
setting.
0x1: 1.8 V (reverse bias photodiode by 550 mV. This
setting may add noise).
AFE channels power-down select.
0x38: powers down AFE Channel 2, Channel 3, and
Channel 4.
0x0: keeps all channels on.
[2:0]
0x6
R/W
Reserved
Write 0x6.
15
0x0
R/W
SLOTB_SINGLE_CH_DIG_INT
14
0x0
R/W
SLOTA_SINGLE_CH_DIG_INT
[13:12]
0x0
R/W
SLEEP_V_CATHODE
0: In Time Slot B, use all four photodiode channels in
parallel for digital integration (default setting for highest
dynamic range)
1: In Time Slot B, use only Channel 1 for digital
integration. This limits connection to PD1 or PD5.
0: In Time Slot A, use all four photodiode channels in
parallel for digital integration (default setting for highest
dynamic range)
1: In Time Slot A, use only Channel 1 for digital
integration. This limits connection to PD1 or PD5.
If Bit 7 = 1; this setting is applied to the cathode voltage
while the device is in sleep mode. The anode voltage is
always set to the cathode voltage in sleep mode.
0x0: VDD (1.8 V).
0x1: 1.3 V.
0x2: 1.55 V.
0x3: 0.0 V.
[11:10]
0x0
R/W
SLOTB_V_CATHODE
If Bit 7 = 1; this setting is applied to the cathode voltage
while the device is in Time Slot B operation. The anode
voltage is always 1.3 V in Time Slot B mode.
0x0: VDD (1.8 V).
0x1: 1.3 V.
0x2: 1.55 V.
0x3: 0.0 V (this forward biases a diode at the input).
[9:8]
0x0
R/W
SLOTA_V_CATHODE
If Bit 7 = 1; this setting is applied to the cathode voltage
while the device is in Time Slot A operation. The anode
voltage is always 1.3 V in Time Slot A mode.
0x0: VDD (1.8 V).
0x1: 1.3 V.
0x2: 1.55 V.
0x3: 0.0 V (this forward biases a diode at the input).
7
0x0
R/W
REG54_VCAT_ENABLE
[6:0]
0x20
R/W
Reserved
0: use the cathode voltage settings defined by Register 0x3C,
Bit 9.
1: override Register 0x3C, Bit 9 with cathode settings
defined by Register 0x54, Bits[13:8].
Write 0x20.
Rev. B | Page 41 of 52
ADPD103
Address
0x58
0x5A
Data Sheet
Data Bit
[15:14]
13
Default
Value
0x0
0x0
Access
R/W
R/W
Name
Reserved
SLOTB_DIGITAL_INT_EN
12
0x0
R/W
SLOTA_DIGITAL_INT_EN
[11:0]
[15:8]
7
0x000
0x00
0x0
R/W
R/W
R/W
Reserved
Reserved
DIG_INT_GAPMODE
6
0x0
R/W
SLOTB_DIG_INT_SAMPLE_MODE
5
0x0
R/W
SLOTA_DIG_INT_SAMPLE_MODE
[4:0]
0x00
R/W
Reserved
Description
Write 0x0.
0x0: Time Slot B operating in normal mode.
0x1: Time Slot B operating in digital integration mode.
0x0: Time Slot A operating in normal mode.
0x1: Time Slot A operating in digital integration mode.
Write 0x000.
Write 0x00.
Digital integrate gapped mode enable.
0: no gap between negative and positive sample regions.
1: use SLOTA_AFE_FOFFSET for Time Slot A or SLOTB_AFE_
FOFFSET for Time Slot B to specify the gap in μs.
Digital integrate single sample pair mode for Time Slot B.
0: double sample pair mode.
1: single sampled pair mode.
Digital integrate single sample pair mode for Time Slot A.
0: double sample pair mode.
1: single sampled pair mode.
Write 0x00.
Table 22. AFE Configuration Registers, Time Slot A
Address
0x39
0x42
Data
Bit
[15:11]
[10:5]
Default
Value
0x4
0x17
Access
R/W
R/W
Name
SLOTA_AFE_WIDTH
SLOTA_AFE_OFFSET
[4:0]
0x1C
R/W
SLOTA_AFE_FOFFSET
[15:8]
0x1C
R/W
SLOTA_AFE_MODE
7
6
0x0
0x0
R/W
R/W
Reserved
SLOTA_TIA_IND_EN
[5:2]
[1:0]
0xE
0x0
R/W
R/W
Reserved
SLOTA_TIA_GAIN
Description
AFE integration window width (in 1 μs step) for Time Slot A.
AFE integration window coarse offset (in 1 μs step) for
Time Slot A.
AFE integration window fine offset (in 31.25 ns step) for
Time Slot A.
0x1C: Time Slot A AFE setting for normal mode. All four blocks
of the signal chain are in use during normal mode (the TIA,
the BPF, followed by the integrator (INT), and finally the ADC).
0x1D: Time Slot A AFE setting for digital integrate mode.
Write 0x0.
Enable Time Slot A TIA gain individual settings. When it is
enabled, the Channel 1 TIA gain is set via Register 0x42,
Bits[1:0], and the Channel 2 through Channel 4 TIA gain is set
via Register 0x55, Bits[5:0].
0: disable TIA gain individual setting.
1: enable TIA gain individual setting.
Reserved. Write 0xD.
Transimpedance amplifier gain for Time Slot A. When
SLOTA_TIA_IND_EN is enabled, this value is for Time Slot B,
Channel 1 TIA gain. When SLOTA_TIA_IND_EN is disabled, it is
for all four Time Slot A channel TIA gain settings.
0: 200 kΩ.
1: 100 kΩ.
2: 50 kΩ.
3: 25 kΩ.
Rev. B | Page 42 of 52
Data Sheet
ADPD103
Address
0x43
Data
Bit
[15:0]
Default
Value
0xADA5
Access
R/W
Name
SLOTA_AFE_CFG
0x55
[15:13]
0x0
R/W
DIGINT_POWER
12
[11:10]
0x0
0x0
R/W
R/W
Reserved
SLOTB_TIA_GAIN_4
[9:8]
0x0
R/W
SLOTB_TIA_GAIN_3
[7:6]
0x0
R/W
SLOTB_TIA_GAIN_2
[5:4]
0x0
R/W
SLOTA_TIA_GAIN_4
[3:2]
0x0
R/W
SLOTA_TIA_GAIN_3
[1:0]
0x0
R/W
SLOTA_TIA_GAIN_2
Description
AFE connection in Time Slot A.
0xADA5: analog full path mode (TIA_BPF_INT_ADC).
0xB065: TIA_ADC mode.
0xAE65: digital integration mode.
Others: reserved.
Power-down for Channel 2, Channel 3, and Channel 4 in
digital integration mode.
0: keep all channels powered up.
7: powers down Channel 2, Channel 3, and Channel 4.
Write 0x0.
TIA gain for Time Slot B, Channel 4 (PD4).
0: 200 kΩ
1: 100 kΩ.
2: 50 kΩ.
3: 25 kΩ.
TIA gain for Time Slot B, Channel 3 (PD3).
0: 200 kΩ
1: 100 kΩ.
2: 50 kΩ.
3: 25 kΩ.
TIA gain for Time Slot B, Channel 2 (PD2).
0: 200 kΩ
1: 100 kΩ.
2: 50 kΩ.
3: 25 kΩ.
TIA gain for Time Slot A, Channel 4 (PD4).
0: 200 kΩ
1: 100 kΩ.
2: 50 kΩ.
3: 25 kΩ.
TIA gain for Time Slot A, Channel 3 (PD3).
0: 200 kΩ
1: 100 kΩ.
2: 50 kΩ.
3: 25 kΩ.
TIA gain for Time Slot A, Channel 2 (PD2).
0: 200 kΩ
1: 100 kΩ.
2: 50 kΩ.
3: 25 kΩ.
Rev. B | Page 43 of 52
ADPD103
Address
0x5A
Data Sheet
Data
Bit
[15:8]
7
Default
Value
0x0
0x0
Access
R/W
R/W
Name
Reserved
DIG_INT_GAPMODE
6
0x0
R/W
SLOTB_DIG_INT_SAMPLEMODE
5
0x0
R/W
SLOTA_DIG_INT_SAMPLEMODE
[4:0]
0x0
R/W
Reserved
Description
Write 0x0.
Digital integration gapped mode enable.
0: no gap between negative and positive sample regions.
1: use SLOTA_AFE_FOFFSET for Time Slot A or SLOTB_AFE_
FOFFSET for Time Slot B to specify the gap in μs.
Digital integration single-sample pair mode for Time Slot B.
0: double sample pair mode.
1: single-sampled pair mode.
Digital integration single-sample pair mode for Time Slot A.
0: double sample pair mode.
1: single-sampled pair mode.
Write 0x0.
Table 23. AFE Configuration Registers, Time Slot B
Address
0x3B
0x44
0x45
Data
Bit
[15:11]
[10:5]
Default Value
0x4
0x17
Access
R/W
R/W
Name
SLOTB_AFE_WIDTH
SLOTB_AFE_OFFSET
[4:0]
0x1C
R/W
SLOTB_AFE_FOFFSET
[15:8]
0x1C
R/W
SLOTB_AFE_MODE
7
6
0x0
0x0
R/W
R/W
Reserved
SLOTB_TIA_IND_EN
[5:2]
[1:0]
0xE
0x0
R/W
R/W
Reserved
SLOTB_TIA_GAIN
[15:0]
0xADA5
R/W
SLOTB_AFE_CFG
Rev. B | Page 44 of 52
Description
AFE integration window width (in 1 μs step) for Time Slot B.
AFE integration window coarse offset (in 1 μs step) for
Time Slot B.
AFE integration window fine offset (in 31.25 ns step) for
Time Slot B.
0x1C: Time Slot B AFE setting for normal mode
(TIA_BPF_INT_ADC).
0x1D: Time Slot B AFE setting for digital integrate mode.
Write 0x0.
Enable Time Slot B TIA gain individual settings. When it is
enabled, the Channel 1 TIA gain is set via Register 0x44,
Bits[1:0], and the Channel 2 through Channel 4 TIA gain is
set via Register 0x55, Bits[11:6].
0: disable TIA gain individual setting.
1: enable TIA gain individual setting.
Write 0xD.
Transimpedance amplifier gain for Time Slot B. When
SLOTB_TIA_IND_EN is enabled, this value is for Time Slot B,
Channel 1 TIA gain. When SLOTB_TIA_IND_EN is disabled,
it is for all four Time Slot B channel TIA gain settings.
0: 200 kΩ.
1: 100 kΩ.
2: 50 kΩ.
3: 25 kΩ.
AFE connection in Time Slot B.
0xADA5: analog full path mode (TIA_BPF_INT_ADC).
0xB065: TIA_ADC mode.
0xAE65: digital integration mode.
Others: reserved.
Data Sheet
Address
0x58
ADPD103
Data
Bit
[15:14]
13
Default Value
0x0
0x0
Access
R/W
R/W
Name
Reserved
DIG_INT_EN_B
12
0x0
R/W
DIG_INT_EN_A
[11:0]
0x0000
R/W
Reserved
Rev. B | Page 45 of 52
Description
Write 0x0.
Digital integration mode, enable Time Slot B.
0: disable.
1: enable.
Digital integration mode, enable Time Slot A.
0: disable.
1: enable.
Write 0x0000.
ADPD103
Data Sheet
SYSTEM REGISTERS
Table 24. System Registers
Address
0x00
0x01
0x02
0x06
0x08
Data
Bit
[15:8]
Default
0x00
Access
R/W
Name
FIFO_SAMPLES
7
6
0x0
0x0
R/W
R/W
Reserved
SLOTB_INT
5
0x0
R/W
SLOTA_INT
[4:0]
[15:9]
8
0x00
0x00
0x0
R/W
R/W
R/W
Reserved
Reserved
FIFO_INT_MASK
7
6
0x1
0x1
R/W
R/W
Reserved
SLOTB_INT_MASK
5
0x1
R/W
SLOTA_INT_MASK
[4:0]
[15:3]
2
0x1F
0x0000
0x0
R/W
R/W
R/W
Reserved
Reserved
INT_ENA
1
0x0
R/W
INT_DRV
0
0x0
R/W
INT_POL
[15:14]
[13:8]
0x0
0x00
R/W
R/W
Reserved
FIFO_THRESH
[7:0]
[15:8]
[7:0]
0x00
0x04
0x16
R/W
R
R
Reserved
REV_NUM
DEV_ID
Description
FIFO status. Number of available bytes to be read from the FIFO. When
comparing this to the FIFO length threshold (Register 0x06, Bits[13:8]),
note that the FIFO status value is in bytes and the FIFO length threshold
is in words, where one word = two bytes.
With the FIFO_ACCESS_ENA bit set, write 1 to Bit 15 of FIFO_SAMPLES to
clear the contents of the FIFO.
Write 0x1 to clear this bit to 0x0.
Time Slot B interrupt. Describes the type of interrupt event. A 1 indicates
an interrupt of a particular event type has occurred. Write a 1 to clear the
corresponding interrupt. After clearing, the register goes to 0. Writing a 0
to this register has no effect.
Time Slot A interrupt. Describes the type of interrupt event. A 1 indicates an
interrupt of a particular event type has occurred. Write a 1 to clear the
corresponding interrupt. After clearing, the register goes to 0. Writing a 0
to this register has no effect
Write 0x1F to clear these bits to 0x00.
Write 0x00.
Sends an interrupt when the FIFO data length has exceeded the FIFO
length threshold in Register 0x06, Bits[13:8]. A 0 enables the interrupt.
Write 0x1.
Sends an interrupt on the Time Slot B sample. Write a 1 to disable the
interrupt. Write a 0 to enable the interrupt.
Sends an interrupt on the Time Slot A sample. Write a 1 to disable the
interrupt. Write a 0 to enable the interrupt.
Write 0x1F.
Write 0x0000.
INT pin enable.
0: disable the INT pin. The INT pin floats regardless of interrupt status.
The status register (Address 0x00) remains active.
1: enable the INT pin.
INT drive.
0: the INT pin is always driven.
1: the INT pin is driven when the interrupt is asserted; otherwise, it is left
floating and requires a pull-up or pull-down resistor, depending on
polarity (operates as open drain). Use this setting if multiple devices
need to share the INT pin.
INT polarity.
0: the INT pin is active high.
1: the INT pin is active low.
Write 0x0.
FIFO length threshold. An interrupt is generated when the number of
data-words in the FIFO exceeds the value in FIFO_THRESH. The interrupt pin
automatically deasserts when the number of data-words available in the
FIFO no longer exceeds the value in FIFO_THRESH.
Write 0x00.
Revision number.
Device ID.
Rev. B | Page 46 of 52
Data Sheet
ADPD103
Data
Bit
[15:8]
[7:1]
0
[15:12]
[11:0]
Default
0x0
0x64
0x0
0x0
0x000
Access
W
R/W
R
R
R
Name
ADDRESS_WRITE_KEY
SLAVE_ADDRESS
Reserved
Reserved
CLK_RATIO
0x0D
[15:0]
0x0
R/W
SLAVE_ADDRESS_KEY
0x0F
[15:1]
0
0x0000
0x0
R
R/W
Reserved
SW_RESET
0x10
[15:2]
[1:0]
0x000
0x0
R/W
R/W
Reserved
Mode
0x11
[15:14]
13
0x0
0x0
R/W
R/W
Reserved
RDOUT_MODE
12
0x1
R/W
FIFO_OVRN_PREVENT
[11:9]
[8:6]
0x0
0x0
R/W
R/W
Reserved
SLOTB_FIFO_MODE
5
0x0
R/W
SLOTB_EN
Address
0x09
0x0A
Description
Write 0xAD when writing to SLAVE_ADDRESS. Otherwise, do not access.
I2C slave address.
Do not access.
Reserved. Read only.
When the CLK32M_CAL_EN bit (Register 0x50, Bit 5) is set, the device
calculates the number of 32 MHz clock cycles in two cycles of the 32 kHz
clock. The result, nominally 2000 (0x07D0), is stored in the CLK_RATIO bits.
Enable changing the I2C address using Register 0x09.
0x04AD: enable address change always.
0x44AD: enable address change if INT is high.
0x84AD: enable address change if PDSO is high.
0xC4AD: enable address change if both INT and PDSO are high.
Reserved. Read only.
Software reset. Write 0x1 to reset the device. This bit clears itself after a
reset. This command does not return an acknowledge because the
command is instantaneous.
Write 0x000.
Determines the operating mode of the ADPD103.
0x0: standby.
0x1: program.
0x2: normal operation.
Write 0x0.
Readback data mode for extended data registers
0x0: Block sum of N samples
0x1: Block average of N samples
0x0: wrap around FIFO, overwriting old data with new.
0x1: new data if FIFO is not full (recommended setting).
Write 0x0.
Time Slot B FIFO data format.
0: no data to FIFO.
1: 16-bit sample in digital integration mode or 16-bit sum of all
4 channels when not in digital integration mode.
2: 32-bit sample in digital integration mode or 32-bit sum of all
4 channels when not in digital integration mode.
3: 16-bit sample and 16-bit background value in digital integration mode
4: 32-bit sample and 32-bit background value in digital integration
mode or 4 channels of 16-bit sample data for Time Slot B when not in
digital integration mode.
6: 4 channels of 32-bit extended sample data for Time Slot B when not in
digital integration mode.
Others: reserved.
The selected Time Slot B data is saved in the FIFO. Available only if Time
Slot A has the same averaging factor, N (Register 0x15, Bits[10:8] = Bits[6:4]),
or if Time Slot A is not saving data to the FIFO (Register 0x11, Bits[4:2] = 0).
Time Slot B enable. 1: enables Time Slot B.
Rev. B | Page 47 of 52
ADPD103
Data Sheet
Data
Bit
[4:2]
Default
0x0
Access
R/W
Name
SLOTA_FIFO_MODE
1
0
15
14
0x0
0x0
0x0
0x0
R/W
R/W
R/W
R/W
Reserved
SLOTA_EN
Reserved
EXT_SYNC_ENA
[13:0]
[15:9]
8
0x0
0x13
0x0
R/W
R/W
R/W
Reserved
Reserved
CLK32K_BYP
7
0x0
R/W
CLK32K_EN
6
[5:0]
0x0
0x12
R/W
R/W
Reserved
CLK32K_ADJUST
0x4D
[15:8]
[7:0]
0x42
0x5E
R/W
R/W
Reserved
CLK32M_ADJUST
0x4E 1
[15:0]
0x0060
R/W
ADC_TIMING1
Address
0x38
0x4B
Description
Time Slot A FIFO data format.
0: no data to FIFO.
1: 16-bit sample in digital integration mode or 16-bit sum of all
4 channels when not in digital integration mode.
2: 32-bit sample in digital integration mode or 32-bit sum of all
4 channels when not in digital integration mode.
3: 16-bit sample and 16-bit background value in digital integration mode
4: 32-bit sample and 32-bit background value in digital integration
mode or 4 channels of 16-bit sample data for Time Slot B when not in
digital integration mode.
6: 4 channels of 32-bit extended sample data for Time Slot B when not in
digital integration mode.
Others: reserved.
Write 0x0.
Time Slot A enable. 1: enables Time Slot A.
Write 0x0.
Enables external sampling trigger.
0x0: samples triggered internally.
0x1: samples triggered externally. Must be set to 1 if EXT_SYNC_SEL is b01 or
b10.
Write 0x0.
Write 0x13.
Bypass internal 32 kHz clock oscillator.
0x0: normal operation.
0x1: use an external clock on the PDSO pin.
Sample clock power-up. Enables the data sample clock.
0x0: clock disabled.
0x1: normal operation.
Write 0x0.
Data sampling (32 kHz) clock frequency adjust. This register is used to
calibrate the sample frequency of the device to achieve high precision
on the data rate as defined in Register 0x12. Adjusts the sample master
32 kHz clock by 0.6 kHz per LSB. For a 100 Hz sample rate as defined in
Register 0x12, 1 LSB of Register 0x4B, Bits[5:0], is 1.9 Hz.
Note that a larger value produces a lower frequency. See the Clocks and
Timing Calibration section for more information regarding clock
adjustment.
00 0000: maximum frequency.
10 0010: typical center frequency.
11 1111: minimum frequency.
Write 0x42.
Internal timing (32 MHz) clock frequency adjust. This register is used to
calibrate the internal clock of the device to achieve precisely timed LED
pulses. Adjusts the 32 MHz clock by 109 kHz per LSB.
See the Clocks and Timing Calibration section for more information
regarding clock adjustment.
0000 0000: minimum frequency.
0101 1110: default frequency.
1111 1111: maximum frequency.
0x0040: ADC clock speed = 1 MHz.
0x0060: ADC clock speed = 500 kHz.
Rev. B | Page 48 of 52
Data Sheet
Address
0x4F
0x50
0x5F
1
ADPD103
Data
Bit
[15:7]
6
5
4
[3:2]
Default
0x41
0x0
0x0
0x1
0x0
Access
R/W
R/W
R/W
R/W
R/W
Name
Reserved
PDSO_OE
PDSO_IE
Reserved
EXT_SYNC_SEL
1
0
[15:7]
6
0x0
0x0
0x000
0x0
R/W
R/W
R/W
R/W
INT_IE
Reserved
Reserved
PDSO_CTRL
5
0x0
R/W
CLK32M_CAL_EN
[4:0]
[15:3]
2
0x0
0x0000
0x0
R/W
R/W
R/W
Reserved
Reserved
SLOTB_DATA_HOLD
1
0x0
R/W
SLOTA_DATA_HOLD
0
0x0
R/W
FIFO_ACCESS_ENA
Description
Write 0x41.
PDSO pin output enable.
PDSO pin input enable.
Write 0x1.
Sample sync select.
00: use the internal 32 kHz clock with FSAMPLE to select sample timings.
01: use the INT pin to trigger sample cycle.
10: use the PDSO pin to trigger sample cycle.
11: reserved.
INT pin input enable.
Write 0x0.
Write 0x000.
Controls the PDSO output when the PDSO output is enabled (PDSO_OE =
0x1).
0x0: PDSO output driven low.
0x1: PDSO output driven by the AFE power-down signal.
As part of the 32 MHz clock calibration routine, write 1 to begin the clock
ratio calculation. Read the result of this calculation from the CLK_RATIO
bits in Register 0x0A.
Reset this bit to 0 prior to reinitiating the calculation.
Write 0x0.
Write 0x0000.
Setting this bit prevents the update of the data registers corresponding to
Time Slot B. Set this bit to ensure that unread data registers are not updated,
guaranteeing a contiguous set of data from all four photodiode channels.
1: hold data registers for Time Slot B.
0: allow data register update.
Setting this bit prevents the update of the data registers corresponding to
Time Slot A. Set this bit to ensure that unread data registers are not updated,
guaranteeing a contiguous set of data from all four photodiode channels.
1: hold data registers for Time Slot A.
0: allow data register update.
Set to 1 twice to enable FIFO access. It is necessary to write 1 to the
FIFO_ACCESS_ENA bit in two consecutive write operations in order to
read data from the FIFO. During clock calibration, set to 1 to force the
32 MHz clock to run. For power savings, reset to 0 when the previously
described operations are complete.
Clock speed setting is only relevant during digital integrate mode.
Rev. B | Page 49 of 52
ADPD103
Data Sheet
ADC REGISTERS
Table 25. ADC Registers
Data
Bits
[15:0]
Default
0x0028
Access
R/W
Name
FSAMPLE
[15:11]
[10:8]
0x0
0x6
R/W
R/W
Reserved
SLOTB_NUM_AVG
7
[6:4]
0x0
0x0
R/W
R/W
Reserved
SLOTA_NUM_AVG
0x18
[3:0]
[15:0]
0x0
0x2000
R/W
R/W
Reserved
SLOTA_CH1_OFFSET
0x19
[15:0]
0x2000
R/W
SLOTA_CH2_OFFSET
0x1A
[15:0]
0x2000
R/W
SLOTA_CH3_OFFSET
0x1B
[15:0]
0x2000
R/W
SLOTA_CH4_OFFSET
0x1E
[15:0]
0x2000
R/W
SLOTB_CH1_OFFSET
0x1F
[15:0]
0x2000
R/W
SLOTB_CH2_OFFSET
0x20
[15:0]
0x2000
R/W
SLOTB_CH3_OFFSET
0x21
[15:0]
0x2000
R/W
SLOTB_CH4_OFFSET
Address
0x12
0x15
Description
Sampling frequency: fSAMPLE = 32 kHz/(Register 0x12, Bits[15:0] × 4).
For example, 100 Hz = 0x0050; 200 Hz = 0x0028.
Write 0x0.
Sample sum/average for Time Slot B. Specifies the averaging factor, NB,
which is the number of consecutive samples that is summed and averaged
after the ADC. Register 0x70 to Register 0x7F hold the data sum. Register 0x64
to Register 0x6B and the data buffer in Register 0x60 hold the data average,
which can be used to increase SNR without clipping, in 16-bit registers.
The data rate is decimated by the value of the SLOTB_NUMB_AVG bits.
0: 1.
1: 2.
2: 4.
3: 8.
4: 16.
5: 32.
6: 64.
7: 128.
Write 0x0.
Sample sum/average for Time Slot A. NA: same as Bits[10:8] but for
Time Slot A. See description in Register 0x15, Bits[10:8].
Write 0x0.
Time Slot A Channel 1 ADC offset. The value to subtract from the raw ADC
value. A value of 0x2000 is typical.
Time Slot A Channel 2 ADC offset. The value to subtract from the raw ADC
value. A value of 0x2000 is typical.
Time Slot A Channel 3 ADC offset. The value to subtract from the raw ADC
value. A value of 0x2000 is typical.
Time Slot A Channel 4 ADC offset. The value to subtract from the raw ADC
value. A value of 0x2000 is typical.
Time Slot B Channel 1 ADC offset. The value to subtract from the raw ADC
value. A value of 0x2000 is typical.
Time Slot B Channel 2 ADC offset. The value to subtract from the raw ADC
value. A value of 0x2000 is typical.
Time Slot B Channel 3 ADC offset. The value to subtract from the raw ADC
value. A value of 0x2000 is typical.
Time Slot B Channel 4 ADC offset. The value to subtract from the raw ADC
value. A value of 0x2000 is typical.
Rev. B | Page 50 of 52
Data Sheet
ADPD103
DATA REGISTERS
Table 26. Data Registers
Address
0x60
Data
Bits
[15:0]
Access
R
Name
FIFO_DATA
0x64
0x65
0x66
0x67
0x68
0x69
0x6A
0x6B
0x70
0x71
0x72
0x73
0x74
0x75
0x76
0x77
0x78
0x79
0x7A
0x7B
0x7C
0x7D
0x7E
0x7F
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
[15:0]
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
SLOTA_CH1_16BIT
SLOTA_CH2_16BIT
SLOTA_CH3_16BIT
SLOTA_ CH4_16BIT
SLOTB_ CH1_16BIT
SLOTB_CH2_16BIT
SLOTB_CH3_16BIT
SLOTB_CH4_16BIT
SLOTA_CH1_LOW
SLOTA_CH2_LOW
SLOTA_CH3_LOW
SLOTA_CH4_LOW
SLOTA_CH1_HIGH
SLOTA_CH2_HIGH
SLOTA_CH3_HIGH
SLOTA_CH4_HIGH
SLOTB_CH1_LOW
SLOTB_CH2_LOW
SLOTB_CH3_LOW
SLOTB_CH4_LOW
SLOTB_CH1_HIGH
SLOTB_CH2_HIGH
SLOTB_CH3_HIGH
SLOTB_CH4_HIGH
Description
Next available word in FIFO. Prior to reading this register, set the FIFO_ACCESS_ENA bit
to 0x1 twice with two consecutive write operations to enable FIFO access (Register 0x5F,
Bit 0). Reset this bit to 0 when the FIFO access sequence is complete.
16-bit value of Channel1 in Time Slot A.
16-bit value of Channel 2 in Time Slot A.
16-bit value of Channel 3 in Time Slot A.
16-bit value of Channel 4 in Time Slot A.
16-bit value of Channel 1 in Time Slot B.
16-bit value of Channel 2 in Time Slot B.
16-bit value of Channel 3 in Time Slot B.
16-bit value of Channel 4 in Time Slot B.
Low data-word for Channel 1 in Time Slot A.
Low data-word for Channel 2 in Time Slot A.
Low data-word for Channel 3 in Time Slot A.
Low data-word for Channel 4 in Time Slot A.
High data-word for Channel 1 in Time Slot A.
High data-word for Channel 2 in Time Slot A.
High data-word for Channel 3 in Time Slot A.
High data-word for Channel 4 in Time Slot A.
Low data-word for Channel 1 in Time Slot B.
Low data-word for Channel 2 in Time Slot B.
Low data-word for Channel 3 in Time Slot B.
Low data-word for Channel 4 in Time Slot B.
High data-word for Channel 1 in Time Slot B.
High data-word for Channel 2 in Time Slot B.
High data-word for Channel 3 in Time Slot B.
High data-word for Channel 4 in Time Slot B.
Table 27. Required Start-Up Load Sequence
Step Number
1
2
3
4
Address
0x4B, Bit 7
0x10
Other registers
0x10
Comment
Write to 0x1 to enable the clock that drives the state machine.
Write 0x0001 to enter program mode.
Register order is not important while the device is in program mode.
Write 0x0002 to start normal sampling operation.
Rev. B | Page 51 of 52
ADPD103
Data Sheet
OUTLINE DIMENSIONS
0.25
0.20
0.15
4.10
4.00 SQ
3.90
PIN 1
INDICATOR
22
0.40
BSC
PIN 1
INDICATOR
28
1
21
2.70
2.60 SQ
2.50
EXPOSED
PAD
7
15
0.80
0.75
0.70
14
8
0.05 MAX
0.02 NOM
COPLANARITY
0.08
0.20 REF
PKG-003523
SEATING
PLANE
0.20 MIN
BOTTOM VIEW
FOR PROPER CONNECTION OF
THE EXPOSED PAD, REFER TO
THE PIN CONFIGURATION AND
FUNCTION DESCRIPTIONS
SECTION OF THIS DATA SHEET.
06-23-2015-B
0.45
0.40
0.35
TOP VIEW
COMPLIANT TO JEDEC STANDARDS MO-220-WGGE.
Figure 31. 28-Lead Lead Frame Chip Scale Package [LFCSP_WQ]
4 mm × 4 mm Body, Very Very Thin Quad
(CP-28-5)
Dimensions shown in millimeters
1.44
1.40
1.36
0.225
3
2
1
A
BALL A1
IDENTIFIER
B
2.00
REF
2.50
2.46
2.42
C
D
0.40
BSC
E
F
TOP VIEW
0.235
0.300
(BALL SIDE DOWN)
0.560
0.500
0.440
0.300
0.260
0.220
1
0.230
0.200
0.170
02-03-2015-B
COPLANARITY
0.05
PKG-004659
Model 1
ADPD103BCPZ
ADPD103BCPZRL
ADPD103BCBZRL7
EVAL-ADPD103Z-GEN
(BALL SIDE UP)
0.330
0.300
0.270
END VIEW
SEATING
PLANE
ORDERING GUIDE
BOTTOM VIEW
Figure 32. 16-Ball Wafer Level Chip Scale Package [WLCSP]
(CB-16-18)
Dimensions shown in millimeters
Temperature Range
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
Package Description
28-Lead LFCSP_WQ
28-Lead LFCSP_WQ
16-Ball WLCSP
Generic ADPD103 Evaluation Board
Z = RoHS Compliant Part.
I2C refers to a communications protocol originally developed by Philips Semiconductors (now NXP Semiconductors).
©2015–2016 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D12722-0-2/16(B)
Rev. B | Page 52 of 52
Package Option
CP-28-5
CP-28-5
CB-16-18
Similar pages