Arduino library for I2C ADC ADS1015, ADS1115, and similar.
For using I2C ADC with Raspberry pi or other SBC with Linux OS, you can check similar library here.
This library should work for the devices mentioned below, although not all sensors support all functionality.
| Device | Channels | Resolution | Max SPS | Comparator | Interrupts | ProgGainAMP | Notes |
|---|---|---|---|---|---|---|---|
| ADS1013 | 1 | 12 | 3300 | N | N | N | |
| ADS1014 | 1 | 12 | 3300 | Y | Y | Y | |
| ADS1015 | 4 | 12 | 3300 | Y | Y | Y | |
| ADS1113 | 1 | 16 | 860 | N | N | N | |
| ADS1114 | 1 | 16 | 860 | Y | Y | Y | |
| ADS1115 | 4 | 16 | 860 | Y | Y | Y | Tested |
As the ADS1015 and the ADS1115 are both 4 channels these are the most interesting from functionality point of view as these can do differential measurements.
Besides polling the ADS1x14 and ADS1x15 support interrupts to maximize throughput with minimal latency. For this these device has an ALERT/RDY pin. This pin can be used both for interrupts or polling, see table of examples below.
| example | Interrupts | notes |
|---|---|---|
| ADS_1114_two_continuous.ino | Y | |
| ADS_continuous_3_channel.ino | Y | |
| ADS_continuous_4_channel.ino | Y | |
| ADS_continuous_8_channel.ino | Y | |
| ADS_continuous_differential.ino | Y | |
| ADS_high_speed_differential.ino | Y | |
| ADS_read_async_rdy.ino | polling | |
| ADS_read_RDY.ino | polling |
The examples of this library all use the RISING edge for the interrupt detection of the ALERT / RDY pin. In #87 it is observed that the FALLING edge gave far more stable results for the application used (determine True RMS). This effect can not be explained as the edges are only 8 us apart. Thus changing the edge to FALLING might improve your measurements too.
Datasheet section 7.3.8 Conversion ready pin, figure 7-8 indicates using the FALLING edge as the moment the conversion is ready.
ejdp62 explained the detailed behaviour of the ALERT / RDY pin See #87 (comment)
As the ALERT/RDY pin is open collector, the rising edges of the output signal and the falling edges of the output signal are quite different! All the scope pictures show this difference. The falling edges are much more clearly defined, because the signal is actively pulled low, whereas it is only passively pulled high. And that is true regardless of the polarity chosen, as it is a consequence of the output cell in the chip (open collector driver), and not of the signal that the I/O cell is passing to the ready pin. I think this does explain all observations made.
Now that the data sheet has been updated to explicitly show that the conversion is ready on the second slope of the pulse, it is clear what to do best: set the polarity such that that second slope is falling (as that is the superior direction), and use that one to trigger the interrupt. It may be appropriate to adapt the code in the library to implement only that choice, as it will indeed give the best results.
Fixed #80, setComparatorPolarity() and setComparatorLatch() as these inverted the setting.
Version 0.4.0 introduced a breaking change. You cannot set the pins in begin() any more. This reduces the dependency of processor dependent Wire / I2C implementations. The user has to call Wire.begin() and can optionally set the I2C pins before calling begin().
- https://github.com/RobTillaart/AD7367 2 channel simultaneous 14 bit ADC.
- https://github.com/RobTillaart/ADC081S 10-12 bit, single channel ADC
- https://github.com/RobTillaart/ADC08XS 10-12 bit, 2 + 4 channel ADC
- https://gammon.com.au/adc tutorial about ADC's (UNO specific)
- https://github.com/RobTillaart/MCP_ADC 10-12 bit, 1,2,4,8 channel ADC
- https://github.com/RobTillaart/ADS1x15
- https://github.com/RobTillaart/PCF8591 8 bit single ADC (+ 1 bit DAC)
- https://github.com/RobTillaart/AD5593R 8 channel ADC / DAC / GPIO device.
The I2C address of the ADS1113 /14 /15 is determined by to which pin the ADDR is connected to:
| ADDR pin connected to | Address | Notes |
|---|---|---|
| GND | 0x48 | default |
| VDD | 0x49 | |
| SDA | 0x4A | |
| SCL | 0x4B |
Sometimes you need to control more devices than possible with the default address range the device provides. This is possible with an I2C multiplexer e.g. TCA9548 which creates up to eight channels (think of it as I2C subnets) which can use the complete address range of the device.
Drawback of using a multiplexer is that it takes more administration in your code e.g. which device is on which channel. This will slow down the access, which must be taken into account when deciding which devices are on which channel. Also note that switching between channels will slow down other devices too if they are behind the multiplexer.
#include "ADS1X15.h"To initialize the library you must call a constructor as described below.
- ADS1x15() base constructor, should not be used.
- ADS1013(uint8_t address, TwoWire *wire = &Wire) Constructor with device address, and optional the Wire interface as parameter.
- ADS1014(uint8_t address, TwoWire *wire = &Wire) Constructor with device address, and optional the Wire interface as parameter.
- ADS1015(uint8_t address, TwoWire *wire = &Wire) Constructor with device address, and optional the Wire interface as parameter.
- ADS1113(uint8_t address, TwoWire *wire = &Wire) Constructor with device address, and optional the Wire interface as parameter.
- ADS1114(uint8_t address, TwoWire *wire = &Wire) Constructor with device address, and optional the Wire interface as parameter.
- ADS1115(uint8_t address, TwoWire *wire = &Wire) Constructor with device address, and optional the Wire interface as parameter.
After construction the ADS.begin() must be called, typical in setup().
- bool begin() Returns false if an invalid address is used.
- bool isConnected() is used to check if the device address is visible on I2C.
- void reset() sets the internal parameters to their initial value as in the constructor.
For example.
#include "ADS1X15.h"
// initialize ADS1115 on I2C bus 1 with default address 0x48
ADS1115 ADS(0x48);
void setup()
{
if (!ADS.begin())
{
// invalid address ADS1115 or 0x48 not found
}
if (!ADS.isConnected())
{
// address 0x48 not found
}
}The function void setWireClock(uint32_t speed = 100000) is used to set the clock speed in Hz of the used I2C interface. Typical value is 100 KHz.
The function uint32_t getWireClock() is a prototype. It returns the value set by setWireClock(). This is not necessary the actual value. When no value is set getWireClock() returns 0. Need to implement a read / calculate from low level I2C code (e.g. TWBR on AVR), better the Arduino Wire lib should support this call (ESP32 does).
See - arduino/Arduino#11457
Question: Should this functionality be in this library?
- void setGain(uint8_t gain) set the gain value, indicating the maxVoltage that can be measured Adjusting the gain allowing to make more precise measurements. Note: the gain is not set in the device until an explicit read/request of the ADC (any read call will do). See table below.
- uint8_t getGain() returns the gain value (index).
| define | PGA value | Max Voltage | Notes |
|---|---|---|---|
| ADS1X15_GAIN_6144MV | 0 | ±6.144V | default |
| ADS1X15_GAIN_4096MV | 1 | ±4.096V | |
| ADS1X15_GAIN_2048MV | 2 | ±2.048V | |
| ADS1X15_GAIN_1024MV | 4 | ±1.024V | |
| ADS1X15_GAIN_0512MV | 8 | ±0.512V | |
| ADS1X15_GAIN_0256MV | 16 | ±0.256V |
- float getMaxVoltage() returns the max voltage with the current gain.
- float toVoltage(float raw = 1) converts a raw measurement to a voltage. Can be used for normal and differential measurements. The default value of 1 returns the conversion factor for any raw number. The parameter is made float to allow the average of multiple raw measurements to be converted.
The voltage factor can also be used to set HIGH and LOW threshold registers with a voltage in the comparator mode. Check the examples.
float f = ADS.toVoltage();
ADS.setComparatorThresholdLow( 3.0 / f );
ADS.setComparatorThresholdHigh( 4.3 / f );The ADS sensor can operate in single shot or continuous mode. Depending on how often conversions needed you can tune the mode.
- void setMode(uint8_t mode) 0 = CONTINUOUS, 1 = SINGLE (default) Note: the mode is not set in the device until an explicit read/request of the ADC (any successful read() call will do).
- uint8_t getMode() returns current mode 0 or 1, or ADS1X15_INVALID_MODE = 0xFE.
| define | value | Notes |
|---|---|---|
| ADS1X15_MODE_CONTINUOUS | 0 | |
| ADS1X15_MODE_SINGLE | 1 | default |
- void setDataRate(uint8_t dataRate) Data rate depends on type of device. For all devices the index 0..7 can be used, see table below. Values above 7 ==> will be set to the default 4. Note: the data rate is not set in the device until an explicit read/request of the ADC (any read call will do).
- uint8_t getDataRate() returns the current data rate (index).
The library has no means to convert this index to the actual numbers as that would take 32 bytes.
Data rate in samples per second, based on datasheet is described on table below.
| define | data rate | ADS101x | ADS111x | Notes |
|---|---|---|---|---|
| ADS1X15_DATARATE_0 | 0 | 128 | 8 | slowest |
| ADS1X15_DATARATE_1 | 1 | 250 | 16 | |
| ADS1X15_DATARATE_2 | 2 | 490 | 32 | |
| ADS1X15_DATARATE_3 | 3 | 920 | 64 | |
| ADS1X15_DATARATE_4 | 4 | 1600 | 128 | default |
| ADS1X15_DATARATE_5 | 5 | 2400 | 250 | |
| ADS1X15_DATARATE_6 | 6 | 3300 | 475 | |
| ADS1X15_DATARATE_7 | 7 | 3300 | 860 | fastest |
Reading the ADC is very straightforward, the readADC() function handles all in one call. Under the hood it uses the asynchronous calls.
- int16_t readADC(uint8_t pin = 0) normal ADC functionality, pin = 0..3. If the pin number is out of range, this function will return 0 (seems safest). Default pin = 0 as this is convenient for the single channel devices.
// read ADC in pin 2
ADS.readADC(2);
// read ADC in pin 0 - two ways
ADS.readADC();
ADS.readADC(0);See examples.
The readADC() can return ADS1X15_ERROR_TIMEOUT (-101) which is an errorcode. This may conflict with a possible actual value of -101. Therefore the user should check with getError() if an error has occurred after reading the ADC.
Value = ADS.readADC()
if (ADS.getError() == ADS1X15_OK)
// Use value
else
// handle errorThe error handling within the library need to be improved, see also issue #84.
To read the ADC in an asynchronous way (e.g. to minimize blocking) you need call three functions:
- void requestADC(uint8_t pin = 0) Start the conversion. pin = 0..3. Default pin = 0 as this is convenient for 1 channel devices.
- bool isBusy() Is the conversion not ready yet? Works only in SINGLE mode!
- bool isReady() Is the conversion ready? Works only in SINGLE mode! (= wrapper around isBusy() )
- int16_t getValue() Read the result of the conversion.
in terms of code:
void setup()
{
// other setup things here
ADS.setMode(ADS1X15_MODE_SINGLE);
ADS.requestADC(pin);
}
void loop()
{
if (ADS.isReady())
{
value = ADS.getValue();
ADS.requestADC(pin); // request new conversion
}
// do other things here
}See examples.
For reading the ADC in a differential way there are 4 calls possible.
- int16_t readADC_Differential_0_1() returns the difference between 2 ADC pins.
- int16_t readADC_Differential_0_3() ADS1x15 only
- int16_t readADC_Differential_1_3() ADS1x15 only
- int16_t readADC_Differential_2_3() ADS1x15 only
- int16_t readADC_Differential_0_2() ADS1x15 only - in software (no async equivalent)
- int16_t readADC_Differential_1_2() ADS1x15 only - in software (no async equivalent)
// read differential ADC between pin 0 and 1
ADS.readADC_Differential_0_1(0);The differential reading of the ADC can also be done with asynchronous calls.
- void requestADC_Differential_0_1() starts conversion for differential reading
- void requestADC_Differential_0_3() ADS1x15 only
- void requestADC_Differential_1_3() ADS1x15 only
- void requestADC_Differential_2_3() ADS1x15 only
After one of these calls you need to call
- int16_t getValue() Read the result of the last conversion.
See examples.
Since 0.3.12 the library tracks the last request mode, single pin or differential. This variable is set at the moment of request, and keeps its value until a new request is made. This implies that the value / request can be quite old.
Values >= 0x10 are differential, values < 0x10 are single pin.
- uint8_t lastRequest() returns one of the values below.
| Value | Description | Notes |
|---|---|---|
| 0xFF | no (invalid) request made | after call constructor. |
| 0x00 | single pin 0 | |
| 0x01 | single pin 1 | |
| 0x02 | single pin 2 | |
| 0x03 | single pin 3 | |
| 0x10 | differential pin 1 0 | |
| 0x30 | differential pin 3 0 | |
| 0x31 | differential pin 3 1 | |
| 0x32 | differential pin 3 2 |
Please note that (for now) the function does not support a descriptive return value for the following two requests:
- readADC_Differential_0_2() ADS1x15 only - in software (no async equivalent)
- readADC_Differential_1_2() ADS1x15 only - in software (no async equivalent)
As these are emulated in software by two single pin calls, the state would be one of the two single pin values.
To use the continuous mode you need call three functions:
- void setMode(0) 0 = CONTINUOUS, 1 = SINGLE (default). Note: the mode is not set in the device until an explicit read/request of the ADC (any read call will do).
- int16_t readADC(uint8_t pin) or void requestADC(uint8_t pin) to get the continuous mode started.
- int16_t getValue() to return the last value read by the device. Note this can be a different pin, so be warned. Calling this over and over again can give the same value multiple times.
void setup()
{
// configuration things here
ADS.setMode(ADS1X15_MODE_CONTINUOUS);
// request on pin 0
ADS.requestADC(0);
}
void loop()
{
value = ADS.getValue()
sleep(1)
}See examples . By using bool isBusy() or bool isReady() one can wait until new data is available. Note this only works in the SINGLE_SHOT modus.
In continuous mode, you can't use isBusy() or isReady() functions to wait until new data available. Instead you can configure the threshold registers to allow the ALERT/RDY pin to trigger an interrupt signal when conversion data ready.
When switching the operating mode or the ADC channel in continuous mode, be aware that the device will always finish the running conversion. This implies that after switching the mode or channel the first sample you get is probably the last sample with the previous settings, e.g. channel. This might be a problem for your project as this value can be in an "unexpected" range (outlier).
The robust way to change mode or channel therefore seems to be:
- stop continuous mode,
- wait for running conversion to be ready,
- reject the last conversion or process it "under old settings",
- change the settings,
- restart (continuous mode) with the new settings.
This explicit stop takes extra time, however it should prevent "incorrect" readings.
(need to be verified with different models)
(datasheet 9.3.8)
Conversion Ready Pin (ADS1114 and ADS1115 Only)
The ALERT/RDY pin can also be configured as a conversion ready pin. Set the most-significant bit of the
Hi_thresh register to 1 and the most-significant bit of Lo_thresh register to 0 to enable the pin as a conversion
ready pin.
If the thresholdHigh is set to 0x8000 and the thresholdLow to 0x0000 the ALERT/RDY pin is triggered when a conversion is ready.
- void setComparatorThresholdLow(int16_t lo) writes value to device directly.
- void setComparatorThresholdHigh(int16_t hi) writes value to device directly.
- int16_t getComparatorThresholdLow() reads value from device.
- int16_t getComparatorThresholdHigh() reads value from device.
See examples.
Please read Page 15 of the datasheet as the behaviour of the comparator is not trivial.
NOTE: all comparator settings are copied to the device only after calling readADC() or requestADC() functions.
When configured as a TRADITIONAL comparator, the ALERT/RDY pin asserts (active low by default) when conversion data exceed the limit set in the high threshold register. The comparator then de-asserts when the input signal falls below the low threshold register value.
- void setComparatorMode(uint8_t mode) value 0 = TRADITIONAL 1 = WINDOW,
- uint8_t getComparatorMode() returns value set.
| define | value | Notes |
|---|---|---|
| ADS1x15_COMP_MODE_TRADITIONAL | 0 | |
| ADS1x15_COMP_MODE_WINDOW | 1 |
If the comparator LATCH is set, the ALERT/RDY pin asserts and it will be reset after reading the sensor (conversion register) again. An SMB alert command (00011001) on the I2C bus will also reset the alert state. Not implemented in the library (yet)
In WINDOW comparator mode, the ALERT/RDY pin asserts if conversion data exceeds the high threshold register or falls below the low threshold register. In this mode the alert is held if the LATCH is set. This is similar as above.
Default state of the ALERT/RDY pin is LOW, which can be to set HIGH.
- void setComparatorPolarity(uint8_t pol) Flag is only explicitly set after a readADC() or a requestADC()
- uint8_t getComparatorPolarity() returns value set.
| define | value | Notes |
|---|---|---|
| ADS1x15_COMP_POL_FALLING_EDGE | 0 | |
| ADS1x15_COMP_POL_RISING_EDGE | 1 |
From tests (see #76) it became clear that the behaviour of the ALERT/RDY pin looks ambiguous. Further investigation eventually showed that the behaviour is logical but one should not think in "pulses", more in levels and edges.
In the continuous mode it looks like an 8us pulse, however this "pulse" is actual a short time (8 us) of IDLE followed by a long time pulse of converting.
In the single shot mode it looks like the converting time is the pulse as that is the only single change visible. This is IMHO the correct view.
| MODE | COMP_POL | CONVERT | COMPLETING | READY |
|---|---|---|---|---|
| 0 = continuous | 0 = LOW | LOW | RISING | FALLING |
| 0 = continuous | 1 = HIGH | HIGH | FALLING | RISING |
| 1 = single | 0 = LOW | LOW | RISING | FALLING |
| 1 = single | 1 = HIGH | HIGH | FALLING | RISING |
See issue #76 and #87 for some screenshots with an oscilloscope. Furthermore #87 gives an explanation why to prefer the FALLING edges (from ejdp62). See also above in description section.
See also Rev. D data sheet, Page 17 Figure 7-8 Conversion Ready Pulse in Continuous-Conversion Mode
| data rate | convert time | Notes |
|---|---|---|
| 0 | 125 ms | |
| 1 | 62 ms | |
| 2 | 32 ms | |
| 3 | 16 ms | |
| 4 | 8 ms | default, see in table above. |
| 5 | 4.4 ms | |
| 6 | 2.4 ms | |
| 7 | 1.2 ms |
Times are estimates from scope.
- Conversion generates a conversion pulse with length depending on the data rate.
- In continuous mode it looks like there is a short pulse, but actual the long period is the conversion pulse.
In short:
-
if COMP_POL == 0,
- a LOW level indicates converting.
- a RISING edge indicates conversion completing.
- a FALLING edge indicates conversion ready.
-
if COMP_POL == 1,
- a HIGH level indicates converting.
- a FALLING edge indicates conversion completing.
- a RISING edge indicates conversion ready.
This interpretation is in line with all tests done in #76 and #87.
See also Rev. D data sheet, Page 17 Figure 7-8 Conversion Ready Pulse in Continuous-Conversion Mode
Holds the ALERT/RDY to HIGH (or LOW depending on polarity) after triggered even if actual value has been 'restored to normal' value.
- void setComparatorLatch(uint8_t latch) 0 = NO LATCH, not 0 = LATCH
- uint8_t getComparatorLatch() returns value set.
| define | value | Notes |
|---|---|---|
| ADS1x15_COMP_POL_NOLATCH | 0 | |
| ADS1x15_COMP_POL_LATCH | 1 |
The (no-)latch is not verified in detail yet.
Set the number of conversions before trigger activates.
The void setComparatorQueConvert(uint8_t mode) is used to set the number of conversions that exceed the threshold before the ALERT/RDY pin is set HIGH. A value of 3 (or above) effectively disables the comparator. See table below.
See examples.
- void setComparatorQueConvert(uint8_t mode) See table below.
- uint8_t getComparatorQueConvert() returns the value set.
| value | meaning | Notes |
|---|---|---|
| 0 | trigger alert after 1 conversion | |
| 1 | trigger alert after 2 conversions | |
| 2 | trigger alert after 4 conversions | |
| 3 | Disable comparator | default |
| other | Disable comparator |
To enable the conversion-ready function of the ALERT/RDY pin, it is necessary to set the MSB of the Hi_threshold register to 1 (value 0x8000) and the MSB of the Lo_threshold register to 0. See section Threshold registers above.
Depending on the comparator mode TRADITIONAL or WINDOW the thresholds registers mean something different see - Comparator Mode above or datasheet.
- void setComparatorThresholdLow(int16_t lo) set the low threshold; take care the hi >= lo.
- void setComparatorThresholdHigh(int16_t hi) set the high threshold; take care the hi >= lo.
- int16_t getComparatorThresholdLow() reads value from device.
- int16_t getComparatorThresholdHigh() reads value from device.
This section has to be elaborated.
Some functions return or set an error value. This is read and reset by getError().
| Value | Define | Description |
|---|---|---|
| 0 | ADS1X15_OK | idem. |
| -100 | ADS1X15_INVALID_VOLTAGE | getMaxVoltage() |
| -101 | ADS1X15_ERROR_TIMEOUT | readADC() device did not respond in time. |
| -102 | ADS1X15_ERROR_I2C | I2C communication failure. |
| 0xFF | ADS1X15_INVALID_GAIN | getGain() |
| 0xFE | ADS1X15_INVALID_MODE | getMode() |
- uint16_t getMaxRegValue() returns 32767 or 2047 depending on ADC bits. (See #91).
- Improve documentation (always)
- split off separate topics?
- Investigate ADS1118 library which should be a similar SPI based ADC.
- improve error handling
- refactor values to be more logic.
- improve API. #84
- (big) breaking change!
- SMB alert command (00011001) on I2C bus?
- Sync code order .h / .cpp
- Type flag?
- Constructor for ADS1X15? No as all types are supported.
- Remove getWireClock() as this is not an ADC library function but a responsibility of the I2C library. Until fixed in Wire I leave it.
If you appreciate my libraries, you can support the development and maintenance. Improve the quality of the libraries by providing issues and Pull Requests, or donate through PayPal or GitHub sponsors.
Thank you,