drivers: add support for ADS1115 ADC #21613

baptleduc · 2025-07-17T13:48:33Z

Contribution description

This PR adds initial support for the [Texas Instruments ADS1115 analog-to-digital converter (ADC)] (https://www.ti.com/lit/ds/symlink/ads1115.pdf?ts=1752581207840&ref_url=https%253A%252F%252Fro.mouser.com%252F) by introducing a dedicated RIOT driver.

The driver enables:

Basic device initialization and communication over I2C
Configuration of key ADS1115 parameters (PGA, MUX, mode, data rate, comparator) according to pages 25-26 of the [datasheet] (https://www.ti.com/lit/ds/symlink/ads1115.pdf?ts=1752581207840&ref_url=https%253A%252F%252Fro.mouser.com%252F)
Reading raw ADC conversion results
Converting raw readings into millivolts based on the selected full-scale voltage range (PGA)

This implementation lays the groundwork for further extensions such the other ADS111x, comparator handling etc...

Testing procedure

Issues/PRs references

#21612

benpicco

Hello and welcome to RIOT!
Nice first driver, I added some comments - the static test scripts also added their 2¢

drivers/include/ads1115.h

drivers/ads1115/ads1115.c

benpicco · 2025-07-17T14:49:38Z

drivers/include/ads1115.h

+ */
+typedef struct {
+    i2c_t i2c;                              /**< I2C device */
+    uint8_t addr;                           /**< I2C address */


Is this configurable?

What do you mean by ‘configurable’? This should depend on the board you're using. If you mean that it can be modified dynamically after init, no, that's not the case.

Ok, some devices have a fixed I2C address (then there is no need to make it configurable in the code), for others the I2C address can be set via strap resistors.

Ok so should I keep this like that ?

benpicco · 2025-07-17T14:51:52Z

drivers/include/ads1115.h

+    ads1115_comp_mode_t comp_mode;          /**< Comparator mode */
+    ads1115_comp_polarity_t comp_polarity;  /**< Comparator polarity */
+    ads1115_comp_latch_t comp_latch;        /**< Comparator latch */
+    ads1115_comp_queue_t comp_queue;        /**< Comparator queue */


Does the device only have a single input channel / input pin?
Otherwise it could make sense to set those at run-time for the different channels depending on what one wants to measure.

No the device can handle four different channel. So maybe I would have to add 4 set functions these four parameters so it can be configure at run-time. What do you think ?

benpicco · 2025-07-17T14:52:58Z

drivers/include/ads1115.h

+typedef struct {
+    i2c_t i2c;                              /**< I2C device */
+    uint8_t addr;                           /**< I2C address */
+    ads1115_mux_t mux;                      /**< Input multiplexer configuration */


You also have ads1115_set_ain_ch_input(), why also the compile-time config?

It's just to make sure we have a default configuration with a default channel entry. Should I remove it?

A single way to set the configuration is enough.

but the user may want to get the inputs from differents channels during run-time so the ads1115_set_ain_ch_input() should be necessary, right ?

benpicco · 2025-07-17T14:53:36Z

drivers/ads1115/ads1115.c

+    i2c_acquire(DEV);
+
+    // Read conversion register
+    if (i2c_read_regs(DEV, ADDR, ADS1115_REG_CONVERSION, buf, 2, 0) < 0) {


Do you need to wait for a converstion to finish?

Honestly I don´t know. Do you think the call to i2c_read_regs is a blocking call or we directly read the conversion register ?

The data sheet should be able to tell you 😉

Devices can implement clock stretching if they are not ready yet, but this is rather rare.

How is the sampling started?

In single-shot mode (when bit[8] MODE = 1), writing a 1 to the OS bit (bit[15]) starts a conversion.
In continuous mode (when bit[8] MODE=0), the device get the input value continuously and store it into the conversion register

I personally think that reading in the conversion register is not a blocking action and we shouldn't wait

benpicco · 2025-07-17T14:54:47Z

drivers/include/ads1115.h

+ */
+typedef enum {
+    ADS1115_COMP_MODE_TRADITIONAL   = 0,    /**< Traditional comparator (default) */
+    ADS1115_COMP_MODE_WINDOW        = 1     /**< Window comparator */


That sounds like it would generate an interrupt if the comperator matches, but you don't have anything like that in the code - if you don't need it, just leave it out.

I don't actually use it, I just add it to have a fully customisable configuration as explained on pages 25-26 of the datasheet.

But what good is it to configure something that you then can not use?

Yes you right, it for the scalability and to be sure that the default config is loaded when calling the init function

@benpicco What do you think about that ?

riot-ci · 2025-07-17T18:11:19Z

Murdock results

✔️ PASSED

6496914 Merge branch 'RIOT-OS:master' into add-ads1115-driver

Success	Failures	Total	Runtime
10536	0	10536	22m:21s

Artifacts

Documentation preview

Co-authored-by: benpicco <[email protected]>

crasbe

Thank you for your addition of the ADS1115 driver. In the review I have addressed some general styling related things and Doxygen related issues.

Also, as in the other PR, please adapt the title and commit message according to our commit convention: https://github.com/RIOT-OS/RIOT/blob/master/CONTRIBUTING.md#commit-conventions

A possible title and commit message could be: "drivers: add support for ADS1115 ADC" or similar.

drivers/ads1115/ads1115.c

drivers/include/ads1115.h

drivers/ads1115/include/ads1115_params.h

drivers/ads1115/ads1115.c

baptleduc · 2025-07-22T19:28:04Z

@crasbe Do you know why the Uncrustify formatter doesn't reformat lines longer than 100 characters? It causes the static test to fail, and manually fixing those lines is quite tedious.

crasbe · 2025-07-22T19:30:50Z

Unfortunately I am not super familiar with uncrustify and what it can and cannot do :(

drivers/include/ads1115.h

drivers/ads1115/include/ads1115_params.h

crasbe

You don't have to re-request a review on every change, I get an e-mail notification anyway 😅

crasbe · 2025-07-28T10:48:29Z

drivers/ads1115/include/ads1115_params.h

+    ADS1115_PARAMS
+};
+
+/** @} */  /* close @ingroup drivers_ads1115 */ 


Suggested change

/** @} */ /* close @ingroup drivers_ads1115 */

/** @} */ /* close @ingroup drivers_ads1115 */

crasbe · 2025-07-28T10:48:40Z

drivers/include/ads1115.h

+    ADS1115_PGA_0_512V, /**< FSR = ±0.512V */
+    ADS1115_PGA_0_256V, /**< FSR = ±0.256V */
+
+    /* Aliases (same behavior) */ 


Suggested change

/* Aliases (same behavior) */

/* Aliases (same behavior) */

crasbe · 2025-07-28T10:54:34Z

drivers/include/ads1115.h

+ * @brief      I2C Analog-to-digital converter with programmable gain amplifier.
+ *             The device has four input channels: AIN0, AIN1, AIN2, and AIN3,
+ *             that can be selected by using @ref ads1115_set_ain_ch_input.
+ *


Suggested change

*

* @{

*

Without the opening bracket, Doxygen seems to ignore the content of this file:

crasbe · 2025-07-28T10:55:28Z

drivers/include/ads1115.h

+
+#ifdef __cplusplus
+}
+#endif


Suggested change

#endif

#endif

/** @} */ /* close @defgroup drivers_ads1115 */

baptleduc · 2025-07-28T17:41:54Z

You don't have to re-request a review on every change, I get an e-mail notification anyway 😅

I'm sorry, I didn't know that it sends you an email on every change, it's my first open-source contribution. I'm going to pay attention for the future

crasbe · 2025-07-29T22:53:53Z

I just searched for something else in the RIOT code and discovered that there already is a driver that supports the ADC111x ADCs 🤔 😅
https://github.com/RIOT-OS/RIOT/tree/master/drivers/ads101x
https://github.com/RIOT-OS/RIOT/tree/master/tests/drivers/ads101x

Your driver seems to have the nicer code and more documentation, but having two drivers for the same purpose is a bit redundant.

Perhaps you could take a look at the ads101x driver and give an estimate what the best path forward would be here.

baptleduc · 2025-07-30T08:34:24Z

I just searched for something else in the RIOT code and discovered that there already is a driver that supports the ADC111x ADCs 🤔 😅

https://github.com/RIOT-OS/RIOT/tree/master/drivers/ads101x

https://github.com/RIOT-OS/RIOT/tree/master/tests/drivers/ads101x

Your driver seems to have the nicer code and more documentation, but having two drivers for the same purpose is a bit redundant.

Perhaps you could take a look at the ads101x driver and give an estimate what the best path forward would be here.

That's a shame..., in fact the integrated driver supports ads111x despite its lack of documentation. What do you suggest ? Should I abandon this driver or update the existant code to document it better?

crasbe · 2025-07-31T13:48:51Z

I'll have to check what the best approach is.

Enoch247 · 2025-07-31T16:53:13Z

That's a shame..., in fact the integrated driver supports ads111x despite its lack of documentation. What do you suggest ? Should I abandon this driver or update the existant code to document it better?

I'm sorry to be the bearer of bad news. Yes we should not have duplicate drivers, without some compelling reason. We would very much appreciate if you would open a PR to port any improvements in code and documentation over to the existing driver from your work here though.

Enoch247

This driver duplicates an existing one already in RIOT. See here.

baptleduc · 2025-08-26T08:09:51Z

@Enoch247 @crasbe The problem is that the existing driver seems to be made for ADS101x and it not fully support the ADS111x family. For example, the data rate configuration description of the ADS101x is not the same as the ADS111x here. For me, it would make sense to write a dedicated driver for the ADS111x even if it means writing a common file for commons functions (ads_init for example)

Enoch247 · 2025-08-28T01:36:10Z

It is not uncommon to write drivers to have a shared common part in one module along with device specific parts that are compiled on demand as separate modules or pseudomodules. The ST77xx is one example. Is this the approach you have in mind?

baptleduc · 2025-08-28T07:16:22Z

Yes, exactly! We could use a public header ads1x1x and separate private headers for each variant (ads1114, ads1013, etc.). What do you think?

Enoch247 · 2025-09-01T14:56:15Z

I'm not familiar with the devices, but seems like a reasonable approach.

crasbe · 2025-09-04T14:05:49Z

Replaced by ~~#21693~~ #21694.

baptleduc · 2025-09-04T14:26:58Z

Replaced by #21693.

in fact it is #21694

github-actions bot added the Area: drivers Area: Device drivers label Jul 17, 2025

baptleduc changed the title ~~feat(ads1115): Write an ads1115 drive~~ feat(ads1115): Write an ads1115 driver Jul 17, 2025

benpicco reviewed Jul 17, 2025

View reviewed changes

crasbe added Type: new feature The issue requests / The PR implemements a new feature for RIOT CI: ready for build If set, CI server will compile all applications for all available boards for the labeled PR labels Jul 17, 2025

crasbe linked an issue Jul 17, 2025 that may be closed by this pull request

feat: Add support for ADS1115 ADC (I2C) #21612

Open

6 tasks

baptleduc force-pushed the add-ads1115-driver branch 3 times, most recently from 2e9f348 to 9d7167f Compare July 18, 2025 12:40

baptleduc requested review from MrKevinWeiss, aabadie, jia200x and leandrolanzieri as code owners July 18, 2025 12:40

github-actions bot added Area: boards Area: Board ports Area: Kconfig Area: Kconfig integration labels Jul 18, 2025

baptleduc and others added 6 commits July 18, 2025 15:26

feat: initial driver with init()

196bcda

feat: add read, set input and convert to mV funcs

cc79b87

style: uncrustify formatting

0f1490f

refactor: return PGA values as int mV instead of float

cde168d

Co-authored-by: benpicco <[email protected]>

refactor: use htons/ntohs for byte swapping

4a409b1

refactor: move a private inline static function to .c file

b71cc2f

baptleduc force-pushed the add-ads1115-driver branch from 9d7167f to b71cc2f Compare July 18, 2025 13:27

github-actions bot removed Area: boards Area: Board ports Area: Kconfig Area: Kconfig integration labels Jul 18, 2025

crasbe reviewed Jul 19, 2025

View reviewed changes

baptleduc changed the title ~~feat(ads1115): Write an ads1115 driver~~ drivers: add support for ADS1115 ADC Jul 21, 2025

baptleduc added 3 commits July 21, 2025 21:04

driver: change C++ comments style to C style

99230a9

driver: remove two consecutives empty lines

e24b925

driver: adapt comments to Doxygen convention

8d07b22

baptleduc requested a review from crasbe July 21, 2025 19:23

baptleduc added 3 commits July 22, 2025 00:50

driver: indent #define macro and add comments to them

7913060

driver: change copyright statements to SPDX format

9ebfa9d

driver: remove leading empty line

915158c

baptleduc force-pushed the add-ads1115-driver branch from a4c2fe4 to 915158c Compare July 21, 2025 22:50

drivers: add missing doxygen docstring

5c30d11

driver: reformat comments to be under 100 characters

8ee90ac

crasbe reviewed Jul 22, 2025

View reviewed changes

drivers/include/ads1115.h Outdated Show resolved Hide resolved

drivers/ads1115/include/ads1115_params.h Outdated Show resolved Hide resolved

driver: change missing C++ comments style to C style

b4b33ab

baptleduc requested a review from crasbe July 27, 2025 14:41

crasbe reviewed Jul 28, 2025

View reviewed changes

Enoch247 requested changes Jul 31, 2025

View reviewed changes

baptleduc requested a review from Enoch247 August 27, 2025 07:24

Merge branch 'RIOT-OS:master' into add-ads1115-driver

6496914

baptleduc mentioned this pull request Sep 4, 2025

drivers: add a common driver for ADS1X1X #21693

Closed

crasbe closed this Sep 4, 2025

baptleduc mentioned this pull request Sep 4, 2025

drivers: add a common driver for ADS1X1X #21694

Open

	/** @} / / close @ingroup drivers_ads1115 */
	/** @} / / close @ingroup drivers_ads1115 */

-#endif
+#endif
+/** @} */ /* close @defgroup drivers_ads1115 */

drivers: add support for ADS1115 ADC #21613

drivers: add support for ADS1115 ADC #21613

Uh oh!

Conversation

baptleduc commented Jul 17, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Contribution description

Testing procedure

Issues/PRs references

Uh oh!

benpicco left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

benpicco Jul 18, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

benpicco Jul 18, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

riot-ci commented Jul 17, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Murdock results

Artifacts

Uh oh!

crasbe left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

baptleduc commented Jul 22, 2025

Uh oh!

baptleduc commented Jul 17, 2025 •

edited

Loading

benpicco Jul 18, 2025 •

edited

Loading

benpicco Jul 18, 2025 •

edited

Loading

riot-ci commented Jul 17, 2025 •

edited

Loading

baptleduc commented Jul 30, 2025 •

edited

Loading

baptleduc commented Aug 26, 2025 •

edited

Loading

crasbe commented Sep 4, 2025 •

edited

Loading