low cost power shield solution #95056

hakehuang · 2025-08-27T11:24:06Z

Create adc_power_shield sample for any board with adc can be used for power measurement.
Add general_adc_platform power shiled, which can be used for power measurement with the sample application.
the power data is stored in the power_shield folder for further analysis

hakehuang · 2025-08-27T11:25:03Z

log for mimxrt595_evk board

((.venv) ) ubuntu@ubuntu-OptiPlex-7050:/home/shared/disk/zephyr_project/zephyr_test/zephyr$ scripts/twister --device-testing --hardware-map /home/ubuntu/nxp/mimxrt595_evk_cm33/map.yaml -T samples/boards/nxp/mimxrt595_evk/system_off
ZEPHYR_BASE unset, using "/home/shared/disk/zephyr_project/zephyr_test/zephyr"
Renaming previous output directory to /home/shared/disk/zephyr_project/zephyr_test/zephyr/twister-out.1
INFO    - Using Ninja..
INFO    - Zephyr version: v4.2.0-2411-g0f6f916cbaa4
INFO    - Using 'zephyr' toolchain.
INFO    - Building initial testsuite list...
INFO    - Writing JSON report /home/shared/disk/zephyr_project/zephyr_test/zephyr/twister-out/testplan.json

Device testing on:

| Platform                      |        ID | Serial device                                         |
|-------------------------------|-----------|-------------------------------------------------------|
| mimxrt595_evk/mimxrt595s/cm33 | 727562924 | /dev/serial/by-id/usb-SEGGER_J-Link_000727562924-if00 |

INFO    - JOBS: 8
INFO    - Adding tasks to the queue...
INFO    - Added initial list of jobs to queue
INFO    - Total complete:    1/   1  100%  built (not run):    0, filtered:    0, failed:    0, error:    0
INFO    - 1 test scenarios (1 configurations) selected, 0 configurations filtered (0 by static filter, 0 at runtime).
INFO    - 1 of 1 executed test configurations passed (100.00%), 0 built (not run), 0 failed, 0 errored, with no warnings in 29.99 seconds.
INFO    - 1 of 1 executed test cases passed (100.00%) on 1 out of total 1154 platforms (0.09%).
INFO    - 1 test configurations executed on platforms, 0 test configurations were only built.

JarmouniA · 2025-08-27T11:52:20Z

samples/drivers/adc/adc_power_shield/sample.yaml

Please take a look at #94585, and only keep 1 platform per vendor.

thanks @JarmouniA PR #94585 looks good, I will update according after your PR merged

There is no need to wait for it to be merged, just apply the same filtering and only keep overlay/conf of 1 platform per vendor in the new sample.

nordicjm · 2025-08-27T12:49:21Z

samples/drivers/adc/adc_power_shield/Kconfig

+
+config SEQUENCE_32BITS_REGISTERS
+	bool "ADC data sequences are on 32bits"
+	default n


Suggested change

default n

samples/drivers/adc/adc_power_shield/boards/arduino_due.overlay

samples/drivers/adc/adc_power_shield/boards/frdm_mcxc444.overlay

nordicjm · 2025-08-27T12:50:35Z

samples/drivers/adc/adc_power_shield/boards/s32k148_evb.overlay

+	pinctrl-0 = <&adc0_default>;
+	pinctrl-names = "default";


move above child node

samples/drivers/adc/adc_power_shield/boards/s32z2xxdc2_s32z270_rtu0.overlay

dsseng

Thanks, this is a really great addition

dsseng · 2025-08-29T17:04:09Z

scripts/pylib/power-twister-harness/requirements-dev.txt

+# Additional development tools
+ipython>=8.0.0
+jupyter>=1.0.0


Are these actually used here?

right, it is added by my local setting, will remove it

dsseng · 2025-08-29T17:05:29Z

scripts/pylib/power-twister-harness/utils/UartSettings.py

+# Set UTF-8 encoding for stdout/stderr
+os.environ['PYTHONIOENCODING'] = 'utf-8'
+
+# For Python 3.7+, you can also reconfigure stdout/stderr
+if hasattr(sys.stdout, 'reconfigure'):
+    sys.stdout.reconfigure(encoding='utf-8')
+    sys.stderr.reconfigure(encoding='utf-8')


Move this under def main or __name__ == "__main__" so that the module is cleanly importable?

hakehuang · 2025-09-03T03:34:24Z

@nordicjm, @dsseng I have updated according to your comments please help to review. Thanks

DTS format fixed

hakehuang · 2025-09-04T14:05:23Z

I define a dft (device for test) which can also extend to multi-application build

kartben · 2025-09-04T14:19:16Z

samples/drivers/adc/adc_power_shield/README.rst

+   :name: Analog-to-Digital Converter (ADC) power shield sample
+   :relevant-api: adc_interface
+
+   Read analog inputs from ADC channels, using a sequence.
+
+Overview
+********
+
+This sample is to enable general power shield for platfroms that have ADC support.
+
+
+Building and Running
+********************
+
+Make sure that the ADC is enabled (``status = "okay";``) and has each channel as a
+child node, with your desired settings like gain, reference, or acquisition time and
+oversampling setting (if used). It is also needed to provide an alias ``adc0`` for the
+desired adc. See :zephyr_file:`boards/nrf52840dk_nrf52840.overlay
+<samples/drivers/adc/adc_dt/boards/nrf52840dk_nrf52840.overlay>` for an example of
+such setup.


All of this is very stale and copied from the original sample (related: do you really need to duplicate the 100s of overlays too??). I have to admit I don't really understand what this is trying to accomplish and the fact that the README hasn't been updated is not really helping.
Just a thought but, if you are calling this a "shield" then maybe you want to create an actual shield (or snippet?), and not a sample?

@kartben, the sample application is modified from adc_sequence sample. The idea here is to create a general power shield, which means any adc featured zephyr platform can be used as a power shield. so we can take any of those 100s boards and flashed with this sample, you can use them as general_power_shield fixture for measuring power, see the https://github.com/zephyrproject-rtos/zephyr/blob/af6248f8b7509f505982b8fe5d6e9a6efd8bc0f5/scripts/pylib/power-twister-harness/README.rst this is a full picture for the usage.
the sample readme is just describe how the sample runs.

I want to provide a platform agonic solution, such as any same feature Zephyr platform can function as the same.
the is a demo setup for general power measurement.

This comment thread still applies IMO. I am sorry if I am missing something but simply copying an existing code sample (and its dozens of overlays!) and then calling this a "shield" while shields are something completely different in Zephyr is really incredibly confusing.
Please don't ask for re-review when comments havne't been addressed - the README being an exact copy-paste of the original sample is also pretty problematic. Thanks!

hakehuang · 2025-09-26T02:48:03Z

@nashif please take a look. Thanks

enable power shield testing on mimxrt595_evk ``` scripts/twister --device-testing --device-serial /dev/ttyACM0 -T samples/boards/nxp/mimxrt595_evk/system_off -X pm_probe:/dev/ttyACM1,115200 ``` an power_shield folder will be created in the build path whith measured voltage/current/power Signed-off-by: Hake Huang <[email protected]>

add power_shield sample for general_adc_platform power shiled, to record power/voltage/current Signed-off-by: Hake Huang <[email protected]>

1. general power shield work with the adc_power_shield sample 2. config samples to config probe 3. add readme Signed-off-by: Hake Huang <[email protected]>

enable power measure on power_mgmt_soc for mimxrt1060_evk@C Signed-off-by: Hake Huang <[email protected]>

sonarqubecloud · 2025-09-30T05:07:36Z

Quality Gate passed

Issues
9 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
2.4% Duplication on New Code

See analysis details on SonarQube Cloud

bjarki-andreasen

Has it been considered to just use the ADC shell and a "blank" application? Setting up and performing a reading can be done directly using the shell, no custom parsing of console input/output should be needed.

hakehuang · 2025-10-01T08:16:21Z

Has it been considered to just use the ADC shell and a "blank" application? Setting up and performing a reading can be done directly using the shell, no custom parsing of console input/output should be needed.

the main purpose of this harness is for CI. As we want to a unified test approach, if in a blank application, you will have to customize it by you own, in this case, you are not using CI. for example, you will not know which ADC channel is connect to which power rails, you have parse them according how you connect your ADC probes, and then you have to write your own script to do so. This is what I try to avoid. Please take a look at https://github.com/zephyrproject-rtos/zephyr/pull/95056/files#diff-9e48191a8fd4eac8e1fb6d4f1c73c35c3cdc65720c3eb6264964eb104e612534
you will need a config parser to link the shell output with your real probe settings, this is what I am trying to unify. @bjarki-andreasen, please be aware this is mainly for CI purpose.

bjarki-andreasen · 2025-10-01T09:05:07Z

Has it been considered to just use the ADC shell and a "blank" application? Setting up and performing a reading can be done directly using the shell, no custom parsing of console input/output should be needed.

the main purpose of this harness is for CI. As we want to a unified test approach, if in a blank application, you will have to customize it by you own, in this case, you are not using CI. for example, you will not know which ADC channel is connect to which power rails, you have parse them according how you connect your ADC probes, and then you have to write your own script to do so. This is what I try to avoid. Please take a look at https://github.com/zephyrproject-rtos/zephyr/pull/95056/files#diff-9e48191a8fd4eac8e1fb6d4f1c73c35c3cdc65720c3eb6264964eb104e612534 you will need a config parser to link the shell output with your real probe settings, this is what I am trying to unify. @bjarki-andreasen, please be aware this is mainly for CI purpose.

We already have a twister harness for the shell, can it not be reused? and regarding the specific ADC channel, that can be added in an overlay when building surely, like this PR is already doing? We can also add a custom shell command to use specifically if needed, no need to implement command line parsing in the sample

hakehuang · 2025-10-03T02:19:37Z

We already have a twister harness for the shell, can it not be reused? and regarding the specific ADC channel, that can be added in an overlay when building surely, like this PR is already doing? We can also add a custom shell command to use specifically if needed, no need to implement command line parsing in the sample

@bjarki-andreasen if you look at my PR, I add a 'dft' which reuse the whole pytest serial handler see

zephyr/scripts/pylib/power-twister-harness/conftest.py

Line 130 in 815f2b6

    
           def dft(request: pytest.FixtureRequest, dft_object) -> Generator[DeviceAdapter, None, None]:

. ADC channel is already controlled by overlay, but the problem is that we need map the channel to the target power rails see below definition:
the channels_p and channel_n to the power rail "VDD_CORE" / "VDD_IO" is by your board setting definition, this could be different pre your test harness setting, and this has to be customized, here is use a yaml to control it.

device_id: "adc_power_monitor_01"
routes:
  - id: 0
    name: "VDD_CORE"
    shunt_resistor: 0.1  # ohms
    gain: 1.0
    type: "single"
    channels:
      channels_p: 4
      channels_n: 0
  - id: 1
    name: "VDD_IO"
    shunt_resistor: 0.1  # ohms
    gain: 1.0
    type: "single"
    channels:
      channels_p: 7
      channels_n: 3
calibration:
  offset: 0.0
  scale: 1.0

so nowhat where to add, we either add shell command or use a dedicated power shield sample application, we have to parse the shell/command output to get the board settings on its probe capacities, see below.

{
  "channel_count": 2,
  "resolution": 12,
  "channels": {
    "0": {
      "mode": "single",
      "verf_mv": 3300
    },
    "3": {
      "mode": "single",
      "verf_mv": 3300
    },
    "4": {
      "mode": "single",
      "verf_mv": 3300
    },
    "7": {
      "mode": "single",
      "verf_mv": 3300
    }
  }
}

and we need map the probe capacities to the target measure power rails as I mentioned above. My understanding is that you propose to use a series of shell command to replace the power measure sample which will be running on power shield. this will add adc shell linked with power calculations I am not sure this will accepted by ADC shell owner? As in this PR, we have an independent sample to host everything, move those to shell command needs approval, can you help to confirm on my understanding and if accepted by ADC shell group I can move the implement to shell.

bjarki-andreasen · 2025-10-04T06:41:48Z

We already have a twister harness for the shell, can it not be reused? and regarding the specific ADC channel, that can be added in an overlay when building surely, like this PR is already doing? We can also add a custom shell command to use specifically if needed, no need to implement command line parsing in the sample

@bjarki-andreasen if you look at my PR, I add a 'dft' which reuse the whole pytest serial handler see

zephyr/scripts/pylib/power-twister-harness/conftest.py

Line 130 in 815f2b6

def dft(request: pytest.FixtureRequest, dft_object) -> Generator[DeviceAdapter, None, None]:

. ADC channel is already controlled by overlay, but the problem is that we need map the channel to the target power rails see below definition:
the channels_p and channel_n to the power rail "VDD_CORE" / "VDD_IO" is by your board setting definition, this could be different pre your test harness setting, and this has to be customized, here is use a yaml to control it.
device_id: "adc_power_monitor_01"
routes:
  - id: 0
    name: "VDD_CORE"
    shunt_resistor: 0.1  # ohms
    gain: 1.0
    type: "single"
    channels:
      channels_p: 4
      channels_n: 0
  - id: 1
    name: "VDD_IO"
    shunt_resistor: 0.1  # ohms
    gain: 1.0
    type: "single"
    channels:
      channels_p: 7
      channels_n: 3
calibration:
  offset: 0.0
  scale: 1.0
so nowhat where to add, we either add shell command or use a dedicated power shield sample application, we have to parse the shell/command output to get the board settings on its probe capacities, see below.
{
  "channel_count": 2,
  "resolution": 12,
  "channels": {
    "0": {
      "mode": "single",
      "verf_mv": 3300
    },
    "3": {
      "mode": "single",
      "verf_mv": 3300
    },
    "4": {
      "mode": "single",
      "verf_mv": 3300
    },
    "7": {
      "mode": "single",
      "verf_mv": 3300
    }
  }
}
and we need map the probe capacities to the target measure power rails as I mentioned above. My understanding is that you propose to use a series of shell command to replace the power measure sample which will be running on power shield. this will add adc shell linked with power calculations I am not sure this will accepted by ADC shell owner? As in this PR, we have an independent sample to host everything, move those to shell command needs approval, can you help to confirm on my understanding and if accepted by ADC shell group I can move the implement to shell.

No changes to the adc shell is needed. We would just create a set of custom shell commands as part of the power shield sample. The issue is not with having a sample at all, we need that, but with how the serial communication is implemented. We can freely create sample specific shell commands, so why not use that?

hakehuang · 2025-10-05T01:57:27Z

No changes to the adc shell is needed. We would just create a set of custom shell commands as part of the power shield sample. The issue is not with having a sample at all, we need that, but with how the serial communication is implemented. We can freely create sample specific shell commands, so why not use that?

ok, I see, I will update this PR to add a sample specific shell command. but I have some concerns on the code size, I enable the stm32f030dlk as the low cost shield, which code size is very critical. Let me try and feedback.

bjarki-andreasen · 2025-10-06T06:59:53Z

No changes to the adc shell is needed. We would just create a set of custom shell commands as part of the power shield sample. The issue is not with having a sample at all, we need that, but with how the serial communication is implemented. We can freely create sample specific shell commands, so why not use that?

ok, I see, I will update this PR to add a sample specific shell command. but I have some concerns on the code size, I enable the stm32f030dlk as the low cost shield, which code size is very critical. Let me try and feedback.

For this, enable SHELL_MINIMAL, the entire shell should "only" take up 6-8K ROM, if that is too much we may have to go with a custom solution

zephyrbot added area: ADC Analog-to-Digital Converter (ADC) platform: NXP NXP area: Twister Twister area: Samples Samples labels Aug 27, 2025

zephyrbot requested review from anangl, butok, decsny, DerekSnell, dleach02, EmilioCBen, gchwier, golowanow, JarmouniA, KamilxPaszkiet, kartben, LukaszMrugala, mmahadevan108, nashif and PerMac August 27, 2025 11:25

zephyrbot assigned kartben Aug 27, 2025

hakehuang force-pushed the lc_power_shield branch 4 times, most recently from 59e3ca2 to 8f0f3f4 Compare August 27, 2025 12:24

JarmouniA reviewed Aug 27, 2025

View reviewed changes

nordicjm previously requested changes Aug 27, 2025

View reviewed changes

hakehuang marked this pull request as draft August 28, 2025 03:32

hakehuang force-pushed the lc_power_shield branch 2 times, most recently from c19dcd6 to 78d25a9 Compare August 28, 2025 05:03

dsseng reviewed Aug 29, 2025

View reviewed changes

hakehuang force-pushed the lc_power_shield branch from b5f1df0 to af6248f Compare August 31, 2025 04:10

hakehuang mentioned this pull request Sep 4, 2025

twister: scripts: sharing of built applications #94167

Merged

kartben requested changes Sep 4, 2025

View reviewed changes

nashif assigned nashif and unassigned kartben Sep 5, 2025

hakehuang requested a review from kartben September 26, 2025 02:47

zephyrbot added area: Power Management platform: NXP MCU area: Tests Issues related to a particular existing or missing test labels Sep 29, 2025

zephyrbot requested review from bjarki-andreasen and ceolin September 29, 2025 16:56

hakehuang added 4 commits September 30, 2025 13:02

samples: adc: add power shield application

ecc136b

add power_shield sample for general_adc_platform power shiled, to record power/voltage/current Signed-off-by: Hake Huang <[email protected]>

tests: scripts: add general power shield support

3b1bf2e

1. general power shield work with the adc_power_shield sample 2. config samples to config probe 3. add readme Signed-off-by: Hake Huang <[email protected]>

tests: pm: power_mgmt_soc: enable measurement on rt1060

815f2b6

enable power measure on power_mgmt_soc for mimxrt1060_evk@C Signed-off-by: Hake Huang <[email protected]>

hakehuang force-pushed the lc_power_shield branch from abd034c to 815f2b6 Compare September 30, 2025 05:03

bjarki-andreasen requested changes Sep 30, 2025

View reviewed changes

hakehuang marked this pull request as draft October 5, 2025 01:57

low cost power shield solution #95056

Are you sure you want to change the base?

low cost power shield solution #95056

Conversation

hakehuang commented Aug 27, 2025

Uh oh!

hakehuang commented Aug 27, 2025

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!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

dsseng left a comment

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!

hakehuang commented Sep 3, 2025

Uh oh!

hakehuang commented Sep 4, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

kartben Sep 4, 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!

hakehuang Sep 5, 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!

hakehuang Sep 5, 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!

hakehuang commented Sep 26, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

sonarqubecloud bot commented Sep 30, 2025

Quality Gate passed

Uh oh!

bjarki-andreasen left a comment

Choose a reason for hiding this comment

Uh oh!

hakehuang commented Oct 1, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

bjarki-andreasen commented Oct 1, 2025

Uh oh!

hakehuang commented Oct 3, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

bjarki-andreasen commented Oct 4, 2025

Uh oh!

hakehuang commented Oct 5, 2025

Uh oh!

bjarki-andreasen commented Oct 6, 2025

Uh oh!

Uh oh!

hakehuang commented Sep 4, 2025 •

edited

Loading

kartben Sep 4, 2025 •

edited

Loading

hakehuang Sep 5, 2025 •

edited

Loading

hakehuang Sep 5, 2025 •

edited

Loading

hakehuang commented Sep 26, 2025 •

edited

Loading

hakehuang commented Oct 1, 2025 •

edited

Loading

hakehuang commented Oct 3, 2025 •

edited

Loading