NRF5340 low-power measurement with PPK2

RE: NRF5340 low-power measurement with PPK2

Simonr — Tue, 13 Sep 2022 09:19:56 GMT

Thanks!

My initial guess is that the solder bridge(s) (SB40 and SB41) are cut, and not shorted to resume normal DK operation. If that's not it, and all the switches are in default mode, please try enabling Verbose logging and upload that log.

Best regards,

Simon

RE: NRF5340 low-power measurement with PPK2

Bas van den Berg — Tue, 13 Sep 2022 04:38:51 GMT

I can only Reply to this message (not the last for some reason).

I have measured board 1 in two setups:

PPK as power supply.

Left picture: USB of PPK connected to PC directly

Right picture: USB of PPK connected to Monitor (=USB hub)

Otherwise exactly the same. Note that no USB is connected to the DK (only the 3 power wires from the PPK)

So the conclusion is that the USB connection to the PPK causes this noise somehow...

RE: NRF5340 low-power measurement with PPK2

Jester — Mon, 12 Sep 2022 15:44:34 GMT

BTW. I am now trying one of our newer DK boards.

Al these boards are of 2.0.0 type.
Together with these boards i get :

2022-09-12T15:41:06.586Z DEBUG Application data folder: C:\Users\0000\AppData\Roaming\nrfconnect\pc-nrfconnect-programmer
2022-09-12T15:41:06.662Z INFO Using nrf-device-lib-js version: 0.4.13
2022-09-12T15:41:06.662Z INFO Using nrf-device-lib version: 0.12.8
2022-09-12T15:41:06.662Z INFO Using nrfjprog DLL version: 10.16.0
2022-09-12T15:41:06.663Z INFO Using JLink version: JLink_V7.66a
2022-09-12T15:41:06.700Z DEBUG App pc-nrfconnect-programmer v3.0.4 official
2022-09-12T15:41:06.700Z DEBUG App path: C:\Users\0000\.nrfconnect-apps\node_modules\pc-nrfconnect-programmer
2022-09-12T15:41:06.700Z DEBUG nRFConnect 3.12.0, required by the app is (^3.12.0)
2022-09-12T15:41:06.701Z DEBUG nRFConnect path: C:\Users\0000\AppData\Local\Programs\nrfconnect\resources\app.asar
2022-09-12T15:41:06.701Z DEBUG HomeDir: C:\Users\0000
2022-09-12T15:41:06.701Z DEBUG TmpDir: C:\Users\0000~1\AppData\Local\Temp
2022-09-12T15:41:10.566Z ERROR Unsupported device.
The detected device could not be recognized as
neither JLink device nor Nordic USB device.

0000 is replacement of real user/company name.

Any ideas ?

RE: NRF5340 low-power measurement with PPK2

Jester — Mon, 12 Sep 2022 14:36:44 GMT

RE: NRF5340 low-power measurement with PPK2

Simonr — Mon, 12 Sep 2022 13:44:13 GMT

Hi again,

Can you upload a screenshot of the noisy test samples, as I tried again on another DK and PPK combo and I still see the same "tall and clean" toggling on my end. If you'd like I can set up a lower power sample as well that puts the device to a proper low power mode when the LED is off for example.

Best regards,

Simon

RE: NRF5340 low-power measurement with PPK2

Simonr — Mon, 12 Sep 2022 11:24:29 GMT

Thank you for the clarification:

Here is a .hex file of the Blinky sample in Zephyr, should toggle up and down every second.

devzone.nordicsemi.com/.../0333.zephyr.hex

Should look something like this power profiler trace. Tested with two different PPK2s on our end.

Best regards,

Simon

RE: NRF5340 low-power measurement with PPK2

Bas van den Berg — Mon, 12 Sep 2022 10:14:15 GMT

My PPK serial is:

DF222A76BBE4

RE: NRF5340 low-power measurement with PPK2

Bas van den Berg — Mon, 12 Sep 2022 09:33:28 GMT

We have 2x PPK and 4x NRF5340 DK

RE: NRF5340 low-power measurement with PPK2

Jester — Mon, 12 Sep 2022 08:44:47 GMT

Can you send a modified version of your software so that you make one of the leds blink in lets say 10Hz.
Then at least, when we program your code, we know its running.

RE: NRF5340 low-power measurement with PPK2

Simonr — Mon, 12 Sep 2022 08:03:48 GMT

So there are three PPK II devices that are not working correctly now, since there are three separate serial numbers here?

Best regards,

Simon

RE: NRF5340 low-power measurement with PPK2

Jester — Mon, 12 Sep 2022 07:05:59 GMT

At least one PPK is working correct.

3V/3301 ohm should read 908,815510451 uA
3V/328.850 ohm should read 9,122700319 uA

So that seems to be OK at least for the source meter part although i would not want to call it accurate.
But let's ignore the accuracy for now. We more or less have a decent measurement with values that can be explained.

RE: NRF5340 low-power measurement with PPK2

Bas van den Berg — Sat, 10 Sep 2022 06:40:08 GMT

The serial numbers I have here are:

2020.51 960191560

2020.51 960190044

RE: NRF5340 low-power measurement with PPK2

Jester — Fri, 09 Sep 2022 13:30:47 GMT

My serial sticker reads : PCA63100, 1.0.1, 2022.13, ECESD384; Is that what you're looking for ?

The power profiler v3.4.4 talks about serial : ECE5D3B48A27

The other one is at Bas' place.

The proposed time, 14:00 - 15:30, we will be there. Already discussed in the company mail.

I will see what resistors I have here and try to validate the device itself.

RE: NRF5340 low-power measurement with PPK2

Simonr — Fri, 09 Sep 2022 11:30:10 GMT

We can do a meeting if you'd like, but that would have to wait until Monday (I will be free from 10:30 to 11:30 and 14:00 to 15:30). Unfortunately it seems like you have two defect PPK2's though, if the behavior is like this on both PPKs, where all the maximum spikes are exactly the same (713.24µA on the one you sent to us).

If you have a resistor available (3kOhm or larger) it would be helpful to see what you measure over a known load in Source Meter mode with 3V applied. 3V on a 3kOhm should show a 1mA current consumption. It would also help if you have a resistor large enough to get under 10µA available (so larger than 300kOhm).

Finally, if we could get the serial numbers of the PPKs you have we can check the test logs of them as well.

Best regards,

Simon

RE: NRF5340 low-power measurement with PPK2

Jester — Fri, 09 Sep 2022 09:43:44 GMT

And... you rule out the "retained registers" as problem ?

RE: NRF5340 low-power measurement with PPK2

Jester — Fri, 09 Sep 2022 09:43:00 GMT

To double check we verified the flash for net core, which is empty.
We erased again. Then programmed the hex file you attached above.
Verification with a read in nrfCOnnect indeed shows content in the app core.

Again. Can we schedule a call ? Then you can see what's going on.

RE: NRF5340 low-power measurement with PPK2

Simonr — Fri, 09 Sep 2022 09:23:44 GMT

The current consumption you refer to here is the HF crystal ramp up of an nRF9160 DK I believe, and should definitely not be seen if you only have the empty app core running on an nRF5340 Dk on your end. Can you use the nRF Programmer to check what is on the APP and NET core of your DK? Both have the bare minimum on my end which should result in what I showed below.

RE: NRF5340 low-power measurement with PPK2

Jester — Fri, 09 Sep 2022 09:10:26 GMT

Hi Simon,

I'm a collegue of Bas. Can we schedule a telco to talk about this ?
This is taking a lot of time doing it this way.

We have 4 DK boards, they can't be all broken.
The USB cable is ok of course, because the board is on, and we're able to program it.

If you look at this video:

https://www.youtube.com/watch?v=B42lPvkUSoc&t=228s

we found that here :

www.nordicsemi.com/.../GetStarted

The current we're stuck at is the current you see in that video on 3 min 45 sec.
With our setups it stays on that level, even with your code.

Could it be that we are having problems with the "retained registers" ?
Things that might not be initialised on your code, and stuck on a value we have tried in our code ?

RE: NRF5340 low-power measurement with PPK2

Simonr — Fri, 09 Sep 2022 08:44:05 GMT

Hi again

Sounds very strange that you have two malfunctioning PPKs.

SW6 is set to default, SW10 is OFF, Power source is set to VDD. The DK is powered with the USB cable to J2.

Vout on the PPK is connected to the top pin on P22 (nRF current measurement), and Vin on the bottom one. Either GND pin is connected to (-) on P21 (External supply). We even have the same revision of the nRF53 DK, this is very strange. Can you double check that the USB cable is working correctly. Do you have another nRF53 DK you can try with both PPKs as well to check if the consumption looks the same there as well. Also double check that all connections and that the cables from PPK to DK works correctly.

Best regards,

Simon

RE: NRF5340 low-power measurement with PPK2

Bas van den Berg — Thu, 08 Sep 2022 10:04:10 GMT

The programming of the hex-file in the Programmer (along with erase-all) went fine. But the result is the same. I tried another PPK2 (I have 2), same.

Do you have your other settings, as they are just as important. How do you supply the board, SW6, SW10 positions, etc

SW6 - nRF_ONLY / Default

SW10 - On/Off

Power source - Li-Po / VDD / USB

Did you connect the USB cable from the DK to your PC?

RE: NRF5340 low-power measurement with PPK2

Simonr — Thu, 08 Sep 2022 09:05:42 GMT

Okay, this sounds like the PPK2 is not working properly then. I have tested the empty APP core sample that should provide very low power as it just keeps the system OFF. If your PPK2 is working correctly you should see a power profiler trace similar to this one (this was using Ampere meter mode):

You can use the nRF Connect Programmer app to first do an "erase all", then flash the empty_app_core onto the nRF5340 DK. Please try and see if you get a similar current consumption or not.

devzone.nordicsemi.com/.../2821.zephyr.hex

Best regards,

Simon

RE: NRF5340 low-power measurement with PPK2

Bas van den Berg — Wed, 07 Sep 2022 07:44:05 GMT

This is what the PPK gives me... 😞

I somehow cannot get the Source mode to work, because I keep measuring 6.5 mA then (the iMCU I suspect) even when trying different switch modes. The Measurement above was done using the following settings:

SW10 ON
SW6 DEFAULT
Both boards connected to PC over USB

The software does nothing except use RTC0 to give a 100 Hz tick, with the main loop doing only WFI().

Do you have a binary image I can use, along with the switch+Solder-Bridge settings? We should be able to get to the same result then..

RE: NRF5340 low-power measurement with PPK2

Simonr — Tue, 06 Sep 2022 13:23:53 GMT

Is the .ppk file the same for both Source and Ampere meter mode, because the current consumption seems very strange. Does it look like the snippet below on your end as well or did something go wrong during export/import? No current consumption should actually look like this. What is your application supposed to do, and does it work as expected except for the current consumption?

Best regards,

Simon

RE: NRF5340 low-power measurement with PPK2

Bas van den Berg — Mon, 05 Sep 2022 13:10:42 GMT

I did follow the guide, although since we're not using Zephyr, only the bottom half was useful. With everything I found the current consumption is around 250 uA (+/- 10 uA). Since I don't have the net-core running for this measurement, I cannot turn the VREGRADIO.VREGH off. I don't know if this runs if the net-core is off...

Please see the power measurement attached.devzone.nordicsemi.com/.../low_5F00_power_5F00_250_5F00_uamp.zip

RE: NRF5340 low-power measurement with PPK2

Simonr — Mon, 05 Sep 2022 11:43:27 GMT

Okay, that looks correct. Can you export the current consumption trace from the power Profiler and upload it here so I can take a look? Please note that .ppk files are not okay to upload as is in DevZone so you need to zip the file. 640µA is way too high and points to a peripheral for example being left running for some reason. Please also double check that you followed the "power optimizing guide" correctly.

Best regards,

Simon