Huge difference on the results on measurements of different PPK firmwares.

[Cássio Miranda Coelho Barbosa]

Hello,

We are about to release a Firmware update for our vibration sensors and we are now testing the consumption of the new firmware.

Our sensor is a nRF51822 + Accelerometer + CR2477.
On the previous firmware version we were using TWI to communicate with the Accelerometer.
O the new version we changed to SPI for a faster buffer reading and therefore, lower consumption (at least in theory).

Our development is currently happening with all the engineers working at home, so everybody has the necessary hardware to develop (on this case nRF52DK + PPK I).
The problem is that we are getting very different results with the same DKs, PPKs, hardware and Firmware. The only difference between both setups is the "nRFconnect for desktop" and "power profiler".

We are getting a HUGE difference between both setups. Which rises a few big questions/concerns:

1. Are the new software versions more or less reliable?
2. Should we trust on either of them, even with such a big difference between updates?
3. If we are terribly wrong about our interpretation of facts, can you help us with a solution for the awkward results we are getting?
4. Can we roll back to previous versions of the softwares?

The following images shows the difference of the readings on both setups with the same HW and FW.

Thanks a lot!

[Øyvind]

Hello, 

Due to Easter holiday we are currently lower staffed than usual, and some extra time must be expected to provide an answer. We apologize for the inconvenience this may cause.  

Are you using nRF51 or nRF52? 

Can you please provide more information on the test setup you are using?  What kind of USB cables are you using? Are you connecting through a USB hub?

Thank you.

Kind regards,
Øyvind

[Cássio Miranda Coelho Barbosa]

Hello Øyvind,

Our sensor is a nRF51822 + Accelerometer(KX122) + CR2477.

We are using a regular 15cm micro USB cable. No hubs.

The only difference is the version of the power profiler Kit.

Over the weekend we continued to run some tests trying to eliminate variables and the results are still the same.

Thanks for helping us!

[Øyvind]

I've talked to our SW team to check if there are any changes in the Power Profiler app. They ask you to use "Shift + Left click and drag" to mark the area of interest. Then click on "Save / Export" and save it as a .ppk file. Please do this for both measurements. 

Let me know if anything is unclear. 

Kind regards,
Øyvind

[Øyvind]

Also, can you provide a screenshot of the whole Power Profile window to include all settings that you are using? In v3.1.1 are you enabling averaging? Please disable this and do a measurement with highest possible frequency in v3.1.1. 

In v3.0.3 the averaging button is not available, and this may be the root cause for the difference. To enable "Save / Export" in v3.0.3, you can use CTRL+ALT+SHIFT+L.

Unfortunately, downgrading (roll back) the Power Profiler app is not a straight forward task. Will get back to you on that one.

-Øyvind