Power Profiling Variance on BLE Transmission using PPK2

[jake11212]

Hello,

I'm using the PPK2 to evaluate a DUT which is transmitting over BLE using NUS in SDK 17.0.2.  I'm seeing very large variability over each transmission and was wondering if there were any ways to optimize this.  Here is one graph for example showing 5 transmissions of the same packet.  The middle peak average was 0.33uA while the 5th peak was 0.13uA.  While these values are close the percent difference is huge and I was wondering what might cause this.

Also just as a sanity check when the device is powered off the PPK2 reads 0.11uA.  Is this a typical leak current for an nRF52840?

I also had even greater variation when power profiling the advertising routine.  Each peak represents 10 seconds of advertising.  The middle peak was 3.2uA average whereas the first peak was only 0.18 uA average.  Is there a way to optimize this, or an explanation of what may be happening to cause such huge differences in an advertising routine which should be similar between trials?

[Vidar Berg]

Hello,

There is something wrong with the measurements. The peak currents should have been well above 5 mA for if you were measuring a BLE advertisement event. For comparison, the screenshot below is of an adv. event I measured here on my desk with a nRF52840DK +PPK2 in Source Meter mode.

This also matches the profile shown in the  Online Power Profiler for Bluetooth LE  if I input the same parameters. 

Could you maybe describe your measurement setup in a bit more detail? For instance, if you are using the PPK in Source Meter mode or Ampere mode, and if your DUT is a DK or custom board, etc.

Best regards,

Vidar

[jake11212]

I'm using the PPK2 in Ampere mode.  The DUT is a custom board.  I connected the Vdd of the custom board to the Vin/Vout circuit of the PPK2 and gnd on the custom board to ground of the PPK2.  

[Vidar Berg]

OK, thanks for confirming. It sounds like you have set up everything correctly, but the measurements results are wrong for some reason.

Could you please try to use it in source mode instead to see if you can get some more meaningful results?

[jake11212]

I changed to source mode and this was the behavior which is more in line with what you saw.  How did you change to the line graph?  

Also in my setup I have the same point of Vdd connected to both vin and vout on the ppk2.  Do I need two separate points of Vdd for Vin and Vout?  I didn't think it would make a difference but I wasn't sure.

Is there a reason why source mode would cause the results to change so much?  The peaks that I obtained earlier did line up with when the device began advertising confirmed on the nRF toolbox app, so it is definitely reading values from the DUT.

A last question I have is about the enable power output toggle.  What does it do when the device is in ampere meter mode?  And if it is toggled off during source mode does that cause source mode to act as ampere meter mode?

[Vidar Berg]

jake11212 said:

I changed to source mode and this was the behavior which is more in line with what you saw.  How did you change to the line graph?  

You should see the same if you zoom more in on the waveform. The length of an Advertising event is about 5 ms.

jake11212 said:

Also in my setup I have the same point of Vdd connected to both vin and vout on the ppk2.  Do I need two separate points of Vdd for Vin and Vout?  I didn't think it would make a difference but I wasn't sure.

OK, this explains the strange measurements. The ammeter must be connected in series with the PPK. Otherwise, the current will just flow past the ammeter inside the PPK2.

How to connect VDD in series with the PPK ammeter:

jake11212 said:

A last question I have is about the enable power output toggle.  What does it do when the device is in ampere meter mode?  And if it is toggled off during source mode does that cause source mode to act as ampere meter mode?

As illustrated by the silkscreen on the backside of the PPK2, the power source is not connected in ampere mode. In other words, toggling the power button will have not have any effect in this mode.