Totally contradicting results measuring a Low Power DCDC design

Our OTIIs (we’ve got a dozen with latest firmware and Windows GUI) are showing strange results when I try to measure power consumption of a wireless, battery powered sensor. The sensor is using a DCDC converter from Maxim which internally works with 1 MHz and above. External battery voltage is 3 VDC, a 10uF CAP is connected in parallel to the Battery before the DCDC, the internal VCC is 1.8VDC. The firmware runs multiple higher current actions (~5…60mA) followed by a deep sleep phase which takes 20 seconds. Then everything starts again.

When I measure the sensor directly with OTIIs main power terminals I see about 10 uA average in the deep sleep phases and an average of 330 uA over a full interval of all actions of a 20 seconds interval. When I add a 4700 uF cap to the battery terminals of the sensor then OTII measures still 10 uA sleep current but 100 uA average instead of 330 uA over an 20 seconds interval. The sensor is sourced from OTII. The images illustrate the issue.

One would think that the consumed energy shall stay the same either with or without an additional 4700uF cap in parallel. Whats going on here? Thanks a lot!

Schematic of power supply via DCDC converter:
2020-02-25 10_32_17-P377_ActiveTag_Default_Schema_PCB_V8.1 - PDF-XChange Viewer

330 uA over 9 iterations of 20 seconds each (no external Cap connected):

330 uA over 10 iterations of 20 seconds each (Zoom, no external Cap connected):

100 uA over 9 iterations of 20 seconds each (4700uF external Cap connected):

100 uA over 9 iterations of 20 seconds each (4700uF external Cap and series resistor of 10 Ohms connected):

100 uA over 1 iteration of 20 seconds each (4700uF external Cap and series resistor of 10 Ohms connected):

10 uA deep sleep:

Hi and welcome to the forum!

I am interested in getting the Otii project files to analyze this further.
Could you create a case in our case system and upload the files there?

Best regards,
Björn

Hi Bjorn.

Thanks and sorry, our fault, the devices firmware is not exactly doing every iteration the same. The developers have implemented some magic… So this topic can be closed/deleted.
Many thanks.