Nordic Power Profiler Kit II, inaccurate charge (integral)

[T-moe]

I'm using the nordic ppk II to measure the power consumption of my device.

In order to validate the charge (current integral) displayed by the PPK2 i put two PPK2 daisy-chained into my current path. Both are configured as ampere meter and started at the same time. 
Even after a short while the two devices report different Charge. e.g. after 13 minutes one reports 24 Coulomb and the other 26 Coulomb.  Why is that? Can I somehow avoid this drift?

[Simonr]

Hi

Why exactly are you connecting the PPK2's like this, and we'd not recommend connecting them like this. I don't think that will make sense since they are not actual Ampere meters, but works much like one. How much power are the two PPKs seeing being drawn and what do the application you're measuring doing exactly?

I'm guessing that the second PPK (in the chain) is the one showing the highest draw, and that the second one shows the higher number because it also sees some power being drawn from the prior PPK2.

Best regards,

Simon

[T-moe]

Hi Simon, 

Thank you for your response.

We're building an embedded IoT product and want to validate the battery consumption of the device.

The device consumes between 40 uA (in deepsleep) and 120 mA  ( when communicating) with spikes even higher (1 Ampere).

We thought that we could leave our device running for 24 hours, under predefined operating conditions, and measure the charge it takes using PPK2. 

Is this not a use-case the PPK2 can be used for? (why not?)

The reason why we chain two PPK2's is the following:

We want to measure measure the consumption automatically and use a python library to control the PPK2. In order to validate this setup, we put another PPK2 in the chain, to validate the measured charge. The measurements were off by 17%. But this could be the fault of the unofficial python library. To make sure our setup is not flawed to begin with, I then chained two PPK2's both using the original GUI and ended up with the difference stated in my initial question...

If I understand you right, you say that the PPK2 will leak some current, so that the 2nd PPK2 will measure a higher charge... Shouldnt the PPK just measure the voltage over a shunt resistor? This shouldn't introduce additional current, right? At least not one changing the result by 10%? (26 instead of 24C)

Thank you for your help!!

[Simonr]

Hi again

Thank you for the explanation. So, the reason you shouldn't connect two PPK2s in series like this, is mainly that the range switching of one of the PPKs can cause the other to switch in the wrong place and vice versa. They can switch on different places because they are differently calibrated. To confirm the setup you should do so one by one, with the given accuracy values for each kit. You're in danger of getting extra resistance in your path, and since both PPKs are calibrated down to 0R5, it will mess up the measurements for one of the kits to get extra series resistance in their path.

I double checked with one of the PPK devs, and leakage from one PPK to the other should not be a concern. Sorry about the confusion. But the above could still be an issue.

Best regards,

Simon

[T-moe]

Ok. That makes sense. I'll validate by measuring one by one.