uCurrent Gold range hack + Results + Glitches

[EarthLord]

Hey everyone,
Here's a small hack of uCurrent Gold for the purpose of measuring current of recent MCUs with radios. Thanks to our dear Dave 'Lee' Jones for this wonderful open source instrument.

So I changed R1 resistor from 0.01 Ohm to 1 Ohm.  (Sorry for the too much solder in the blob) You can see the 0.01 Ohm resistor stuck to one of the pins to the switch. Now the conversion should be almost 100 mV/mA. Also a USB cable is attached to the battery connector so that a USB battery bank can be used for power to get higher range for the measurement.


Now for the actual use case. I've connected uCurrent on the high side to measure the power of a nrf51 chip on PCA10028 board when sending out periodic BLE advertisement packets. So in this case the current drawn varies from a couple of uA when sleeping to a maximum of 13-14 mA when both the radio and processor is ON. For this case the uCurrent should give less than a mV output when sleeping and less than the maximum range when fully powered up with this new mode of 100mV/mA.
The uCurrent is powered by a battery bank. The nrf51 DK is powered by a laptop not connected to the charger. And the output of uCurrent is seen on a Hantek DSO5072P scope (1GSa/s, 70 MHz bandwidth).


Here's the screenshot of the beautiful waveform of current consumed when sending out the advertisement packets in the three advertisement channels. The only blemish in the whole setup are these periodic glitches that you can see in the waveform. Any thoughts on why this could be happening?


And any feedback on this hack? Cheers!

[SeanB]

Glitches are probably the internal converter in the battery bank putting noise on the supply lines.  Try using a 10R series resistor on the USB supply line and 100uF 10V capacitor across the ucurrent supply rails. Also change the resistor for the low battery/power led to 4k7, so you do not kill the LED and the battery detector. If you want better simply remove the USB and stack 2 CR2016 cells in the holder ( will fit and does work on a lot of these holders) to give a 6V supply rail with the only drawback being shorter battery life.

[EarthLord]

Thanks for your suggestions. I'll see how the power bank supply rail is and add better decoupling for the uCurrent.
Isn't the absolute maximum voltage of the maxim opamp used is 6V, this cutting it too close for using coin cells in series?