Page 1 of 1

### Measuring Pi power with a Keweisi USB tester

Posted: Fri Mar 29, 2019 9:21 am
I bought a Keweisi USB tester to measure the current of my Pi + Arduino project.
When I measure a Pi Zero I get a reading of 20mA, which is way to low. I put my MultiMeter in between and measure 82mA average, which is about what it should be. Keweisi USB tester broken? Maybe, but when I measure the current of a charging phone or powerbank, I get about the same value on the Keweisi and my MultiMeter. Does this mean that the current that the Pi is drawing is not flat causing the Keweisi to miss-sample? Could I fix this by adding a Capacitor in parallel over the USB black and red wire to flatten out the current? If so, what would be appropriate value of the Capacitor?

Thanks

### Re: Measuring Pi power with a Keweisi USB tester

Posted: Fri Mar 29, 2019 10:42 am
Assuming that the phone charging is at a much higher current than 100mA, the Keweisi USB tester may not be accurate at lower current values. Large variation in accuracy over the measurement range is a common accuracy problem of low cost measurement devices.

### Re: Measuring Pi power with a Keweisi USB tester

Posted: Fri Mar 29, 2019 3:10 pm
These things work by measuring the voltage drop across a small value series resistor (current sense resistor) in line with the ground side of USB power. As a design decision one wants the voltage drop across this resistor at the maximum current (maybe 3 Amps) to be relatively small so that the voltage to the device under test is not too large. Thus the voltage drop at small currents is that much smaller.

In better versions of USB power testers there is an amplifier to increase this voltage to better match the range scale of the embedded microcontroller that is reading the current sense resistor voltage drop which helps mitigate inaccuracy. Cheaper units simply sense directly with the microcontroller A/D input throwing away a great deal of the scale/resolution of the microcontroller A/D so they don't do well at low current levels. Also A/D linearity near the ground rail isn't going to be great.

I have a number of these cheap USB power testers and I've considered putting a larger value current sense resistor (say 10x whatever's there) on one of them which would shift the output reading proportionately, but I've not gotten around to it. The trade off is that one would lose the top end current capability and would have to remember to scale the output reading mentally, but it would make it more useful for monitoring currents typical of Arduino sorts of things. On the other hand it probably wouldn't work with circuits that have large intermittent current spikes such as a processor with a WiFi (e.g. RPi Zero-W or ESP32).

### Re: Measuring Pi power with a Keweisi USB tester

Posted: Fri Mar 29, 2019 8:43 pm
I have some of those

Tried to measure power consumption of RPi3B, RPiA+, Arduino Uno as I wanted to compare real world power consumption for 1-day use

I find that these units introduces voltage drop to the supply to RPi and I always get the lightning icon on my RPi3B, and sometimes on my RPiA+.