diy lipo capacity tester

[reyntjensm]

I've got my hands on a broken (the previous owner told me one of the cells was broken) e bike battery with 7 lipo cells. I'm pretty sure the BMS was the culprit, but nevertheless i would like to measure the capacity of the battery's. Sadly enough, no specs are listed on the cells. So i did some research and constructed an arduino capacity tester with a load resistor ( like this https://create.arduino.cc/projecthub/mircemk/lithium-battery-capacity-tester-b65d7c). I have used a N channel mosfet instead of the relay. To test the accuracy of my setup, i powered the load circuit with my RD6006 since this is the only device i have with an Ah readout and compared both results. The arduino readings are always 12-15% higher compared to the RD6006 results. I would like to increase the accuracy of my test setup but i don't really know how to do this. Since i'm now only measuring twice a second, doing more measurments a second should increase the accuracy. But i'm wondering about the accuracy of the RD6006? Is this a good device to compare with? So first of all i'm looking for any tips to increase the accuracy of my test setup, on the other hand i would like to have a good way to verify my measurement results. I don't have the ability to use any fancy gear to compare to. I only have some linear power supply's and a basic oscilloscope. Any suggestions?

[vjekobalas]

I'm no expert in batteries but have had to get my hands dirty with
my NimH Toyota hybrid battery and had a few guys over from the Rimac
company (fastest Electric car in the world) where they do testing of
really large batteries, anyway the takeaway for this post (if I have it right -
and a quick look at Lion Battery University web pages confirms this) :
For NimH the charge/discharge curve is rather flat i.e. it is difficult to tell
the battery capacity. You can only tell the capacity by discharging to
minimum voltage and cycling to see if you can improve the capacity.
For Li-on the capacity can be identified more or less directly from the
OC voltage once the battery has been lying around for a while
(see the charge/discharge curve per cell).

[Hiemal]

using just a resistor doesn't give you a stable load over time; as the battery voltage drops, so too does your current which can skew ratings somewhat.

A better way to do it would be to use a constant current load, hook that up to your battery and then measure both current, as well as voltage, from fully charged (4.2v) down to fully drained (3v).

However much current you set it to, multiply by the total time it took for the battery to discharge (if, 1 amp, and it took 2 hours to discharge, then have a 2000 mAh capacity battery) and whammo, battery capacity tester. 

Other thing is, most datasheets specify capacity over different loads. Every li-ion battery will have a smaller useable capacity the more current you draw from it. It's generally useful to have a datasheet for the specific battery type you're testing, but if not, I'd say somewhere between 500 mA and 1 amp is a good place to start for most.

[Doctorandus_P]

Absolute accuracy is irellevant as long as you are comparing measurements made with the same method.

If you start with fully charged cells, then discharging with a simple resistor and a relay to cut of at an (Accurate!) low voltage is sufficient.

If all you want to know is the relevant quality of the cells in a single battery pack there is no need to make it more complicated.