Testing capacity of a battery

[LLee]

I was wondering if someone could provide some input on the following:

I want to set up a testing procedure to determine the capacity of a battery relatively quickly, or to identify defective batteries that have a lower rated capacity than what the manufacturer claims. Therefore, I'm looking for a method that doesn't involve fully charging and discharging the battery, as this is time-consuming.

Does anyone know of any suitable methods for accomplishing this? I've come across Electrochemical Impedance Spectroscopy (EIS), but I'm unsure if it's feasible for a home project. Additionally, I've seen mention of Electrochemical Dynamic Response (EDR), which involves momentarily applying a load to a battery and observing the voltage response.

Has anyone had experience with this? If so, how would you recommend starting such a project, and can I do this with arduino...
Thanks

[Solder_Junkie]

I have a West Mountain battery analyzer, while I appreciate that you want "quick" results, I cannot see how you can obtain mAh (or Ah) measurements without discharging the battery fully... especially if you want to raise the matter with the battery manufacturer.

This is the battery analyzer company http://www.westmountainradio.com/cba.php

SJ

[CaptDon]

Only 'crude' estimates can be obtained without doing full battery cycles. Also, with aged batteries a couple of full charge / discharge cycles will actually show improved results with each successive test showing that the battery can benefit from a few 'full cycles'. There are so many different ways to do calculations during discharge, Constant current until the voltage drops to a specified level or constant power (important when the batteries drive a SMPS and current goes up as voltage goes down). You can't just take a battery of unknown charge condition and tell anything about it. Also, how about batteries that have developed high internal leakage which test reasonably well but go dead if left sitting for a week? Hard to weed them out sometimes.

[ajb]

If you have enough batteries to be worried about testing time, then running tests in parallel should speed things up without sacrificing usefulness of the results.  It would be an interesting project to build a multi-battery test system that could recycle energy by discharging the batteries back into the same DC bus used to charge them.  Basically a 4-switch buck-boost converter for each battery, something to dump excess energy into when the bus voltage gets too high, and then some sort of controller to schedule charge and discharge cycles. 

If you want a simpler arduino-friendly solution, you could use an off-the-shelf charge controller (IC or module) and a basic constant current discharge circuit, and use an MCU to simply turn the charge/discharge circuits on and off and log current/voltage during the tests.