Electronics > Projects, Designs, and Technical Stuff

Determining capacity of a battery using EIS (electro impedance spectroscopy)

(1/1)

LLee:
Hello so I want to determine the capacity of a battery  (NiMh) using EIS technique.

My plan is to put an AC voltage on a battery cell, and measuring the output current with a scope.
I measure the phase and amplitude so I can make a Nyquist plot (imaginairy impedance on y axis and real impedance of the battery on x axis).

I will do this for multiple batteries that I know are good and batteries that I know are bad (lower capacity).
This gives me multiple Nyquist plots.

So whenever I have a new battery I can use this trick to know if its a good or bad battery, depending of his Nyquist plot.
But I can't find a lot of tests and how to do this, most of the info I find are very hard to understand (chemical stuff).

Does someone have any experience with this, and is this even possible?

Someone:

--- Quote from: LLee on February 25, 2024, 03:50:00 pm ---Does someone have any experience with this, and is this even possible?
--- End quote ---
The characteristic curves will vary between both brands and models. So while it can work for a single-source or captive battery (say a particular line and model variant of tool batteries) its not a general solution which can say things about an arbitrary battery or cell.

moffy:
You could do some basic experiments with a PC, soundcard and a few components. Might be worthwhile as it would cheaply give you a way of seeing if it is a relevant technique for what you want to achieve. Use Kelvin connections. Not sure how useful it would be apart from measuring internal series resistance. Might it also be affected by the level of charge in the cell?

RoGeorge:
The battery impedance depends a lot with the battery's temperature, charge level, load, frequency, brand, age, etc., so not very useful as a general diagnostic method.  Then, there is the self-discharge rate, when a charged battery is waiting to be used.  In the recent years, a new type of NiMH appeared, the ones with low self discharge rate, significantly better than the first NiMH.

Best test is to charge the battery, store it unused for a month, then discharge the battery with a known load while measuring the time until the voltage drops to 1V or so per element.

For outdoor use, lower temperatures will diminish the battery performance, particularly for the NiMH type.

max_torque:
The "advantage" of EIS is the ability to estimate (note estimate,not measure!) a batteries useable SoH (note not capacity!) in a short time period.  This is clearly useful where you haven't got the time to actually do enough of a charge / discharge cycle to otherwise estimate capacity (in Ah) from the relaxed Open Circuit Voltage and delta Ah.  The downside is the large amount of characterisation you have to do to get to a set of correction tables and characterisitic surfaces to allow that short EIS measurement to be converted into a useable capacity estimate.

There is plenty of reading and papers on the subject, and it's necessarily quite complex and inter-dependant on a large range of factors and conditions.

For most people it's not worth the effort and it often leads to highly eroneous estimation because not enough training data is available to characterise the module with sufficient fidelity.