Measuring the characteristics of a cell

[Ohms_Cousin]

I am interested in measuring cells so that I can batch cells together for making 48vnom batteries - batteries will be in parallel connected to a bussbar.

I am looking at importing a large number of cells so that I can test them all and keep the best cells and sell the rest.

I need to "Match and Batch" for real.





Is this tester "overkill" for my needs or is an instrument like this very good for getting down into the weeds with cells to be able to match them all very closely. End use is a lot of batteries on a boat so need the cells to be as closely matched and binned for making into batteries, as possible. I want to do everything I can, while still DIY'ing, to make sure my battery setup is as well setup and as safe as possible. We dont get a second chance out there on the high seas, so this is really really critical to bin those cells to make battey packs as closely matched as possible.

I know of the ZKetech - and its available and about $100 locally (not worth importing 1 of these from China so better to buy locally) - but is this going to report on a cell within very fine tolerances of each cell ?

Or is it just not worth binning cells to that level ?

 

[sunshine_eggo]

That device does not report cell capacity, so it and the ZKetech do different things.

The 1kHz signal referenced is also used by the YR-1035+ for ~$50 to measure cell internal resistance. It's also what the prismatic cell resellers in China use to confirm cell resistances.

Using a YR-1035+ to measure cell resistance is commonly used on this forum. I have been using its predecessor, the YR-1030, to measure the internal resistances of 6S batteries for commercial purposes for 8 years.

 

[Ohms_Cousin]

How does the accuracy of the YR-1035 compare to the instrument I have linked?

In general is it worthwhile getting super accurate with a higher end instrument to be able to BIN cells for resistance? How tight do we really need the binning to be?

Despite costs and time issues to do it, is matching cells down to a really fine level something that pays off or there is definite diminishing returns at a certain point and its not worth going any further and if so at what point does the diminishing returns kick in?

From quite some time reading this forum, my understanding is the cells we are buying at our price level at not very well BINNED - iirc the cost per cell to properly diagnosis it and BINN it at a very high level was something like $60 per cell in labor charges alone because of the time it takes to do this - if one is willing to undertake this work themselves and work from a larger pool of cells in order to find those closest to each other (thus, tight Binning), which instruments are recommended ? The instrument linked above runs about $1800 locally.

If it means having some years relatively trouble free with the batteries keeping tight voltage differences (as a result of the tight Binning) when being charged and discharged (I think I would be discharging at a 0.5 C rate most of the time as I am driving a motor in a boat and pulling from several batteries connected to a Bussbar) then the time, trouble and cost to get this right up front and out the gate is worth it (at least to me...ymmv)

 

[RCinFLA]

How does the accuracy of the YR-1035 compare to the instrument I have linked?

Apples to Oranges comparison. YR 1035 is only measuring the Rs component. It is mostly just cell ohmic resistance.

The parallel Rct and Cdt is ion migration path component. It is the important indicator of cell condition. It gets worse as the cell ages.

 

[Ohms_Cousin]

Apples to Oranges comparison. YR 1035 is only measuring the Rs component. It is mostly just cell ohmic resistance.

The parallel Rct and Cdt is ion migration path component. It is the important indicator of cell condition. It gets worse as the cell ages.

View attachment 251426

Hi RC,

Thank you for the informative post. Could you tell me in English what you just said

I'll take a stab but please tell me if I am interpreting correctly. I think what you are saying is that the YR1035 is not a very sophisticated measurement and that the much higher end instrument is reading extra data from the cell (that the YR1035 does not give) that gives a much more in depth look at the cell. But I could be wrong and that is not what you meant at all.

 

[RCinFLA]

You don't need the fancy BK meter shown.

You can do close to same thing with ZKETech tester.

The initial voltage slump from open circuit rested cell state is Rs.

In the programming sequence, put in 5-minute zero current rest points at approximately predicted 25%, 50%, and 75% SoC during discharge and charge curves. During the 5-minute recovery to no-load rested state represents the Rct-Cdt component. The amount of voltage delta in the slump/recovery/bump is most important.

 

[bestconcreteblock]

Apples to Oranges comparison. YR 1035 is only measuring the Rs component. It is mostly just cell ohmic resistance.

The parallel Rct and Cdt is ion migration path component. It is the important indicator of cell condition. It gets worse as the cell ages.

View attachment 251426

I'm finally starting to understand this a little better. So the R_ohmic shouldn't change much over time because the physical resistance of foil and connections shouldn't change too much. Obviously this graphic is for display purposes, but is it common for the R_ionic to be almost double the R_ohmic?

 

[RCinFLA]

I'm finally starting to understand this a little better. So the R_ohmic shouldn't change much over time because the physical resistance of foil and connections shouldn't change too much. Obviously this graphic is for display purposes, but is it common for the R_ionic to be almost double the R_ohmic?

Correct.

If a cell is subjected to high currents with a lot of temp cycling of electrode material to neg copper and pos aluminum current collectors, the printed on electrode material can delaminate from metal foil increasing Rs (R_ohmic).


Severe degradation of electrolyte can also increase Rs. but R_ionic will suffer more from degraded electrolyte.

With moderate cell current, R_ionic is the more important factor. It determines the overpotential voltage slump during discharge and bump during charging. It can increase 3 to 5x over lifetime of cell. Depending on peak current demand, an old cell will eventually have too much terminal voltage slump to be useable in the application.

 

[bestconcreteblock]

With moderate cell current, R_ionic is the more important factor.

So the majority of the internal resistance values advertised by sellers are most R_ohmic and only a fraction of the full story. I guess they do have some value for comparison purposes?

It can increase 3 to 5x over lifetime of cell.

Wow, that is significant.

Depending on peak current demand, an old cell will eventually have too much terminal voltage slump to be useable in the application.

This is the part I am interested in. As the battery ages, R_ionic will increase. For the same load current more stored energy will be made into heat(I assume?) My question is can we temporarily ignore safe low cell terminal voltage(2.5V) during these high load situations as long as the heat is being dealt with?

 

[sunshine_eggo]

IMHO, accuracy is almost irrelevant provided it's "sane" and measurements are consistent and repeatable. I have found a high degree of correlation between various instruments, and the #1 factor in matching is consistency.

You're comparing values measured by the same instrument looking for outliers to discard. Do you think the capacity tester is 99.9% accurate? Doubt it. You're using it for comparison purposes. Does it matter if it's 10% off if it's consistently 10% off? I say no, and that worked for me commercially.

 

[bolek_bolek]

Just asking, @RCinFLA, taking the graph in your post took like 7 hours = 280AH/40A ?
Or the instrument is using some "cheats" for a quicker result?

​

 

[RCinFLA]

Just asking, @RCinFLA, taking the graph in your post took like 7 hours = 280AH/40A ?
Or the instrument is using some "cheats" for a quicker result?

Graph done with ZKETech EBC-40 tester. It takes C / 40A hours in time. Important to do the charge and discharge process sequentially successive to get the full capacity span for overpotential voltage during discharge voltage slump and voltage bump during charging.

Overpotential voltage is very much affected by temperature, so you want to keep to 20-25 degs C to have apples to apples comparison.

For LFP cells, the charge bump and discharge slump are very close to symmetrical for same charge and discharge cell current (charging has slightly lower overpotential). Split the difference between the charge and discharge curves and you get close to OCV rested cell voltage curve. You do have to do some math manipulation to normalize charge and discharge data to 0 to 100% state of charge/discharge before you plot the two together.

AH efficiency is higher than true wH efficiency due to the higher cell voltage during charging. This is the overpotential voltage overhead loss.

I rarely run full discharge/charge curves because of the time required. Almost as much info is achieved by just doing 3-5 minute discharge test on new cells to measure their overpotential voltage slump at 40-50 amp current, 0.2 to 0.4 C(A) range, which is the most important parameter.

BTW, variations in the curve in the 90-100% SoC is due to variations in electrode thickness between anode graphite and cathode LFP. This is normal build tolerance for printing the electrodes on metal foil. Normally, graphite is intentionally made a little thicker to have graphite anode capacity be 15-20% more capacity than LFP cathode. This is done because graphite degrades more during life cycle of cell. New cells may take 5-10 cycles before the overlap of anode and cathode capacity settles down to final overlap relationship.

LFP SoC cathode electrode potential curve is almost dead flat at 3.43 vdc. The slight voltage bumps in overall LFP cell discharge curve is due primarily to anode graphite potential. Cell voltage is cathode potential minus anode potential. LTO cell anode has a very high potential yielding a very low cell overall cell terminal voltage of 2.1 to 2.8v.