48V LiFEPO4 battery possible problem, shuts down after short distance

[Fiddlerpaul]

I put in a used battery in my new ebike build (Magic Pie Edge).   It ran about 10 kms and then quit, with lights saying full charge still.  Recharged and it went about 6 kms the second time.   Took apart the battery and all cells measured 3.3 V except one which measured 2.91.  The battery is a little over a year old, was used for three months and then sat for 10 months and was charged every 2-3 months in that period.
Not sure if the sag factor under load would have pushed the battery so low as to shut off or if it's some other issue altogether.   The GM Canada supplier is out of cells as well so I could use advice on picking one up. 
He also told me that you should put similar aged cells in as replacement?  This seems unrealistic, but any advice on this whole thing would be appreciated.
Is it possible to just remove the one bad cell and use the battery? (guessing not but thought I should ask)

thanks Paul

[Bikemad]

Hi Paul andto the forum.

Fully charge the battery and then measure the voltages of the cells again.
If the suspect cell voltage is not at the same voltage as the other cells after a full charge, it could be that the cell is out of balance and in need of a top up charge to bring it up to the same level as all the others, so the first option would be to charge the individual cell after the battery has been fully charged using an old phone charger.

IMPORTANT NOTE: You would have to carefully monitor the voltage of the cell throughout the charge to ensure it doesn't exceed 3.65V, as the phone charger would continue to try and charge it all the way up to ~5V, which could be damaging for a LiFePO₄ cell. 

If the cell voltage was already at the same voltage as the other cells after a full charge, then a replacement cell will be required.

Unless you can find a faulty second hand battery pack going cheap to use as spares for yours, you will have to search online for a suitable 3.2V 10Ah cell, but it needs to be the correct dimensions and preferably have the same type of terminals:


Unfortunately, it is not straightforward to use just 15 of the 16 cells, as the BMS and the Charger would both need to be changed.

Alan
 

[Fiddlerpaul]

Thanks Alan

I didn't know you could rehabilitate a cell like that.  Would be great if I can. 
To use a phone charger, I guess I would have to cut into it and put one wire on each terminal..?

[Bikemad]

I didn't know you could rehabilitate a cell like that.  Would be great if I can.

If the cell voltage is low because it is just out of balance, it should revive, but if it is charging to the correct voltage and then dropping lower than the other cells under load, it will need to be replaced.

To use a phone charger, I guess I would have to cut into it and put one wire on each terminal..?

Depending upon the type of charging plug, you may be able to attach wires to it without removing the plug, but most modern chargers nowadays use a USB socket, which is why I suggested an old phone charger.

If you know anyone who is into R/C Models, they may have a suitable charger that can be used on single cells. But old phone chargers are simply being thrown away nowadays, so cannibalising one is a cheap option.
I like the old Nokia phone chargers with a more standard type of power plug:



You can easily attach a couple of wires to the plug (Positive on the inside and Negative on the outside) without having to chop the plug off.

Alan
 

[Fiddlerpaul]

Ok Alan,  I charged it up and most of them are 3.38.  The bad one is 3.24.   There is one reading 3.46.
Not sure which way I should go from this result but guessing it's within range and is probably needing replacing..? 
I had a look at that site you linked to and there was only a few with screw on bolt heads and they were all the wrong dimension, 26 x 70 x 180 mm.

[Bikemad]

I have just checked my 48V 10Ah GM pack that has not been used since it was last charged 10 days ago, and it is currently reading 55.8V (that's an average of 3.4875V per cell).  Considering that your battery has only just been charged, I am surprised that your cells are all at a lower voltage than mine.

What is the maximum output voltage of your charger?
Mine is regulated to 58.4V (3.65V per cell) and this is the voltage at which my charger terminates the charge.

I wonder if your voltmeter is reading slightly lower than mine? 

Alan
 

[Fiddlerpaul]

Yes I think it's the voltmeter reading differently.  I've got one that reads crazy high and another that is hard to get to work but when it does I think it says that the good cells are slightly over 3.5 (couldn't get it to read on the bad cell).  Still, the proportions are probably the same.    I guess the question is, do I try to boost it or is it not going to work?

[Bikemad]

I guess the question is, do I try to boost it or is it not going to work?

My answer is, if you don't try to boost it, you will never know whether it makes a difference or not! 

Alan
 

[Fiddlerpaul]

Alan,

Ok, boosted it up to  3.36 and it immediately dropped to 3.31.   I was able to go further, somewhere in the neighbourhood of 17-23 kms.  Not good enough for the long term.
I've looked high and low on alibaba and aliexpress and so far everyone is out of the cells that have screws on top.
I also am going to ask Gary at GM Canada if he will cannabalize a battery for me as he said he might.
I also tried charging the cell again singly tonight and was able to get it to go all the way to 3.42 from the initial charge that went to 3.36.   In both cases, however, it returns to 3.31.  But I do wonder if this is progress and I should continue doing this.
Is there anything else I can do at this point?   

[Bikemad]

I have been doing a bit of research and it would appear that,"a LiFePO4 battery can be safely overcharged to 4.2 volts per cell, but higher voltages will start to break down the organic electrolytes".

If you look at the following information I found on PowerStream's website, you can see that the cells need to be charged to 3.6~3.7V to achieve maximum capacity:



The above results were obtained from tests carried out on used 26650 cells from four different manufacturers, and charging was done with a lab power supply, the voltage was set and the power supply current was limited to 1.6 amps. The charging would start out at 1.6 amps until the voltage rose to the set limit, the current would then automatically reduce. In most cases the charge was terminated when the current had dropped to below 30mA.
The cells were then discharged at 2.5A rate (approximately 1C) to 2.6 volts termination voltage to measure their stored capacity.

The bit that really surprised me was the available capacity of the cells when they were only charged to 3.3V, as this was only between 20~31% of their full capacity. 

Let's take another look at your results.
Most cheap digital multimeters are typically accurate to within within 0.5% (but can be more inaccurate if their internal battery is too low), therefore your measured 3.42V ±0.5% could actually be 3.4~3.43V, which is still well below the typical recommended maximum charging voltage of 3.6± 0.05V per cell.

I would try to charge the single cell again and see if you can get it up to at least 3.7V, and then note what the resting voltage is after you have disconnected the makeshift charger. 
You should not cause any damage to the cell as long as you don't exceed 4.2V. 

Please note, if your 10Ah cell was only at 80% of its full capacity, it would take ~8 hours to fully charge the cell @ just 500mA.   
My old Nokia phone charger's rated output is only 355mA @ 3.7V, so it would take at least 11.25 hours to top up the same cell from 80% to 100% capacity.

What is the maximum output voltage of your charger?
Mine is regulated to 58.4V (3.65V per cell) and this is the voltage at which my charger terminates the charge.

Check your charger's output voltage (if you haven't already done so) to ensure it is high enough to fully charge (and correctly balance) the battery pack.
If the charger's voltage is too low (<57.6V) it may not fully charge and balance all of the cells in the pack.

Alan
 

[Fiddlerpaul]

Alan,

So I think what you're saying is my whole battery might be low.  I just had another go at boosting the cell with a different charger that was 5V 2500ma  and it boosted it to 4.2 in about a half hour.   But it has now sunk all the way back to 3.28 where it was last time I boosted it. 
Pretty much resigning myself to buying a new battery and using this one as a spare tire.
Thanks a lot for all the help.

[Fiddlerpaul]

Alan,

Conversing with Gary of GM Canada, it came out that my one cell that is higher than the rest at around 3.63 is stopping the others from going higher.  He mentioned you had used a light bulb as a way to put a load on this cell and bring it down.   In process of doing that now...will see if it increases the ability to go higher for all cells.

[Fiddlerpaul]

By using a light bulb to put load on a few cells that were abnormally high, I am now able to get the following readings after charging.
3.57
3.42 (one knocked down)
3.57
3.31 (the bad cell that has not been able to be raised any higher)
3.56
3.56
3.56
3.57
3.57
3.57
3.56
3.57
3.57
3.56
3.36 (the other one I knocked down)
3.57
Hope this translates to something approaching acceptable distance, but I do fear the weak cell will shut down the system well before the rest are done.

[Bikemad]

Hope this translates to something approaching acceptable distance, but I do fear the weak cell will shut down the system well before the rest are done.

I'm afraid you might be right, as it is not looking so good on that one low cell, but I would expect the BMS to allow those other two cells to accept further charge until they were equal all the others if left on charge for long enough for it to properly balance the cells.

Although, I seem to recall Gary mentioning that some versions of the chargers do not cut back in again (or provide a trickle charge current) to allow the balance process to continue after the green light comes on at the end of the initial charge. 

Alan
 

[Fiddlerpaul]

Alan,  I was reading up on some of the programming settings you recommended that might help a weak battery.
I saw where you mentioned raising the undervoltage cutoff level up to 46-48 (from 42 I presume the default on mine) as a way to prevent the bms from shutting down under heavy load.   I don't quite understand how setting it higher would accomplish that as I would have thought that would cut it off sooner at a higher voltage level.  I think you also mentioned lowering the current draw but not by how much.  I've also set the acceleration to 45. 
Any comments you have on settings to compensate for this situation would be great.