Power cuts out on LiFePo4...

[Supatrike]

Hi all,

A brief history to my problem:

Last year I bought a 36v GM motor and battery from China.  After a long bit of troubleshooting I discovered that I got a 24V LiMn battery in a 36V LiMn case.  After a bit of haggling, China sent me a new battery and charger.  The new battery was a LiFePo4 36V.  It worked fine for a couple of rides and then started cutting out upon full acceleration.  Les (from the list) pointed out that using the LiMn charger on a LiFePo4 battery would send the cells out of balance and lead to the type of problem that I am experiencing.  This leaves me with two questions... 1) is this the case?  (see vimeo link below) and 2) do you have any advice for me to proceed?

This bike performs so much better with the 36V battery!   I can still ride with the 24V but... eh...

Check out the vid...
https://vimeo.com/40197953

and as always thanks!

[Bikemad]

Bart, it is important to know the battery voltage just after the pack is fully charged, which is why it was the first question I asked:

I have some questions for you:

• What is the voltage reading of the fully charged pack?
• How does this reading compare with the voltage stated on the charger?
• Can you try and measure the battery voltage just before it cuts out?
• Did this battery come with a LiFePO₄ specific charger

After watching at your video, the battery appears to be cutting out at around 38.5V, which should not happen if all cells are balanced and working correctly.
38.5V divided by 12 cells = ~3.2V per cell, which would be an acceptable voltage under a slight load.

These LiFePO₄ cells are typically charged to approx 3.65V, which should give a pack voltage of ~43.8V fresh off the charger.
If you have a weak cell, and the charger is only supplying 42V max., it's theoretically possible for 11 cells to be charged up to ~3.6V and the weakest cell to only reach ~2.4V, simply because there is not enough voltage being output from the 42V charger to bring the weakest cell up to the same level as the other cells.
The already low voltage of the weak cell would probably drop low enough to trigger the BMS's LVC as soon as any load was applied.

If you can access the balance lead connections on the BMS board (presumably located at the handle end of the pack), you should hopefully be able to measure the voltage of each individual cell to see if one is substantially lower that the rest.

Be very careful not to cause any short circuits with the meter probes while you're trying to measure the voltages.

It's hard to say whether a charger set at 43.8V would cure the problem, as it could still be due to a faulty cell or even a faulty BMS board.

If you can measure the fully charged pack voltage (and ideally, the individual cell voltages too) it may tell us a lot more.

Alan
 

[Supatrike]

Thanks! At full charge it is 43.1V  I can try to test the cells individually too, but do you think it would be better to first get the right charger?
Bart

[Bikemad]

Bart, I'm not convinced that the charger is responsible for the cutting out problem, I feel it's more likely to be a failed cell.
If it's being charged to 43.1V, I reckon that 3.59V per cell is pretty close to where I would expect them to be.

Was the voltage measured while it was still on charge, or was is checked just after the pack was disconnected from the charger? It would be interesting to see if there is a great difference between the two.

I was looking at your video again, and I thought the voltage dropped far more than it should under a no load situation, so I did a voltage check on my battery as a comparison. The highest voltage drop I measured was ~0.15V when I accelerated to maximum throttle with the wheel off the ground. If I twisted the throttle very slowly it dropped less than 0.1V all the way up to (and whilst holding it at) full throttle. Whereas on your video, the voltage quickly drops from 40.1V to 39.1V with very little throttle, and this was before the throttle was increased to the point where the motor cut out, so your voltage is actually dropping at least ten times lower than mine does!

I now think you may need to check the cell voltages twice, first with no load, and then again while under a slight but constant load.
If you raise the wheel off the ground as before, and then very gently accelerate to a point before the motor cuts out and press the cruise button, it will hopefully maintain a steady load on the battery while you check the voltages of the cells.

I'm expecting 11 cells to remain pretty much the same, and one cell to drop substantially in comparison to the rest.

Please let us know both sets of voltage readings, assuming you are able to measure them.



Alan
 

[Supatrike]

Sorry for the long delay in getting to this problem - busy life.

Alan, thanks for your suggestions.  I took the battery apart yesterday and took pics.  There was quite a bit of moisture trapped in the shrink wrap and the terminals were badly corroded.

Negative terminal:



Positive terminal (this was the worst and a considerable amount flaked off before I took the picture):



All of the cells measured 3.4 or 3.5V except for the last one:



The last cell was the weakest and it was the one attached to the badly corroded positive terminal. 





My first inclination is to just clean up the corrosion, resolder and test it again... but I'd like to hear from you all first.  Could the corrosion be from a bad cell and not from damp conditions?

Bart

[Supatrike]

I cleaned the solder joint and resoldered.  After recharging the dead cell reads 2.7 and the rest read between 3.4 and 3.7.  Anyone have any advice for replacing the cell?  where would I be able to buy one?  and how would I be able to shrink wrap the batteries?

Bart

[Bikemad]

Hi Bart,

I was just looking back through the recent posts and rediscovered this one. I think the corrosion may have been cause by a small leak from the sealed cell. Hopefully GM will be able to supply a replacement cell of the correct Ah capacity to suit your battery. Send an email to David (wyh@goldenmotor.com) and see if they keep these cells in stock as separate items.

Re-wrapping the cells is probably best done with heat shrink tube, but a piece large enough to cover your battery pack will not be cheap. You would need a suitable length of 150mm diameter heat shrink tube which should be the ideal size to fit over your cells before applying the necessary heat to shrink it tightly into place.

A cheaper option would be to bind the cells tightly together with some fibreglass reinforced filament tape.

Alan

[Supatrike]

As always, thanks Alan.  I really can't deal with China anymore - every time I lose money.  I'm seriously thinking of just scrapping it all and going back to driving my car... so sad.  What a waste of time, energy, and money!  This forum at least gave me a shred of hope. I'll email, but I have never gotten an email back from them excepting the time Gary wrote to them for me.

[Supatrike]

As I figured, no response from China.  Does anyone have any advice for building my own battery?

[Bikemad]

Bart, I've just noticed that Gary now sells these cells separately:

10Ah LiFePO₄ cell
12Ah LiFePO₄ cell

This might be your easiest option.

Alan