Possible bad cell in GM 36v bottle battery

[Mrblenny]

Hi guys,

I have owned a SmartPie 5 (front wheel motor) with the small 36v bottle battery from golden motor for a few months now. Initially, pretty positive experience for the 25km pedal assisted commute, although occasionally the battery cut-off would kick in after riding for about an hour, even when drawing less than 5 amps. I found this a little worrisome, as by my back of envelope calc's, it meant I was only getting 4-5ah out of the battery when it is rated for 8ah. However, since my commute is just under an hour, I stopped worrying about it, and was just careful to keep power usage low near end of the ride. 

But the past few days, it is now cutting out when drawing 5amps only a few minutes into a ride  My suspicion is a bad cell in the pack, or perhaps a faulty BMS not load balancing the cells correctly?

Before I open up the pack and start testing cells, can anyone offer me advice on the most likely scenario, and can I pursue an RMA with GM (I bought directly from goldenmotor.com)? The battery cost quite a bit, but return postage to my part of the world would eat up over half price of a new unit so my options are a little screwy - downside of long distance internet purchases I guess...

Any comments appreciated 

Thank you,
Damian

[Bikemad]

Hi Damian andto the forum.

I suggest that you email sales@goldenmotor.com and explain the problem to them and see what they suggest.

If it's a faulty cell group or BMS problem, they may be able to send replacement parts for you to fit yourself (or have fitted by someone locally if you don't have soldering equipment etc.).

However, as the battery is obviously faulty, you may not be able to legally return it to China as the shipping regulations regarding dangerous goods may not allow "faulty" lithium battery packs to be shipped. 

I had a similar problem last year with a Non-GM 48V 20Ah battery that I purchased on AliExpress that only delivered just over 7Ah from fully charged until it cut out. After opening a case with AliExpress, I eventually settled on a partial refund as compensation for the missing battery capacity, as I was not prepared to risk shipping a "faulty" lithium battery back to China (at my expense) in order to obtain a full refund.

Alan
 

[Mrblenny]

Thank you for the suggestion Alan, I have emailed GM and am awaiting reply.

Cheers,
Damian

[Mrblenny]

Update:

I am now in correspondence with GM, sending cell voltages to identify bad cell etc. No word yet on what (if anything) they will replace.

I managed to partially discharge the pack by riding for an hour with a low ~70watt draw. Most cells now read 3.5volts, but cell #1 reads 3.20. An obvious contender for replacement  Cells #4 and #8 read a little high, at 3.6volts. I don't know how good the BMS is meant to be, but the remaining cells are all within 0.02volts of each other.

Does anyone think cells #4 and #8 might be sus as well, or is 0.1volts difference ok?

Cheers,
Damo

PS I have a digital charger that can be programmed for charge/discharge cycles, so I can verify individual cells capacities, but am hoping to narrow it down to a particular cell before testing to minimise how much of the pack I need to pull apart, and solder back together again :-)

[Bikemad]

Hi Damian,

Discharging just 70 Watts from a 288Wh pack should leave a pack with ~75% capacity still remaining (6Ah) and I would therefore still expect to see at least 3.7V per cell.

When you fully charge your pack, does it cut off at the expected 42V, and if so, what does the pack voltage read immediately after the charger cuts out?

It would also be helpful to check that all of the 10 cell groups are reading ~4.2V at the end of the charge (now you are able to measure the individual cell groups).

Most BMS will only balance the individual cell groups at the end of the charge when the battery voltage approaches 42V. If the cells are not being fully charged (or not being correctly balanced) the low usable capacity could simply be due to a faulty BMS (assuming the charger output is regulated to the correct 42V).

Alan
 

[Mrblenny]

Hi Alan,

You are right about the power draw/capacity, I was just going with what the pack gave me when new (it would cut out after about 200w draw down). But of course, with a faulty cell (possibly from outset), the pack never worked to rated capacity.

I did actually measure the cells when fully charged earlier this week, and got 41.85volts for the whole pack - and all the cells averaged 4.18volts with no outliers.

From what you are saying, the BMS does not attempt to balance *discharge* the cells, so the minor differences are most likely not indicative of a faulty cell. I guess the most likely contender is still the cell reading 0.3volts lower than the rest of the pack...

Cheers,
Damian

[Mrblenny]

Update: Good news, GM has offered to replace 2 cells (6x batteries) for free, but I pay postage.

Unfortunately, postage is $60USD, so it might be cheaper for me to buy locally /shrug

I have asked GM what brand/capacity they are using, unless there is a huge difference in price I am thinking it is probably best to get the batteries from them to ensure matching etc?

Cheers,
Damian

[Mrblenny]

Update: I have tested* the capacity of all 10 cells (the pack is a 10S3P unit), and the results are disappointing.

I tested with a smart charger used for my RC models, first discharging each cell at 1A down to 3volts, then charging at 1A back to 4.2volts. At the end, the charger tells me how much current was accepted by the cell.

Only 3 out of the 10 cells accepted over 7000mah - as the pack is rated at 8ah this is a disappointment. Bad enough I have a bad battery in the pack, but even if I replace that, my effective capacity is only 6ah or 75% of rated specification. I have reached out to GM for replacement or refund, will keep the thread updated.

*This took some time, a nominally rated 8000mah cell will take 8 hours to discharge, and another 8 to charge, it took over me over a week to complete the test :-o

[Mrblenny]

Update update:

GM have acknowledged the batteries are unacceptable and we agreed to a refund (less shipping). I am buying a 3rd party battery with samsung cells which should hopefully be a lot better.

Thanks for the help :-)

[Bikemad]

Hi Damian,

It's good to know that you have been given a refund for the faulty battery.

I eventually opened up the 20Ah battery that I purchased from AliExpress to have a look inside and was surprised to see that it was only a 13S4P pack.
I now understand why it doesn't have the described 20Ah capacity, as I have yet to see a 18650 cell with a genuine capacity of 5000mAh.
I'm pretty sure that my battery only has 2000mAh cells making it just 8Ah instead of the stated 20Ah capacity.

Hopefully your new battery will actually contain the correct capacity Samsung cells.



Alan

[Mrblenny]

Hi Alan,

It does seem to be very wild west out there in OEM battery land. A reminder for everyone on this forum that paypal has buyer protection automatically included on most purchases.

Cheers,
Damian