busted lifepo4 cell?

[dingo117]

Hello people, I am from Australia and have recently started myself an e-bike project (in this damn summer heat!) and this forum seems to be plentiful in similar enthusiastic users so i look forward to learning and doing a lot from every one! but to the point i have an issue:

I recently bought a new LiFePo4 48volt, 10amp battery with BMS and charger from ebay, the seller was 3wvpowerhk and seems to be their own "make" of battery pack, clearly with other brand cells and electronics they just seem to have put it all together.
But it was fully charged when i got it, around 56 volts, i have been using it on my 2400 watt electric bike for a few days, i haven't charged it as i was waiting to run it down to see how long it will last, on the last day it shows full power on the e-bike kits volt meter, i got for a ride for 20 or so minutes and it all of a sudden shows empty, and no longer is able to power the motor, i know lifepo4 has a sharp voltage and current drop so i thought it was just that.

I later on proceed to check its "flat" voltage its around 46 volts charge, i try to charge it but on a 6 amp charger after 6 hours it wont charge over the same voltage, i decided to check each cell voltage, by removing the lid from the BMS and using a multimeter one prong on the negative, and the positive on each of the cell balancing wires going to the circuit board.

the cells marked with <<< are the cells of interest, and clearly the negative one i think is stuffed and may need replacing.

cell:
1. 3.29 (3.29)
2. 6.59 (3.3)
3. 9.48 (2.89) <<<
4. 13.08 (3.6) <<<
5. 15.67 (2.59)<<<
6. 19.00 (3.33)
7. 22.2 (3.2)
8. 25.5 (3.3)
9. 28.8 (3.3)
10. 32.1(3.3)
11. 34.8 (2.7)<<<
12. 38.1 (3.3)
13. 41.3 (3.2)
14. 40.9 (-0.4) <<<
15. 44.1(3.2)
16. 47.4 (3.3)

i got the cell voltage by subtracting the voltage from the previous cell from the reading of the current cell, i hope this is right! but one cell seems stuffed, any ideas? and use i have the BMS connected properly most of the time, sometimes i bypass the BMS when using the bike but i always connect it afterward, and i used it for the charging.

 

[circuit]

Well, it looks like you have overdischarged one of your cells (and probably some others) when you had your BMS bypassed. Call it karma..

 

[dingo117]

thing is i have spent plenty of time going up some damn steep hills around 30+ degrees that would have pulled way more current, a few days later i rid it on flats in a bike park voltage was fine then 50+v, then a few days after that was riding on flats again and that's when it all of a sudden went flat, at 2400 watts it would only pull 50 amps if i am maxing it out which is pretty close to its continuous draw rating and way under its maximum burst rating. also do you have any advice or just want to leave a smart remark as if i did something evil to deserving bad karma. only thing i did wrong was buy from chinese sellers lol

 

[Tench]

You can get a more accurate picture of the cell voltages by measuring differently, it may give you the same results but does help to minimise possible errors. measure from the first pin to the second pin to read cell 1 voltage, then move your - probe to pin 2 and put the + on pin 3 to read the next cell, then move - to pin 3 and + to pin 4 etc. Each pin is the positive for the cell below it and the negative for the cell above it in the series.

Add all the cell voltages and check this against the main cable voltage, if its the same your readings are correct, if its different find out where the difference is.

I have a Ping Lifepo4 pack where doing this doesn't read all the cells as one of the main cables is the end of the series and the first pin is the first intermediate cell tap.

When you have all the voltages and they add up to the pack total you can be sure your readings are correct, then you will be able to see the true state of the pack and be able to work towards sorting it. At the moment we don't know for sure if the info you have displayed is 100% correct. using the above method cross references it against the pack voltage as confirmation you have read everything correctly.

 

[Gregory]

Sorry for your loss. There's a lesson in there for you.


Draining cells at 5C (I assume that's why you bypassed the BMS) seems very high for cheap ebay lifepo4. I doubt vpower have 5C lifepo4. Battery is probably more like 1C or 2C?
Got a link to the battery specs?

 

[dingo117]

alright i measured each cell from terminal to terminal, the first cell was from negative on the main output to the first pin on balancing connections, second cell was from first pin to second pin, so on and the last cell was the last pin and the positive main output

1. 3.3
2. 3.3
3. 3.26
4. 3.25
5. 2.69
6. 3.3
7. 3.25
8. 3.27
9. 3.32
10. 3.3
11. 2.84
12. 3.25
13. 3.24
14. -0.5 slowly crept down to +0.03while measuring the whole time i stopped testing then as it took a few minutes to do so, tested again and went from +0.06 back to 0.00
15. 3.24
16. 3.26
total added: 3.3 + 3.3 + 3.26 + 3.25 + 2.69 + 3.3 + 3.25 + 3.27 + 3.32 + 3.3 + 2.84 + 3.25 + 3.24 + 3.24 + 3.26 - 0.5 = 47.57
total measured: 45.7 volts

something really bizarre happened to that dodge 14 cell while measuring it keeps slowly creeping back and forth so it will probably be inaccurate! the main output voltage as you see doesn't equal the adding of the individual cell recordings, also cell 4 seems to have been balanced down to match the rest, but cell 11 and 5 are still low, but i haven't been charging it, thought i should add i dont have the BMS connected while doing these readings so any cell change on 14 is from the multimeter or the cell itself

 

[dingo117]

http://www.batteryspace.com/lifepo426650cell32v3300mah19.8arate10whunapproved.aspx that is the cell there are 3 in parallel per cell giving a peak of 59.5 amps, i also put a fast rupture fuse rated for 32 volt 60 amp on the battery inline so even with out a BMS i have a backup

 

[Tench]

Certainly looks like your 3p cell group #14 has taken a beating. I would say the rest could still be good. The safe option is to replace those 3 cells, then get a 3.7v charger and bring the pack back into balance by charging each 3p group individually or pulling the higher ones down with an halogen bulb until they are all the same. Then charge via the bms and monitor the cell voltages during charging to ensure the bms is not the culprit and draining #14.

Your pack is in reality only a 10ah pack, Lifepo4 really should be limited to max cont drain of 2c with momentary bursts to 3c, regardless of manufacturers claims. If you want it to be reliable you need to lower your current limits or at least double your capacity to 6p.

 

[dogman dan]

How can I put this. Even with the bms, it sounds like you would have quickly killed this battery.

It's dinky, and adequate for a bike with a 15 amps controller. To get 2400w out of 48v, I think you have a 50 amps controller? You don't need 20 ah of this kind of battery, you need 30 ah.

It's battery murder.

However, I'm still surprised one cell gave up the ghost that fast. Likely it was slightly defective. But if it wasn't you'd have seen a lot of good cells doing the same thing pretty soon.

 

[dingo117]

well i can understand it pulling the full 2400 watts if i am going really really fast or going up a steep hill or accelerating really hard but didn't do any of those just cruising along around 30 kph on flat road around the time it ran out of power so fast, would have been using 1000 watts maybe pulling only 20 amps, and my hill climbing was long before any of this and was not a constant climb was 10 or so meters up before reaching flat low gradients

 

[Matt Gruber]

it has LVC low voltage cutoff. if it didn't, you would keep riding it until more cells were killed.
I don't run my car until it is out of gas to see how far it will go. certainly not an ev. My neighbor asked how far it will go on a charge, i said
I HOPE I NEVER FIND OUT

 

[dingo117]

a lot of people here seem to think i have no clue what i am doing, it was reading full on the battery meter, but as i had stated it very rapidly went to fully empty unable to run the motor in a matter of minutes, i had no intention of running it to zero i was expecting the meter to at least show half way or near empty but it didnt, just seems to have snuck up and went right to fully empty , i know lifepo4 cells should not go below 2.8 volts and to be honest they are not below it... it is just this ONE cell that seems to have randomly inverted its voltage, i know running the battery with out a BMS can lead to a damaged battery but here is the thing i am being the BMS in this case when i ride with out it.

 

[Matt Gruber]

need a wattmeter and all good cells.
i use a stopwatch since roads are flat. and a DVM instead of a bms. an odometer helps too.
but those cells are known to fall off a cliff, no warning.

 

[Tench]

The dead cells have caused a break in the pack, a bit like having a string of tree light bulbs in series, if one blows they all go out! those cells inability to pass more than a nominal voltage is why you have no current and the voltage drop caused by them is why your pack reads empty. Replace them and get the pack back in balance and it will probably all be good again. But expect more cell failures if you use it above 2c or 20 amps.

The lifepo4 chemistry does not drop in voltage through the charge in a linear way like lipo does, it tends to hold its voltage up higher till very nearly empty then drop fast which is why losing the voltage of that group had such an effect on your gauge.

 

[dingo117]

The dead cells have caused a break in the pack, a bit like having a string of tree light bulbs in series, if one blows they all go out! those cells inability to pass more than a nominal voltage is why you have no current and the voltage drop caused by them is why your pack reads empty. Replace them and get the pack back in balance and it will probably all be good again. But expect more cell failures if you use it above 2c or 20 amps.

The lifepo4 chemistry does not drop in voltage through the charge in a linear way like lipo does, it tends to hold its voltage up higher till very nearly empty then drop fast which is why losing the voltage of that group had such an effect on your gauge.

but any idea why the 3 parallel cells making up cell 14 have a negative charge? if they are bust shouldn't they just go really low voltage like 0.5 volts.
also i think i will badger the battery seller to get accurate specs on their cells, but if something is genuinely rated to handle 10 amps continuously per cell it should be fine pulling with 3 parallel cells 30 amps continuously right? and for acceleration and short climbs the 50 amps it may pull are within a 20amp x3 burst range.

 

[Matt Gruber]

they overate cells to make money.
just like car makers claim lofty mpg. so you buy a new car.
or the salesman tells you that pine furniture is oak. to make the sale.
so you ask a friend, is this really oak? he says it is pine!
so now you sue!
next time, ask around FIRST.

 

[Gregory]

but any idea why the 3 parallel cells making up cell 14 have a negative charge? if they are bust shouldn't they just go really low voltage like 0.5 volts.

Pushing current through a discharged cell (i.e. in series with others) can cause cell reversal. The cell is permanently damaged.

 

[dnmun]

how many times can you tell people not to do this? even after you learn about batteries and think you know it all, you should never run the battery down to the end, it hurts them a lot. but the BMS should be able to turn off the output mosfets when it reaches the LVC.

if you will post up a picture of the BMS, then we can show you where to examine it to see if the BMS is functional anymore.

take pictures including the balancing shunt resistors and the two row of transistors running across the top of the BMS.

was it reading 56V while on the charger? or did you measure that when you took it off the charger?

 

[7circle]

...
but any idea why the 3 parallel cells making up cell 14 have a negative charge? if they are bust shouldn't they just go really low voltage like 0.5 volts.
also i think i will badger the battery seller to get accurate specs on their cells, but if something is genuinely rated to handle 10 amps continuously per cell it should be fine pulling with 3 parallel cells 30 amps continuously right? and for acceleration and short climbs the 50 amps it may pull are within a 20amp x3 burst range.

The wandering pos/neg voltage maybe due to one of the three parallel cells haveing reverse plated Li ions and then had huge surge currents bounce" through it from the other two cells. the spot welds could be faulty too. this may have caused lots of heat to one cell to vent.
any way plenty of possible failure types.
but bms lvc is critical you can see two other groups are low.
If they are under 3.00V at 0A they are basically flat.
The contoller LVC ... is useless to protect cell groups
A bms buzzer for lvc on your seat might get the riders attention to stop.

Also you might pull 3C or more when the cells are above 1/2 full, but what can you expext when cells are near empty?
Would be good to see a graph of Rint vs DOD.

Good luck with repair. or claim if BMS is faulty.
7C

 

[dogman dan]

:lol: 20 amps max battery, 50 amps controller. But I didn't give it 50 amps, I swear.

Reminds me of the ol, "I promise I won't come inside you honey."

 

[Tench]

We can talk about this until the cows come home, your still going to have to replace those 3 cells but before you do it maybe worth investigating groups 5 and 11, they may have 1 dodgy cell in their 3p group too. It might be best to order 9 new cells, replace all 3 groups and then you can test the removed cells from 5 and 11 to see which ones are good/bad and you may have 4 spares for next time.

 

[fechter]

Try measuring #14 directly on the battery terminals. It's possible the balancing wire for that one broke somewhere, which may cause the BMS to shut down the output.

 

[dingo117]

I cracked open the battery pack and measured the cell directly still negative and stuffed up, cut them out and measured all 3 cells, they all seem stuffed, and it looks like this cell had black permanent marker marks over it while none of the other cells did, odd maybe it was marked as dud but made its way in the pack some how? the seller said they will send replacement cells by express but its still a lot of work for nothing but repairs! also there was SO much hot glue melt in this cell, spent hours with a hair dryer digging it out with a butter knife... but the cool thing is the pack was folded in half and now un-folds to a nice long shape now which i think will mount more subtly on to my bike frame!

 

[k-harvey]

the2.7 v rule didn't work for me BMS kept shutting b pack down after a few k's replaced the 2 cells that got down to 2.8+ & other was higher if I remember then it was ok.