Bad Battery or Bad Cell

[Adamsavage79]

I am once again with a problem. It seems I have plenty of them, way more than than I should. This will be my SECOND battery I've had go on me, and with the EXACT same problem. The first one was brand new out of the box. Soon as I sit on the bike and use the throttle. Bam, the power is cut. If I left the back end of the bike up, and use the throttle the motor will spin just fine, and this because there is much less power being drawn from the battery. The battery is reading 51.6V so It's not a low battery problem. Non less, I'm going to give it a full charge and see if that does anything. I'm highly sceptical this will do anything, but I have nothing to lose as I refuse to soak any substantial money into this bike at this point. However, the bike was working just fine earlier today and yesterday. So my 2nd theory is that I have a weak cell in the pack and once it reaches a certain voltage, it drags the entire pack down. Wouldn't it still go to below 48v and then die, vs dying at 51.6v ?

EDIT: Put it on charger for a hour, and tried again. This time the battery didn't cut. Will charge it overnight and see if it's the same. If it is, then I have a reduced range. It seems I have a cell that is weak ? Not sure why the pack is cutting out @ 51.6v. Never seen the behavior before.

Also want to add I got it up to speed going downhill I was able get the motor going. However, once the road flattened out and the power increased, the pack died again.. Not sure what is going on here. Is this the BMS or a weak cell ??

[Bikemad]

As you haven't mentioned which battery you have it is difficult to say what is causing the cutting out.

I suspect the battery's BMS is not able to supply the 25-30 Amps continuous that the Magic Pie will try to draw under full load, and it is simply disconnecting the power to prevent damage to the battery.

A Magic Pie needs a battery that is capable of delivering a continuous 1500 Watts power output, but many 48V ebike batteries are only rated for 500-1000 Watts.

If a battery's BMS was only rated for 20 Amps continuous (1000 Watts), you would need to set the maximum current setting on the Magic Pie's controller to 20A (or lower) to prevent the BMS from being overloaded and cutting out under sustained full power.

I recently bought a BMS stated as being 40A continuous and 50A max, but the BMS that arrived is marked 30A continuous and 50A max:



Unfortunately, the battery it was to be installed on will be used with a modified Magic Pie that regularly draws up to 45A when working hard, and 30A continuous is not sufficient for my needs.

Purchasing items online can be very frustrating when the descriptions don't match the item being supplied, and battery capacities and power outputs etc. are regularly exaggerated just to make a sale.
I bought two non-genuine DeWalt style battery packs which were supposed to be 6Ah capacity:



When I looked inside, I found that each pack contained ten 2,500mAh cells in a 5S2P configuration, which should have produced a useable capacity of 5Ah - not the stated 6Ah.



I fully charge one of them to 4.1V per cell using the DeWalt charger and then discharged it at 1.5~2A until the lowest of the cells reached 2.5V, and it only produced a capacity of 3.703Ah.
I then fully charged the pack all the way up to to 4.2V per cell using my icharger and the total amount put back was just 4.032Ah!

If I had purchased 4Ah packs, I would have been very happy with these results, but it's very frustrating when you buy two "6Ah packs" which only have 2/3rds of their stated capacity.



The above "18650 High capacity 48V battery" is described as having a capacity of 30,000mAh (30Ah) but it is constructed from 39 18650 cells with 3 parallelled groups of 13 cells connected in series (13S3P).
I've yet to see an 18650 cell with a genuine capacity of more than 3500mAh (3.5Ah) so it's very doubtful that this battery contains 18650 cells with a true capacity of 10,000mAh (10Ah) each. 

Alan
 

[Adamsavage79]

I charged it for several hours, and then let it sit on charger for 3 hours (smart plug timer). I let the voltage settle overnight and sat on bike and the battery had no problem this time. It would seem that 51.6v is the now cut off point for the battery. If it goes past this, then the pack will not work under any sort of real load. So everything you mentioned above makes sense, but it doesn't. It doesn't explain why bringing the packs voltage up solves the problem.

EDIT: I took it out for a test ride of about 11.53km and tried to keep it under 500 watts most of the way. The last 1km coming back home I pushed it to the max for about 500 meters. Still had plenty of juice left after that. This will allow me to get to work and back without worrying about the pack dying on me. It still doesn't answer why the pack was dying at a higher voltage.

[Adamsavage79]

I can confirm this is a now DEAD cell. This one cell was discharging faster than the rest, dragging the pack down. Today it died at around 52.6v roughly give or take .10 of a volt. Kept riding it to work. Left it sit outside for my shift. Checked the voltage again, and just sitting there with battery disconnected from the bike. It now read 44v. It was at this point, I knew there was definitely something going on with the pack. So I cracked it open and un screwed a few cells. Then started to check them. I got to the 2nd cell and I got a reading of 1.5v while the rest of the cells are giving me 3.2v - 3.1v, which would be a normal reading for them. 1.5v however, is NOT a normal reading. This means I have dead or dying cell.

[Bikemad]

That certainly does appear to be a faulty cell, and it would definitely cause the BMS to cut the power when that single cell voltage fell below the minimum safe cell voltage setting in the BMS (probably ~2.5V).

Presumably it is a LiFePO₄ battery pack with 16 cells in series?
However, your voltage readings are still slightly confusing, as 15 cells @ 3.1~3.2V and 1 cell at 1.5V should give a total pack voltage of between 48~49.5V, which is a lot higher than your 44V reading.

I guess that the additional voltage drop could be across the BMS if it has partially disconnected the power.

Hopefully, the pack is still under warranty, but if it's not, you may be able to buy a single replacement cell from your battery supplier to replace the failed one.



Alan
 

[Adamsavage79]

Yeah the pack goes all strange with the bad cell connected. Without the cell I get 50v. With the cell I get 44v, which make no sense really. The cell is now reading 2.0v, but that's still a bad cell. I'm looking into replace the single cell. So far I've only known the one site and it's gonna cost me $122 USD due to the shipping and battery handling fee. However, I want the bike to work and be able to get a higher resale value for it. The cells are from 2018, so while I would be mixing, it's only one cell. It shouldn't be too much of an issue.

[Adamsavage79]

Well I was wrong about the pricing. There was a big GOTCHA at the last stage of checkout. They wanted $105 USD shipping fee. Combined with the handling fee, and cost of the cell and converting US to CDN. I said nope. It's not worth $268 for a single cell. I would be lucky to get $500-$400 for the pack. Maybe less, as it is two years old. It's simply too risk. I rather sell the entire kit with the pack, and they can decide if they want to replace the cell. I will call around local ebike shops and see if somehow they can find me the cell, but it's unlikely.

Also, would running the bike hard at max throttle little over half the time have an effect on the battery ?

EDIT: Now, after being removed from the pack, the cell has gone back up in voltage over the last few days. It's now reading @ 2.36v and the pack is @ 52.5v. This doesn't make sense to me, that this cell would gain voltage over time.

[Bikemad]

They wanted $105 USD shipping fee. Combined with the handling fee, and cost of the cell and converting US to CDN. I said nope. It's not worth $268 for a single cell.

That is definitely an extortionate price! 

I had a look on AliExpress and it seems that the correct size 10Ah cells are impossible to find with threaded terminals.

I found several which were 75mm wide (too wide to fit the 48V 10Ah GM rack pack) but the ones I could find in the correct size did not have the threaded terminals and would probably require a spot welder to install.

I also found some with additional lugs, but I don't know if they would be too long including the additional nuts and bolts to attach the connectors. 

It might be worth contacting one of these manufacturers to see if they could supply cells with the threaded terminals. 

Also, would running the bike hard at max throttle little over half the time have an effect on the battery ?

The more load you put on a battery, the shorter its lifespan is going to be, but the battery's BMS should prevent the cells from being overloaded beyond their rated output.

Now, after being removed from the pack, the cell has gone back up in voltage over the last few days. It's now reading @ 2.36v and the pack is @ 52.5v. This doesn't make sense to me, that this cell would gain voltage over time.

If a weak cell has been pulled down to its lowest permissible operating voltage, it will typically increase in voltage after the load has been removed, but it might be worth connecting it back up again to see if its voltage starts to drop again, as a noticeable voltage drop could be indicative of a faulty BMS incorrectly discharging the cell slowly through the balancing resistors.

Alan
 

[Adamsavage79]

I put the bad cell back in the pack and in about 2-3 hours, the cell dropped down from 2.45v to 1.5v. Soon as I removed the cell from the pack, it started to quickly regain voltage. Also, mypack is a 15ah pack. I came across one site, but it seems sketchy as they were offering $40 shipping from China via UPS in 3-5 working days and non mandatory handing fee, due to it being a battery.

Here is the website in question: https://bmsbattery.com/ebike-battery/392-headway-40152s-15ah-10c-lifepo4-cylindrical-battery-cell-battery.html

[Bikemad]

Can you swap it with one of the other cells and charge the battery and make sure the swapped cells are both reaching ~3.65V at the end of the charge.

Then check to see if the voltage of the suspect cell drops, or whether the swapped cell voltage drops instead, as it might be the BMS that is faulty and the cell is OK. 

You shouldn't encounter extortionate shipping charges purchasing a replacement BMS unit. 

I did find one seller offering those 40152S Headway cells on AliExpress for C$ 25.21 and the standard shipping charge was C$ 16.14.
However, the DHL shipping option was considerably higher:


I would simply order it via standard shipping and see what happens, but only after you have established whether it is a faulty cell or a faulty BMS.

Alan
 

[Adamsavage79]

I have an individual cell charger. I will try to bring cell up to match the rest of the cells. Then put it back into the pack, see if the voltage drops again. As for swapping a cell, I don't think that will matter. This is based on how the pack is made. Here is what the pack looks like. I also attached pictures of the pack without the cell in question.

[Bikemad]

As for swapping a cell, I don't think that will matter.

That really depends on whether you want to locate the cause of the problem.

If you put it back in the same position and its voltage starts to drop, you still won't know whether it is a faulty cell or a faulty BMS.
However, if you swap it with an adjacent cell, you should be able to determine whether it is a faulty cell or a faulty BMS by seeing which of the following observations apply:

• If the voltage of the swapped cell starts to drop, but the voltage of the suspect cell doesn't, then this would indicate that you probably have a faulty BMS that is incorrectly discharging the cell through the BMS via the balance wires.
• If the voltage of the suspect cell starts to drop in its new position, and the voltage of the swapped cell doesn't, then this would indicate that the problem is within the cell itself and not a fault with the BMS.

Alan
 

[Adamsavage79]

Ok I moved the bad cell one over. Basically swapped spots with another cell.

[Adamsavage79]

Update: Moving the cell had no effect. I seen no change in voltage in the pack at all. However, this was after applying a small charge to the cell. The cell is able to hold a charge, and I was able to bring the charge up to 3.23v and then it seemed like I couldn't get anymore charge into it with the charger I was using. I've put in back into the pack, but still in the swapped position and not it's original position. I will let the pack sit over night and see if there is any change. Then I will charge the entire pack with the cell in it. Wrap up the pack, and see if the pack will work under load. If it does, then see at what point the cell cuts off. This could be a weak cell, and it couldn't handle being cold ? I'm not sure. The cell is acting odd. Going up up in voltage by itself without being connected to the pack, is bizarre.

Assuming nothing has changed, then a simple charge to the pack/cell should allow it to handle load.

[Bikemad]

If your suspect cell has a resting voltage of 3.23V, it is probably only 20~25% charged:


How much charging current does your individual cell charger output? if it is 500mA or less, it would have taken 30 hours or more to bring an empty 15Ah cell back up to its fully charged voltage of 3.6~3.65V.

15Ah/0.5A = 30 hours (1.25 Days)

If the other cells in your pack are already almost fully charged, it could take several days to fully charge the suspect cell, as its charging current would be limited by the balance current of the BMS, which is often less than 100mA (0.1A).

15Ah/0.1A = 150 hours (6.25 days)

If your cell is already 25% charged, it should still take over 4.5 days to charge it the remaining 75% with 100mA balance current, over 9 days @ 50mA current, and 15 days @ 30mA current! 

It will be interesting to see what the final outcome is, but I'm placing my bet on a faulty BMS.

Alan