Strange BMS issue

[eikido]

I have a strange BMS issue.

I change a bad BMS with a new one. The bad BMS outputs correct voltage, but as soon as it is loaded, it drops the voltage.

The new one outputs 0.45v so something is bad with it.

I change to another new BMS. I test the new one before soldering it and it outputs correct voltage.

I solder everything and after, it starts behaving strangely. I have to disconnect and connect the load several times until I get proper voltage. I can load the battery pack (4A), but only sometimes.



I take the previous new BMS I changed first time and try it on another battery pack, and viola! It works!



I’ve checked the balancing wires many many times, and they are correct. I measure the voltage directly on the BMS connector with my super thin probes.

All cell groups are balanced.

ATM, they are all around 4v.

I checked the increment, and all add upp correctly.

4,8,12,16…..v

The polarity is correct.

I’m suspecting bad cell wires, but they all measure correctly on the connector.

What can it be?

 

[eikido]

Turns out it was a bad balancing cable.
Unbelievable, I tested the cable a million times. Check continuity, resistance, voltage. Etc.
Don't understand what's bad with it.

 

[amberwolf]

Was continuity tested while it was plugged into the BMS and attached to the cells, by placing probes on the cell and the BMS solder joint for that pin?


If not, then it could be spread pins of the contact so that you can read correctly touching the probe to the metal, but when the contact is slid over the pin of the BMS board connector it doesn't touch, or has very high resistance.

Another thing that can happen with certain types of connectors is the contact can push back out of the shell just enough during connection to make no or poor contact. This you can see visually while it's connnected, if it's an open-frame shell. (the other problme you can't see directly like this).

 

[eikido]

The connector was glued on top which was difficult to break so i don't think it could slide back. So my theory now is that the pins were not crimped correctly.
Thanks.

 

[pishta]

Wait, I had the same thing happen to me! I had a xiomie type 36V 10S3P pack but it had a BMS board soldered to the top of the entire pack with its pick up points folded onto by the nickle strips and/or jumper wires going up the sides to the boards contacts. I was getting 41V at the B- and B10+ points AND 41V at the P+ and P- (pack?) contacts but when I powered up the scooter, the output voltage would drop to 2V AND SLOWLY climb to about 26V and the controller would throw a accelerator code due to low voltage on its POST. There was a tiny resistor that I could jump and it would give me back almost full voltage to the P+ side but I didnt think that was safe to solder bridge. I found a new BMS for it and its in the mail. I have a second BMS issue in that this other "cylinder" 10S2P pack shows 37V at the B-B10+ points but 19V at the pack out (discharge) points when its all assembled. It has been sitting for a year and luckily the guy was trying to charge it with a 12V wall wort charger and sold it to me for $20 complaining it would not charge. I took the pack apart and found 9 of 10 groups at 3.51v and 1 group at 3.1v (?) which I found odd as these were all in series. I took the 2 low cells out and charged them back to 3.5V and put the pack back together (weird modular stackable 4 cell holders, looks like a baseball bat) and now my other 36V charger is charging the entire thing. Ill let it charge for another 3 hours. and pull the plug. If it stays on, Im good. If it shuts down, Im losing the BMS and just going straight 10S2P @ 41V. Nothing on either BMS looks cooked.

 

[amberwolf]

At least for the second pack, your problem isn't the BMS--it's just doing it's job. The low voltage you see is perfectly normal; that's just a consequence of how the FETs leak a tiny bit of current even when off. The voltage drops to "nothing" when a load is placed on it.

If the BMS FETs are damaged, you could have a secondary problem there, but that's unlikely, and the main problem is the cells that were low. They are significantly different capability (capacity, resistance, etc) than the others, and the BMS was probalby trying to protect against a potential fire, because when there is that big a difference between cells, it can mean the different cells have a problem that could be damage (you can't know if there is potential for catastrophic failure until it actually happens, so "overprotectiveness" is better than the alternative).

Some BMS will prevent charge/discharge if there is a difference in voltage beyond a certain point, and some will only do it if the cells are above or below a certain voltage range. Since 3.1v is usually above that minimum voltage, but there's nearly half a volt difference between the readings you had on the majority vs minority, it's very likely your BMS is one of the better-designed ones that will protect against that difference.


Taking the BMS off just removes your protection against overdischarge and overcharge, and increases the likelihood of a fire at some point.


The first pack there isn't enough info to diagnose remotely, but the second type of problem we've seen numerous times.