JBD AP20S006 possible static electricity damage?

[Jon_And_Kate]

I had made a plexiglass (acrylic) cover for a 16s 280AH EVE battery bank. All in a line, about 46" long. It makes sense for the space, but it means the wire lengths are barely able to reach the far terminals, and the BMS needs to sit about in the middle, just above the cells.

I bent the plexiglass sides so that it's resting on top of the cells, containing everything, and I set the BMS on top of that for initial startup, not grounded. It was working fine, showing all 16 voltages with no issues for a day in the overkill app, and in the remote LCD interface. Then yesterday, while it was still not grounded and working fine, I bored a hole into the acrylic cover a bit wider so that the positive lead can stick through the cover easier, about 2 feet away from the BMS on the positive side of the battery bank. Then not long after, the BMS started beeping, showing cells 13 and 14 at 0mV and 5000mV respectively for a second, then they would swap for the next second, and repeat.

I unplugged the harnesses from the BMS to power down, hooked up a grounding wire to the BMS. Then measured cells 13 and 14 at the cells and on the harness with no issues. I started it up again, and it showed voltages correctly again. Then about 5 minutes later, the same issue came back with cells 13 and 14 flipping back and forth, this time with the unit grounded. I haven't grounded the negative pole of the battery yet, but the BMS is grounded. I suppose I'll try that next.

Has anyone else had this issue? Can it be fixed, or did I fry it?

 

[kommando]

If you have fried it then you would be seeing nothing and certainly not 5 minutes of good running. You need to check these wires for 100% I would replace them checking all the connections. If there is a fault on the board it will be a component that is probably taking 5 mins to heat up and then stop working that is common to cells 13 and 14. So checking the board for components that are too hot after 5 mins should narrow it down.

 

[justgary]

I also vote wire connections. Tug on each connection at the cells and make sure that they are all tight.

 

[Jon_And_Kate]

Yes, I did that before I posted. I also loosened and re-torqued the terminal nuts. I also checked all the wire harness voltages with a multimeter before plugging back in to power it up again. I don't know how I would replace the wire besides cutting it near the harness and splicing in a new one, but I haven't done that because I would likely change the resistance of the wire, which might make a difference at the millivolt precision of the BMS.

I'm waiting on a response from JBD at the moment. I've left the unit unplugged now for a few days, and hopefully I'll have time to check it again tomorrow. I'll make sure to post an update when I have one. Thanks for the thoughts folks, I appreciate them! If you have any crazy ideas, please let me know.

 

[justgary]

I don't think that the balance current is large enough on a JBD that wire length is overly critical. I cut mine to length.

 

[Jon_And_Kate]

I don't think that the balance current is large enough on a JBD that wire length is overly critical. I cut mine to length.

hmm, but what about the voltage measurement? You're probably right, since voltage measurement is done with extremely high impedance of say around 10 Mega Ohms, thus requiring very low current to the point that wire resistance will only affect the measurement at around .1 micro volts. I guess I just had to think that through. Ok, I'll try redoing the wires.

 

[Jon_And_Kate]

Ok all, problem fixed!

The problem is somewhere between my chair and my keyboard. Turns out you were right! I just have kid gloves with this stuff still. I re-crimped the cells 13 and 14 connectors, and it started up with errors on cells 8 and 9. Then I went through all wires, re-crimping them all, and it works fine.

So, this whole thread is me putting my foot in my mouth. But It's a good example for others.

 

[Quattrohead]

I like to keep the wires the same length and solder them as well as crimp or instead of.