CellMan A123 Teardown

[llile]

Paul (cellman) at EM3ev.com has helped me diagnose that I have a bad BMS. Although he provides excellent products and helpful service, they aren't indestructible. I thought it might be useful to describe the teardown here in case someone wants to see how the BMS is connected.

Note that this is an older battery, and the company indicates they use much more sophisticated methods on their new batteries.

The setup is a 40 AH 72V A123 battery, split into two packs, with dual motor controllers on dual front motors. Cycle Analyst limits total current to 35 amps. BMS is rated at 80 amps.

How do we know it is a bad BMS? Been working fine, then one day cut out on low voltage, an inconvenient 5 miles from home, when it should have had plenty of charge. Towed it home on a trailer, and started to look into the problem. Normal charged voltage is 83.1V. I read 83.1V at the battery terminals after a night's charge, but as soon as I plug it in, the cycle analyst (and the voltmeter) reads 68 volts. As soon as I touch the throttle with the wheels up on blocks to spin freely, the cycle analyst says the voltage goes down to 50V and cuts out on low voltage.

Checking cell voltages at the BMS, I see between 3.4 and 3.5 volts on every cell at the BMS connector. That says cells are probably OK. However when I try to load the battery, pack voltage drops to half or a quarter of normal, although I know from cell voltages I have a fully charged pack. This would occur if the BMS has a few bad FETs.

Tearing into the Cycle Analyst connectors, I find one connector has a broken wire, which happens to carry full battery voltage, and could have shorted against another wire, killing the BMS. Could have killed a motor controller too, don't know yet.

So here are a few rules about working with batteries that are obvious but I didn't know before ruining one:
1. never have more than one bare conductor at a time. As soon as a plug or wire or other conductor is bare, cover it with tape. Because if you have two bare conductors, Murphy's Law of Conductors says they will touch and go bang! Then your BMS (at a minimum) could be blown. I learnt this on a PING battery, killing the BMS. A big chunk of a FET was simply gone.
2. Wear those safety glasses. When things go bang in your face you want some specs.
3. Use insulated tools when possible. 83V can deliver quite a zap along with a powerful spark.

 

[llile]

So I've peeled back the shrink tube, some Kaptan Tape, and some actual duct tape to reveal the BMS. Pics linked here:

https://www.dropbox.com/sh/vispamwvdjxjjg7/AAC4eucvVX1zWwuzMAisUwvla?dl=0


I'm liking this split pack for maintenance. Once split, there is a lot less chance of things shorting together - there is actually no connection between the positive and negative output terminals since one of them connects directly to the second pack. Dnmun has warned us that the connection between split packs should be robust. Although he likes a soldered connection, this one uses a bullet connector for the power connection between packs, and that has never been a problem. The way I route the wires and protect them, there is very little way that these connections could come loose. I'm fairly warned.

There is some stuff glueing down the BMS board that looks like silicon or hot glue. I verified by melting a chunk with the soldering iron it is hot glue. So, should I use a heat gun to get the hot glued BMS off, or will that cook everything and cause trouble? Or would a good soldering iron loosen this bead of hot glue enough to pull off the board? Perhaps the first step should be to remove the negative wire from the BMS terminal, and insulate it?


View attachment 2

 

[amberwolf]

Hot glue also "fails" when you get it colder, as it becomes fragile. That may be safer than heat. How cold it needs to be you might have to determine experimentally.

 

[llile]

Hot glue also "fails" when you get it colder, as it becomes fragile. That may be safer than heat. How cold it needs to be you might have to determine experimentally.

Wow! I'll look into this. Maybe get a can of freeze spray.

 

[llile]

I was nervous about this, but it was a lot easier than it looked.

First, I insulated the entire heat sink, in case a de-soldered negative lead shorted against it. Although it looks like the heat sink is electrically isolated, I didn't want to take any chances. Removed and insulated the negative wires one at a time.

A cool thing about the split pack: The negative lead is literally the only electrical connection to the battery, once the split pack is disconnected. This wasn't obvious at first.

The hot glue was a lot easier to remove than I thought. The soldering gun plowed through it like butter, then the entire BMS could be peeled off carefully. Voila! Ready for a new BMS.