how to diagnose a battery pack/bms problem

toaroa

100 mW
Joined
Jun 9, 2010
Messages
37
Hi. I'm hoping some of you kind experts could tell me the steps I should go through to diagnose a battery problem. Pack is:

Headway 12Ah 12s1p with a capacity-based bms (pack made by evassemble.com). The pack worked well from day one and has gone through ~20 cycles, usually never below 60-80% capacity. My commute usually takes 4Ah from it. Although i have been careful to keep it well supported and upright, when taking it out of my battery/trunk bag, I have noticed a rattle and I suspect that the bms might not have been stuck down properly to the battery housing (That would fit with a few other half-arsed build problems from evassemble, like using what seems to be second-hand wire for the leads!)

After return from a ride, I put it on to charge (4A charger designed to cut off at 43.7v). At the end of the charge, I plugged in my watt meter and it read 41.7v. It's usually around that mark, so I presume that the bms cut off works before the charger's does. Once or twice I've got it to 42v but nothing higher. I plugged it in to the controller this morning but... nothing. Then I tried with the watt meter and nada. I suspected my crimping, but usually I can get a voltage reading off the in/charge line as well as the out/discharge line so tried the watt meter on there. Nothing. This suggests it wasn't one of my crimps and that it's something between cells and leads. Oddly, after removing the anderson housings and attaching the contacts to a volt meter I got a reading of 41.3v (there's often a slight difference between what my watt meter and volt meter state). The value was fluctuating but that might be because I never remember the right setting to use on the volt meter for DC current. I also don't understand the measurement of amps using a voltmeter, but turned it to 'A' and took a reading at 16.x. I was expecting 0 because it wasn't under load. Perhaps it was mA and I'll need to do test that again. However, I guess the pack could have overcharged and the bms might have tripped????

I've never had the battery pack out of its shrink wrap, so I'd appreciate if you could share the basic steps for diagnosing what the problem might be. Presumably testing each cell would help, along with (somewhow) testing voltage into and out of the bms to see if it's a bms problem?

A 'how to' diagnose a battery pack might be quite a useful thread for other relative noobs too, as the ES info on that seems to be inconsistent and spread across multiple threads. Cheers.
 
when you switched your voltmeter to A, it shoulda made a loud popping sound and a lotta smoke shoulda come up from behind the screen. i doubt if it works anymore.

sounds like you had voltage at the anderson powerpoles, but the plastic housing is not seated properly and so the powerpoles don't make good contact inside?

anybody interested in splitting a big bag of powerpoles?

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=350363931564&ssPageName=STRK:MEWAX:IT#ht_3991wt_932
 
Poorly housed connectors was my first assumption but, as stated in the original post, both my 'IN' leads (for charging the battery) and the 'OUT' leads (for connecting to the controller) previously could be used to give a voltage reading - connecting my turnigy watt meter to either of those would give the same result.

As neither set of leads will now provide a reading on the watt meter and the 'out' leads won't power the controller, doesn't that suggest that something more fundamental is amiss?

Why I got readings on the voltmeter has me beat but they were quite unstable (Perhaps simply because I had it on the AC rather than DC setting?).

It all beats me. Will sit down with a library book about using voltmeters - clearly I need to sort that out before I go much further.
 
You can buy a digatal meter for 1.99 on sale at harbour frieht with a battery. Put the meter on dc 50v or higher for 36v. Or just a higher setting then what you are testing. Do you have any pics.
 
you should never put the voltmeter on the amp scale or on the resistance scale unless you are measuring current through the meter or measuring the resistance of a passive device like a resistor.

when you put the ammeter scale across the voltage source, it is like a direct short through the meter so it usually blows up. voice of experience here. been there done that, 5-6 times so far.

but without more to go on, not much more we can help you with.

yes, you cannot measure DC voltage on the AC scale.
 
Thanks guys. I didn't blow the voltmeter because there was no voltage going through the leads =>

I took off the multiple layers of tape and shrink wrap and the problem was obvious: the plug which connects into the bms and carries the narrow gauge wires from each battery had dislodged. Plugged it back in and all working fine again.

While I had the covers off, I checked the voltage of each cell. My pack is in a 3x4 array, with the bms stuck to a row of three cells. If I number the three furtherest from the bms as 1, 2, 3 etc (so the bms is stuck to 10, 11, 12), I measured...

3.294
3.292
3.400
3.293
3.296
3.389
3.294
3.301
3.373
3.294
3.311
3.350

That doesn't seem as balanced as I expected after ~20 charges. It is a capacity-based bms so it won't behave like the majority bleeder BMSs but ... what do you think? Is it working ok? (NB. pack wasn't at maximum charge)

Thanks for your advice. I still reckon a newbies' guide to pack diagnosis would be a good idea. I found some useful tips about multimeters here: http://www.masonmonitoring.com/pdf/support/UsingMultimeter.pdf
 
Back
Top