How do I wire this BMS?

[tony67]

I received this http://eclipsebikes.com/48v-bms-16-cell-60a-p-1042.html BMS today. I thought it was fairly straight forward to wire but I think I may have blown it. On the board are the following connection points and what I think they mean.

P- Power neg out to controller
B- Battery neg into BMS
C- Charger neg into BMS
P+B+C+ Common connector for the pos leads to the controller/charger/battery pos.

There is a 15 pin connector for what are the balance leads I assume these go on the 15 series link plates between cells but i've no idea what order.

My battery is a 16s 1p 10ah headway pack so what I did was connect my battery neg to connector B- battery pos to B+C+P+ terminal. I measured the voltage at P-B+C+P+ and got 55.5. I then connected the 15 tails to the 15 plates in sequence based on my pack order. When I plugged the 15 pin connector into the board the first and last resistor glowed red hot, i unplugged it within seconds.
Now with the battery pos and neg still connected to B-/ B+C+P+ when I measure the voltage at P+P- im getting 44.5v but measured at the battery its still 55.5. I dont know if its a coincidence or what but if I divide 55.5 by the number of fets 7 and multiply by 6 I get 44.5. Is it possible a fet is dead? Would it cause my voltage drop if it was? Did i mess up somewhere? The resistors all still seem to work and have the same measured value despite visual damage. I only noticed afterwards the second fet is marked (T470 BN1C3N). The other 6 are marked (T470 BV0Y2E) dont know if thats significant.

Whole board showing damaged resistors
http://imageshack.us/photo/my-images/545/photoau.jpg/

Fets
http://imageshack.us/photo/my-images/684/photo2dt.jpg/

Shows second fet is different to the other 6
http://imageshack.us/photo/my-images/221/photo1vtm.jpg/

I cant seem to find any info on the site or indeed the net about wiring a BMS. I hope someone can offer suggestions or its €50 down the drain.

 

[wesnewell]

Go to the website you posted and click on the installation diagram.

 

[tony67]

Thanks for the reply. That's the diagram I followed but it doesn't seem to have worked. I'm gonna give it another go during the week.

 

[amberwolf]

It sounds like:

--your cell taps are not wired in the order you think they are (easy if they are all the same color, which is common),

--or you hooked them up in completely reverse order (backwards connector on your harness, for instance).

--or pack + and - got mixed up (not likely if you use different colors).

--or the instructions are wrong (whcih doesn't seem likely, as the diagram makes sense from a component-layout perspective).

Any of those things could have caused a reversed connection that could cause what amounts to a short thru variosu components on teh BMS.

If something was wired reversed, anything along the current path could be fried; you'd have to test each component to find out. It could be the switching transistors for the balance circuits (since obviously quite a bit of current went thru them if resistors glowed), or the main charge or discharge FETs, since they could have been reverse-biased and essentially shorted the pack (or at least some cell(s)) thru the body diode(s).

 

[tony67]

I think you hit the mark there. I was looking at the diagram again and thinking back on what I done and I remember thinking I'll start with the positive side of the battery when hooking up the balance leads. From what I can get from the diagram I was supposed to start from the neg side. This means I wired the balance leads in reverse order This beIng the case do you think it's safe to say this board is fried? There are hundreds of tiny components that I don't have the skill to test. If it was a fried fet I'd give it a go at replacing it.
The voltage drop I reported in the opening post is exactly 1/7th of the pack voltage. Could this point to a fried fet or am I clutching at straws.

 

[amberwolf]

Most likely all the large FETs are the same and wired in parallel. The one that's differnet is probably just a different date/batch code, and the top number that's the same on all is probably the p/n (can't find any fo them in my part searches). Those FETs you can probably test all at once. but they are probably still good.


The balance channels...one or more of them is probably toast, most likely the ones that had the hot resistors, especially. But if you hook up the pack like it should be, including balance connectors, and the voltages at each cell don't read abnormally, you could then test the balancing functions and see if they still work.

You may be able to use two adjustable current-limited power supplies to test it, if you don't want to risk the pack. Leave all the pack connections off the BMS. Set one PSU for the full pack voltage, and put it across the main battery leads of the BMS. Set the other for a single cell's just-fully-charged voltage (but below the voltage at which that BMS starts balancing). connect that across one balance connector pin pair at a time, starting at the most negative. Keep an eye on the current meter for the PSU, and it's voltage. If either varies when connected, then that balance channel is drawing current, when it should not, and probably needs further diagnosis. if it does not vary, then go to the next channel.

if all channels work ok at a static fully-charged voltage, then bring up the PSU's voltage to that of a fully-charged-needs-bleeding voltage, so the balancers would then be draining current from the cell. Retest all teh channels just like the first time, noting any that fail to draw current at that voltage. Any that dont' draw current need to be fixed.

While doing this test, also check the charging terminals of the BMS for pack voltage to be present. It should cut off the voltage there if the voltage for any cell goes above the HVC point, wich should be at or below the bleeding voltage point.


If all pass those tests, then set the PSU to just below the LVC for each cell. Test just like you did for HVC except now the test is just to see if there is voltage present on the outut terminals of teh BMS. it should cut off as soon as any cell goes below LVC.


If everythign works like it should, then the BMS might be fine. But I doubt that it is. If you want to be sure without all teh testing, you'd be better off replacing it, but if you don't mind the testing there is no reason to just chuck the BMS without finding out how hard a repair job (if one is needed) will be.

 

[tony67]

Thanks for that comprehensive answer AW. The procedure you describe is beyond my ability due to both lack of knowledge and lack of equipment. I intend to try wire it up again the correct way maybe at the weekend. I'll take whatever readings i can and see what happens. Unfortunately i think this board is toast. If nothing else I hope this experience helps another novice avoid my pitfall. Maybe I can salvage one of the resistors for the precharge circuit I'm currently missing. Anyway thanks for the advice lads, guess I'm goin BMSless for a while longer.

 

[amberwolf]

Well, you can use batteries to do the test, too, but for myself I'd rather not risk the cells since I do have the adjustable power supplies and typically I can't afford to buy new cells if I kill existing ones.

So I try not to do my potentially destructive tests on things I can't fix myself

 

[dnmun]

yep, if he reversed polarity, which we still don't know from the info yet, the shunt transistors may have been damaged and could all be turned on so they would just drain the cells when hooked up so he needs to watch the voltages on the cells after he hooks it up again and not leave the sense wire cable connected in between times.