No longer able to charge. Charger or BMS?

Got back from work today and my 3 year-old cheapo 48v lifepo4 charger wouldn't start charging the battery. It is still putting out 60v. The charging FET on the BMS for my Ping battery melted off the plastic a few weeks ago, but it's not nearly as melted as my friend's whose Pings both did that when he first got them and those have been working for years like that. The BMS is still putting out 52v to the charging connection. Is there a way to determine whether it's the BMS's or the charger's fault that it won't charge? I don't have access to any other 48v batteries or chargers to determine it the easy way, so any help is appreciated. Also any advice for how to potentially get the battery charged in the mean time would be great; this weekend is probably gonna be the warmest weather we'll see for months. So the timing is perfect as usual... why couldn't this have happened in that nasty November we had?

you can test the charging mosfet with the diode tester on your DVM. if the mosfet is getting hot, it is shorted. mosfets do not usually get hot when functional.

Hey Dunum,what fet should he use to replace it?

Dan

dnmun said:

you can test the charging mosfet with the diode tester on your DVM. if the mosfet is getting hot, it is shorted. mosfets do not usually get hot when functional.

Click to expand...

Yeah, thinking about it, that seems to make a lot of sense. Strange thought that my friend's got so hot the plastic on the mosfet melted and it's still functional, but mine only melted the shrink wrap and still worked for a few weeks but is now maybe dead? Also strange is that my battery got a full charge the night before and wouldn't even charge for a second when I got home, so essentially what ever broke, broke while not in use. But I hope it's the mosfet, I can probably grab one from another BMS and then be good to go for the lovely weekend weather! A boy can dream...

if the battery had been balanced already before the charging mosfet shorted, then it would continue to charge up and never go to HVC because the pack was balanced, so the fact that mosfet is shorted does not mean the pack will overcharge.

this is a common failure of the ping BMS. i think the problem is that the opto transistors that the gate voltage is produced to the charging mosfet through are only partially turned on that causes the mosfet to operate in the transition so they overheat with the charging current. i do not know the part number for the mosfet.

also the charger will push large amounts of current just as the charging mosfet turns on and off so the high current during this period when the mosfet does not conduct well is how they die imo. DrBass used the 4110s for the charging mosfet and put a heat sink on his too as i recall.

Welp, it's definitely the BMS. I have a old BMS that was DOA when the battery was shipped to my friend. Think had the same problem of not charging, but I was thinking the discharge mosfets on it are probably the same and I'll just harvest one of those to replace this failed charging mosfet. Now to see if I can get any of them off in one piece...

Btw, if something doesn't work out, it's fine to bypass the BMS and connect the charger directly to the battery to get a charge, no?

Hmmm... the heat caused the surface of the circuit board to come detached from the backing, so it came off with the mosfet. Seems like I need a whole new BMS, no?

if your battery is already well balanced from regular charging to the point where the voltage was high enuff to allow the BMS balancing circuits to function, then you can continue charging directly. do not remove the BMS, just use a jumper wire from the C- spot over to the drain of the discharge mosfets. then it will still balance the cells but if the pack is not currently balanced you will not have cell level HVC protection until you replace the charging mosfet.

if you wanna replace that charging mosfet you can solder the new mosfet to the drains for the discharge mosfets and then connect the gate drive from where the original charging mosfet gate trace was with a small (even tiny)wire, and then connect the source leg of the new mosfet to the C- terminal of the BMS where the negative charging lead attaches with another small (large enuff for charger current) wire.

Great success! I ended up soldering the gate and drain legs into their usual circuit board holes and building a fat bridge of solder between the mosfet drains. Thanks for the guidance!

i always would put a piece of 12AWG bare copper wire across all 5 mosfets on those little BMSs. soldered to the ends of the tabs and then soldered the P- lead onto the copper wire.

do you understand how the charger mosfet gate voltage travels through all of those opto transistors down to the gate for the charging mosfet?

i have found that sometimes the opto transistor is not fully turned on because the led inside the opto transistor doesn't turn the transistor on strongly enuff, so the voltage gets 'hung up' and that reduces the voltage on the charging mosfet gate.

the cause: i think the led inside the opto just doesn't work well when the cell voltage drops really low down to 2V. it seems that maybe the current limiting resistor for the opto transistor led may be too large in value and limits the current so much when the voltage of the cell is low that the opto transistor does not turn on.

you can measure the voltage drop across each transistor to see if they are all low and if you find one higher than the others, that could be where the problem is. you can put the red probe on the collector of the top opto, then walk down the others one by one from there and you should see the voltage drop about 1050mV for each one as i recall, but don't quote me on that. the voltage on your meter will be increasing since you are walking the black probe down from the top while the red probe is on that top collector.