BMS Battery Alloy Shell Charger Problems

I have now experienced another issue. Charging works fine with both chargers. The fan is off until the FETs have some 50oC. Then the fan starts, runs for some minutes and then the charger turns off, turning the charging led from red to green.
Battery is not fully charged though. I set the no-load voltage to 87.6v for 24s. But charging stops way lower. It seems timer based and not voltage based. If I pull the plug and start all over again I can charge some minutes more.
 
izeman said:
dnmun said:
GT50J325 is an insulated gate bipolar transistor, IGBT. i got them for $4 each from utsource.com since they are $13 from mouser.
Click to expand...
'Has anyone seen this' was referring to the bent pins. Not the transistor. The local electronics's shop sells them for $20+.
Click to expand...

yes, the bent pins is from someone dropping the charger and the pcb was not glued to the case so it slides while the heat sink that the transistors are mounted to remains stationary. or vice versa.

i repeat. those not mosfets, they are IGBTs.
 
that charger from pucksterpete i thought i had fixed will not work now that i glued it back into the case. just like yours.

but it got worse, i decided to discharge the 103V 28S 12Ah ping pack i built to test it and just destroyed that battery by accidentally over discharging it while trying to scrub off volts to make the charger turn on. now no way to even test the 100V charger. i was even telling myself to put the BMS back on, but no, no time to waste, then spaced it.
 
when you check the IGBT, use the diode tester function and look for open circuit from Collector to Emitter. but the gate is insulated from the device so there is no pn junction to measure like you would do with a regular npn transistor. but there is a 10 ohm resistor on the gate that show up when you measure the gate so if that is what you see, the gate is not shorted. they are fairly strong semiconductor devices so it takes something like a direct short to kill them, so you can check those again to see if they are functional.

i discovered utsource.com on ebay but the have lots of other devices for sale too. the trick is to pile everything into one order because they charge $4 for the first and then 10 cents for each additional item. they have some of the really big schottky diodes too as well as bunches of other stuff.

but if you still have those IGBTs out of the case then check them again with the diode tester to see if you can figure if they should still work, except for no legs.
 
i checked for shorts with the igbt installed, and after i removed them. i checked the pins and the soldering pads on the board. one of thempins had a short, and i also realized the 10ohm resistance on the board, easily explained as i saw the resistor :)
btw: it's utsource.net - and $4 is int shipping. that's good to know - reliable sources are hard to find (if you are short on money)
 
dnmun said:
i think most of the chargers fail because of the inrush current surging on to the npn switching transistors in the front end when the battery is already plugged into the back end. i have tried and tried to explain to people to not plug in the battery first but to them they cannot imagine why it would make any difference and instead they blame poor manufacturing or 'cheap chinese junk'as the failure mode. most of these would never have failed if they had plugged the charger in first. that is why they have the power switch on some of the chargers, to protect the input when it is plugged in to the AC.
Click to expand...

Because it IS shit engineering. You could implement a soft-start using the DTC pin of the TL494 and fix these inrush issues. Another 50c in parts if that.
 
the pcb has a space for the ICL on the small chargers but it is never populated. there is a short wire jumper. on the switchers using the mosfets they always have an ICL.

on the big 240V EMC-1000 there is a huge ICL on one leg, about 25mm in diameter and i bet it handles 12A at least. i am gonna use up my stash of ICLs when i repair the pile of chargers on my counter.

but i still recommend people consider the switch to be a valuable feature, since i added a thermal breaker to that switch line that would shut off the base current when the transformer overheated.
 

Attachments

  • DSC01588.JPG
    DSC01588.JPG
    96.4 KB · Views: 2,377
I would strongly recommend that you use the soft start using the DTC pin - its in the TI 494 document (I'll get you the name - its here somewhere). On the supplies I design, I use a ramp of 250-500 cycles - quite slow - but large explosions due to the power supply starting loaded (which is what they are really designed for!) are unusual.

The ICL is just to stop the bridge rectifiers and high-side electrolytics getting beaten up (and blowing fuses - through inrush) the good designs use a relay that bypasses the ICL after a few seconds - they are usually oversized (ie. too low resistance) to reduce the steady heat that they produce under load - this is good for stopping themselves from burning up and desoldering themselves off the boards, but they do less to protect the components, which is their primary purpose.

On the bigger power supplies, this should be pretty much mandatory (relay bypass) - there is pleanty of information for calculating the correct size of these. These cheap supplies use ordinary electrolytics with a poor ripple capacity, so they end up hot and bothered pretty quickly. It does however mean that you can remove the ICL completely since they don't charge with quite the vengence of a decent cap.
 
i am familiar with using the soft start cap on the dead time pin. albert vandalen has it on his circuit diagram of the charger:

http://www.avdweb.nl/solar-bike/electronics/portable-lightweight-lifepo4-ebike-battery-charger-800g.html#h0-1-2-5-the-battery-charger-circuit
 
Hey All - another sad BMS Battery Alloy Shell Charger owner here. My 74v 600W unit smoked itself last night, and for no good reason. No abuse, kept a room fan blowing on it, they stayed inside and never got exposed to bad weather.

I'm reading over the thread now, will try to figure out if I can fix this thing or not. I'm long-time engineer with average troubleshooting and soldering skills (and tools). Looking for any feedback or thoughts I can get about the BMS Battery chargers.
 
jkbrigman said:
Hey All - another sad BMS Battery Alloy Shell Charger owner here. My 74v 600W unit smoked itself last night, and for no good reason. No abuse, kept a room fan blowing on it, they stayed inside and never got exposed to bad weather.

I'm reading over the thread now, will try to figure out if I can fix this thing or not. I'm long-time engineer with average troubleshooting and soldering skills (and tools). Looking for any feedback or thoughts I can get about the BMS Battery chargers.
Click to expand...

Take a pic of inside and we can help! :wink:
 
bmsbattery_failed_charger.jpg
The full view of the charger. The section that burned is upper edge, just left of center.

https://plus.google.com/photos/103306472219696118223/albums/5875464357932187601/5875464360850667618?authkey=CO291peHzN_XIQ
Here's a close-up of the particular section. Looks like two "power" resistors, two capacitors, maybe a diode or two.

It appears the main damage was confined to this one small section of the battery charger? The "can" of one of the small electrolytic capacitors simply "exploded" off the part and left the innards of the capacitor to unravel. The two power resistors simply burned up.
 
you can post your picture here in your thread. that's bad. usually when the switching transistors blow they only take out the base resistor.

did you have the little switch turned off when you plugged it into the AC? was there a battery attached when you plugged it in?
 
dnmun said:
you can post your picture here in your thread. that's bad. usually when the switching transistors blow they only take out the base resistor.
did you have the little switch turned off when you plugged it into the AC? was there a battery attached when you plugged it in?
Click to expand...

There was no battery attached when I plugged it in. My habit is to plug in first, flip the switch on and then plug it into the battery. Then the fan starts up and the charger starts running.

It wasn't a catastrophic failure - the charger was working, charging a battery, and reached a point where I heard a "sizzle-pow!". I think that's when the capacitor blew.
 
i have just never seen so many parts burned up in the front end like that. when you get it free you can test the transistors but they are toast for sure. the base resistors are 2.2 ohm 2W. and we can find the values for the other ones there. hopefully the diodes did not blow.
 
dnmun said:
i have just never seen so many parts burned up in the front end like that. when you get it free you can test the transistors but they are toast for sure. the base resistors are 2.2 ohm 2W. and we can find the values for the other ones there. hopefully the diodes did not blow.
Click to expand...

Is there any schematic anywhere for this power supply?
 
yep, albert vandalen did a nice drawing in his thread: http://www.avdweb.nl/solar-bike/electronics/portable-lightweight-lifepo4-ebike-battery-charger-800g.html#h0-1-2-5-the-battery-charger-circuit

he also has the DTC slow ramp design that heath was talking about. i have some numbers for the slope of the graph between the resistor value that changes with the charger voltage rating.
 
dnmun said:
yep, albert vandalen did a nice drawing in his thread: http://www.avdweb.nl/solar-bike/electronics/portable-lightweight-lifepo4-ebike-battery-charger-800g.html#h0-1-2-5-the-battery-charger-circuit

he also has the DTC slow ramp design that heath was talking about. i have some numbers for the slope of the graph between the resistor value that changes with the charger voltage rating.
Click to expand...

Wow! BEAUTIFUL schematic. I see right where the blown components are, and I see why you thought it was bizarre - and it was - that those components failed. I think I had some bad parts to begin with and they finally catastrophically failed. One of those 4.7mF electrolytic caps blew the can off and the spiral interior paper was all that was left.

JKB
 
I had one that did exactly the same thing - was charging away, made some strange noises and exploded.

Decided I have had enough of the stupid thing and want to go to something more reliable. CBF fixing it again.
 
heathyoung said:
I had one that did exactly the same thing - was charging away, made some strange noises and exploded.
Decided I have had enough of the stupid thing and want to go to something more reliable. CBF fixing it again.
Click to expand...

IS there something more reliable? I've been researching and it seems everyone's charger blows up now and then.

This is so bizarre to me. Power supplies are common failure points in the industry, but it's possible to build a pretty solid battery charger. I have a 12V SLA charger that's over 10 years old that I've abused the heck out of and it just keeps on coming back for more. Think about it: how often do you hear about a power tool battery charger going bad? I've used the same Ryobi NiMh charger for a decade too and it's still going strong.

Hrm.

JKB
 
they aren't switching 20A at 170V and 28khz either.

those parts that blew up are the passive elements in the high voltage part of that oscillator. usually it is only the base resistor that blows up with the npn transistor.

heath has a good point about the slow start hack where you add a capacitor to the DTC dead time control function as albert sketched it in. the surge in current into the npn transistor at start up is huge and i think that is what causes the big spikes that damage the silicon. but what do i know.. that is why i think the switch is so crucial.
 
dnmun said:
they aren't switching 20A at 170V and 28khz either.

those parts that blew up are the passive elements in the high voltage part of that oscillator. usually it is only the base resistor that blows up with the npn transistor.

heath has a good point about the slow start hack where you add a capacitor to the DTC dead time control function as albert sketched it in. the surge in current into the npn transistor at start up is huge and i think that is what causes the big spikes that damage the silicon. but what do i know.. that is why i think the switch is so crucial.
Click to expand...

dmun, thanks for your good words. I'm not switching 20A at 170v either (more like 9A at 72V) but that's just a detail, I'm certainly not disagreeing with you - I am hoping I can go back with better quality components and end up with a better charger than I would otherwise have had.

My working theory is that the tiny electrolytic capacitor failed first (more current than it could handle) then the resistors failed. Could be the other way around, doesn't matter - I'm thinking I can go back with better components and solve the problem no matter what.

I think the NPN transistor may have gone south with it too. I'll test it to find out.
 
You must log in or register to reply here.

Similar threads

T
warning about ENERPROF.DE / ENERDAN battery&accesories
2
Replies
33
Views
2,548
tomtom81
T
Lucifer
Want a battery spot welder that works for £20?
2
Replies
26
Views
7,417
Daijoubu
D
E
BMS bypass problems
Replies
0
Views
556
EectraTechGlobal
E
Watchamacallit
Burnt BMS (Pics included) Please help?
2
Replies
31
Views
7,099
Watchamacallit
Watchamacallit
electroneg
New to LiFePo4 and New Member - Battery, charging and BMS Questions.
Replies
24
Views
4,999
john61ct
J
Back
Top