stefan, are you still there? do you still wanna try to fix the charger?
i finally got greg's big 900W BMS battery charger fixed. this is the one that burned up the ground lead.
when i got it i found that one of the big switching transistors in the front end was shorted to the case. that is why the ground wire had to carry so much current because it was rectified DC poured directly into the case. too much current (20+A) and it burned up the ground wire connection.
this is why i now recommend that anyone using these metal alloy cased chargers use them only on circuits with ground fault protection. whether kingpan charger or the BMS battery charger, both of them are liable to put 200V DC onto the case when they short out and if you touch the case while grounded you can receive a fatal shock.
200V is not a trivial voltage so i am recommending again that people be aware of this risk and do not handle the charger case while grounded since if it fails while you are holding it and are grounded, the ground will burn out just like this and you will be receiving a high voltage shock and may not be able to release the charger fast enuff to avoid a fatal shock.
it turns out that his charger had shorted the transistor to the case because there was a tiny, as in really tiny, the size of a pin tip, not pin head but the tip. less than .1mm across and .6mm long metal filing embedded in the heat sink insulator. i am certain that it was a piece of aluminum from the hole drilled in the heat sink for the bolt that holds the transistor onto the heat sink.
i postulate that the worker assembling the IGBTs onto the heat sink had pushed the screw through the hole and that had knocked the metal filing outa the hole and right under the rubber insulator just as she was putting it together. she would never have seen it because it was under the transistor.
it was so tiny that it did not short out the collector to the heat sink initially so it woulda passed inspection for continuity that i am sure must be part of their assembly process quality control. someone in their QC must go through every one of these heat sink/ IGBT, schottky diode, rectifier diode bridge assemblies after it is screwed together, checking continuity between the IGBT collector and the heat sink.
that would allow for repair of a short before the assembled heat sink and transistors were then mounted to the pcb and soldered into final position. i am certain they must have this inspection because of how easy it is to catch a short at that point.
this metal filing was not large enuff to cause a short initially so it passed inspection. it finally shorted out after greg had used it several times and it had heated up enuff to allow the insulator to soften so that the metal finally could push through the insulator enuff to short out the collector to the heat sink.
that was when it started smoking and burned up the ground wire connection. later when he fixed the ground, it was a strong enuff short that it blew up the transistor. so if the charger had been tested for a short at the time that the ground wire burned out, then the transistor was still intact and the only intervention would have been to take out the insulator and examine it for the shorted spot. it could been fixed then by removing the metal filing and replacing the insulator back under the IGBT with some more heat sink thermal compound. my TO-3P insulators are not even big enuff for these IGBTs so i put one of the TO-3P insulators under the original perforated insulator.
not many would think of a metal filing as the culprit, instead would think the transistor was just bad, so that is why i am making this warning now so people will know what to expect if they have a ground wire burn out in a similar fashion.
it was such a hassle. the switching transistors for these big 900W chargers switch so much current that they used IGBTs (40A 400V toshiba GT50J325) for the transistors instead of the big npn switching transistors.
they are expensive. $13 each from mouser. i found a hong kong source and got 3 for $16.20 including shipping. i ended up needing all three, even though only one had blown initially.
when i examined it, i had tested the second transistor. i thought maybe it would be dead too just like the one that shorted. but when i tested it the collector-emiiter was open circuit and there was an epitaxial diode between emitter-collector, and the gate, the insulated gate used for a 'base' in these transistors, was also open circuit so it was ok, imo.
my mistake was that i did not test the rectifier diode bridge also.
it had shorted when greg reconnected the ground wire and it then shorted out the diode bridge when he powered it up again after repairing the burned open ground lead.
so test the rectifier if you have this problem in the future. such a bitch to work on that this was a hard lesson to learn since replacing the bridge with the first transistor woulda fixed it then. i would still have the extra two IGBTs for parts too. expensive lesson.
so i powered it up after just replacing just the IGBT and the gate resistors for both transistors. one had blown with the IGBT so i replaced both so the gate resistors would be identical. both from the same batch of resistors. 10 ohm 1/2W 1%.
the fuse blew right away but it had already blown both the IGBTs by then so i had to take it all apart again and replace two transistors the second time. i was really lucky to discover the shorted rectifier diode bridge the second time. of course i should have looked for it the first time. so stupid.
duane 'biohazardman' had accidentally blown up the rectifier diode bridge on a big 900w kingpan charger and ordered two of the big diode bridges when he fixed it. a huge capacity, 25A 1000V bridge. GBJ2510-F, not the cheap $.83 type, the $5 kind.
he gave me the extra one after he was done so i had an exact replacement already in my stuff so i installed it along with the remaining two IGBTs that i had left from hong kong, soldered them all in place and cleaned up the flux with some nail polish remover. plugged the charger in and............... it works!!
i am so happy to get this fixed. whatta struggle today. impossible to reach the IGBTs under the big lip of the heat sink. then having to replace the diode bridge too which was also under the overhang of the heat sink. so glad it is fixed. except now i am gonna see if i can tweak it up to 100V for him so i am gonna order some higher voltage capacitors than the current 100V size so they won't blow up right away and i would have to take the charger apart again. no way. so much hassle.
it looks like i will have to use a 160V nichicon. to get the 7.5mm spacing it will be 18mmx40mm 330uF but that may be too large to fit in the space, so maybe only the 16x35.5mm 220uF which also has the 7.5mm lead spacing. the current ones are 100V470uF so i will be giving up a lot of capacitance but since it is gonna only be exposed to the voltage while attached to the battery pack then the battery will provide a lot of the capacitance when actively charging.
i also have a fix for the other charger where the choke had shorted out to the 12V regulator supply leads and blown up the output capacitor. that was the one that had burned a hole in the pcb and turned the choke winding black as it shunted so much current from the schottky diodes directly to ground and overheated so badly. that actually is kinda easy to fix compared to the shorted IGBT on this 900W BMS battery charger. EMC actually is the manufacturer. i think they are either the same outfit or very close.
i will cover that repair later. i also was able to fix his icharger and found that the fuse had burned open inside the bionyx charger and i am waiting on the leaded fuses for it now.
the icharger had burned up the discharging shunt wire when he used the icharger to discharge his lipo pack. this was a fatal error.
i wanna recommend nobody use these balancing chargers to discharge your battery. they are not designed to handle that kinda power and heat. the shunt in this one was a 4" long nichrome wire and it had melted and twisted into a pretzel from the heat. so discharge into a load that can handle it, not through the internal shunt. you will lose your icharger if you do.
so no more ichargers as dischargers or it will be finit icharger.