Replace MJ10000 NPN darlington with Toshiba K2915 N channel FET

[electronics-whiz]

I'm trying to fix something that had 6 MJ10000 transistors. I looked at identical replacements and to fix this thing would cost like $45 which I don't think is worth it. These were total overkill also as is for 208V VFD drive that maxes at like maybe 2-3 A.
I was running it on 110V worked, not sure was part of why it failed. I'd kind of like to get atleast some of it working to use as a transformer tester, but it's not worth the $45 to fix to me.

I have a bunch of Toshiba K2915 from a gob of old identical SMPS units. I have similar parts from UPS units too, but not sure a MOSFET can replace a Darlington. I don't see Darlingtons much anymore. now it's all MOSFETs or some cases IGBTs. I have 3 large IGBT from another old junk VFD drive, but doubt those would help with this either.

[oldway]

There is no easy way to do that.
You should make a new printboard with Mosfet drivers.

[Ian.M]

At the very least, you'd have to rip out the base drive circuit all the way back to wherever you can get a clean logic signal and make a daughterboard with proper MOSFET gate drivers on it to get suitable gate signals.  Also there may not be any suitable supply rail for the high side gate drive, or if you go N channel all the way, the high side drivers need to be isolated and fully floating.

There's a heck of a lot more than 45 bucks worth of work just designing the replacement driver circuit, and by the time you've ordered the daughterboard from a PCB houser, built it up and tested it etc.  you'll be well over the cost of a new VFD off EBAY.

[electronics-whiz]

Sort of what I was thinking. I was able to find the ouput portion had it's own DC bus off of a large filter cap. I disconnected and the issue went away.

I'm not exactly sure how it works, but if I trace back from the base of the failed Darlington I see this:
Base links to the striped side of a diode (black plastic like rectifier diode), the other side links to a 51 ohm 2-watt resistor. The other end of the resistor links to the collector of a 2N2907 K8619 transistor. The base of this transistor connects to 2 small 1K ohm resistors one connects to pin 4, the other pin 6 of a GE  H11L1 opto isolator.

Seems to me a circuit like that you should be able to throw any proper voltage, current, frequency rated transistor in and have it work.
I think it may just be junk as it 30 or more yrs old while I don't see any sign of it there could be some bad caps on it also. This unit was supposed to be a PLC trainer for a school. All of the parameters are set with potentiometers. I have a small AB 120/240 3-4A PLC with LCD screen, lots more features I could use instead.

The only real reason I desire to fix this is to have something I can use to test transformers or get a variable frequency signal from. The first one I can't do without some larger transistors. The later I could probably just tap off on the base of one of the old darlington transistors and get a small voltage variable frequency signal.

[Ian.M]

VFDs generate a very harsh switching waveform with spikes that can break down transformer or motor winding insulation unless it is rated to be powered by a VFD so I certainly wouldn't use one for transformer testing.   You certainly wont be able to get a clean variable frequency sinewave from the VFD controller H-bridge drive signals as it will be a pair of pulsetrains per phase that will need to be combined then put through a heck of a lot of filtering.

A beefy audio amplifier + a step up transformer fed by a signal generator might be a better option. 

[electronics-whiz]

Good point about the amplifier. I have a cheapy parallax propeller board I'm sure if I want to I can program whatever frequency in I want. Then use that to drive something or use large mosfets to drive something. Then I can put in a lot of frequency values.
At least for mains transformers a variac does fone and you can calculate things out if needed.
So far yet see a real need to deal with SMPS transformers, and unless it has a thermal cutoff or a winding breaks an ohm meter could test that fine.

I opened it up did a few probes just to see what the remainder of it did. I was able to get about 1.8-2.3 V AC on the base of the old darlingtons I removed. Base V increased with increasing speed knob and raising frequency. It didn't appear to be working that stability though. As would show nothing then would start seeing something then repeat.

I'm just going to part it out because has limited use and is about the size of a computer tower. My AB PLC is about the maybe shoebox sized and has so much more potential. Debating keeping the case for a future project of some sort.

[Ian.M]

It was probably faulting out and restarting because the output stage was disabled.  Not worth bothering with figuring out exactly what was going on.   Unless its a particularly nice case or you have a spare barn to stick 'may come in handy' junk in, I wouldn't bother keeping it unless it suits a current project.

[electronics-whiz]

True. I didn't detect any feedback from output, I suspect something either in it fried the transistors or the other way around. Based on age, looking at the unit design I wouldn't be surprised had more issues or a pain to track down any issue. I may keep case a month or 2 and see if can think of any use if not trash it.