IRF431 replacement

[cncjerry]

Hello, I was playing a little too hard with my new Keithley 228a and popped an IRF431.  I bought two from china but was wondering if there was a cheap, readily available substitute?  IT is an n-channel power mostfet in TO3 case.

Thanks

[BennVenn]

Just a word of caution, what you order from china is not always what you get even though its printed on the package. I have a bag full of IRF4110's that say they are but are clearly not IRF4110's. Don't make the same mistake!

[all_repair]

Just a word of caution, what you order from china is not always what you get even though its printed on the package. I have a bag full of IRF4110's that say they are but are clearly not IRF4110's. Don't make the same mistake!

Concurr, 100% QC is needed so they can end up more expensive.  But I can only check for reverse breakdown voltage easily but good enough for my usage.  I mainly use them as "test" FET during troubleshooting because keep blowing geniune parts before all the faults can be cleared are very costly.

[cncjerry]

 yes, I don't always trust batch parts like that though it seems like the older the part the more likely of getting an original.

I am starting to wonder about the mosfets.  I was doing resistance checks across  r305, 306, 307, 308 which are in the A/C oscillator for the two voltages that are out on the upper right of the attached schematic section.  These were all replaced as part of an engineering change and whomever changed them used Dale C-64 1% T1 1/2w resistors.  They changed the pair of 22R0 to 18R0 which matched the doc.  I just find it strange that they are all open.  At first I thought it was my leads or meter but not.  They have no external signs of smoke holes, etc.  They are on the gate and the only thing I can think is that both mosfets failed with an internal short to the gate.  I'm struggling with the odds of that happening.  Any ideas of a failure scenario that would pop all four?  The parts lists suggest 1/2 W carbon comp, nothing fancy so I am going to throw some in there and watch for smoke.

Thanks for the warning on the parts.  the local NTE rep wants $100 per.  Yea, right.

Jerry

[cncjerry]

So I replaced R305  thru r308 with 1% carbon 1/2 watt and the promptly went up in flames, smoke, etc.  but I found the problem.  Whoever put it back together once had run a wire within the frame wrong from the power supply that power supply.  This time when the resistors went up I saw the arc and checking found the heatshrink pulled away and aluminum splatter from where this huge diode made contact with the frame.  There is a cutout in the frame and if the previous hacker had used it this wouldn't have happened.  So now we know that overload of those supplies off T302 will take them all out.

There could still be a mosfet problem.  This time I am going to use 1/4 watt resistors which should at least bring the supply up without load.

[Richard Head]

Use a DC power supply to power the control circuitry before connecting mains to it. Then, if all is ok connect a variac to the input and bring the mains up slowly while monitoring the gate drives etc. This will avoid the possiblity of flames and smoke if there is a problem.

[Seekonk]

A 100W lamp in series always works for me with switching power supplies without damage.

Thought that number looked familiar.  I still have a nearly full tube of IRF433 I bought from Newark in the 90's.  On is 2 vs 1.5 ohm if interested.

[cncjerry]

I've been using that lamp trick for 38yrs to be exact.  When I was a young whipper-snapper I fixed stereos for a while.  This guy that owned the shop had all kinds of tricks like that.

It was strange last night after I put the 1/4 watt resistors in they took about 5minutes to pop.  That board won't take much more of that so I will pull all the loads today.  The +5/-5 come up so I have the logic running. That power supply runs the +/- 15 and the +/- 115 so there could be a load issue.  It is very difficult to service though, all spade lugs and old brittle connectors.

Thanks for the tricks.

Jerry

[cncjerry]

Traced it back to a shorted MUR860 T0220 diode in the -115 supply.  Go figure.  I should have looked at this more before ordering the IRF431 mosfets, but hey, I usually rebuild around the smoke first anyway.

Will report back when I have this running.   If you havent seen a Keithley 228 you should take a look.  This is a cool box.

Jerry