MTH TIU Variable Channels Produce Constant DC Offset

I believe this has been asked before. I found this on a quick search.

https://ogrforum.com/...ariable-tiu-channels

Maybe you could send an email to GGG.

I can't imagine a TVS doing this, it's wired directly across the power.  If it were acting like a diode, it would quickly overload the FET voltage control and likely cook them. 

If one FET fails, you can get a DC offset, I just fixed a pair of TIU's with that exact problem.  I replace all four as you can't really do a test in circuit of the FET, and since I have to remove them anyway, it's easier to chop them out and remove the leads individually.  The parts are only a buck, so saving them isn't worth the labor.

Thanks guys. As my first step, I will remove / replace the FET's and see if that does the trick.

If a FET has failed, I am not sure how you can have full voltage control over the VAR channel unless it is some partial failure/short.  You can diode check the fets, when one is shorted you will see it.  Usually (almost always) only one fet fails.  G

All I can say is I had one FET fail and replacing all of them resolved the oddball DC output indication.  I agree it's usually one, but it's much easier to remove them one leg at a time, so I just chop them and pull the legs individually.  You can't really test them in circuit, so it's pretty hard to determine which one, hence my technique of replacing them all.

Actually you can test them in pairs, and I know the one that typically fails 80% of the time.  So I remove it first.  For me, I bridge the legs with solder on the back of the board, leave heat sink on, and pull the fet out as  unit.  Works great.  Then retest.  Confirm the Fet pulled has shorted.  Replace.   G

Interesting leaving the heatsink on, that saves a little time.  I've only done maybe a dozen of these so far, so I haven't "polished" the technique.  I did note that they seemed to be in pairs, that helps.  Which one fails most of the time, useful information to have.   If I knew that, I could pull the pair and replace, could save time.

The front of the failed pair.  G

The front? 

gunrunnerjohn posted:

I can't imagine a TVS doing this...

But it seems reasonable to assume that the purpose of the TVS is to protect the TIU electronics.  So if a TVS fails (for example from repeated spikes from the track), then the previously protected FETs would be exposed to those spikes and possibly fail.

It appears the 10-cent FETs (free shipping from Asia) are now 16-cents.  

That's certainly true Stan, if the TVS fails open, you'll never know.  Perhaps it's a good idea to replace the TVS whenever you replace the variable channel FETs.