When Insulated is not always Insulated.

I would have never figured that out. It would have ended up on the shelf.

I was ready to put another one in when I figured it out.  Not the first thing that came to mind.  I had the thing out of the heatsink looking all over for something on the board at the edge.  Finally stuck the meter probes in the holes and got continuity from one of them, mystery solved!

gunrunnerjohn posted:

Here's a slightly odd one for you.

I was doing a TMCC upgrade to a 3rd Rail Decapod, and an odd issue surfaced.  When it was finished, it ran 100% on the bench on rollers with the tender just sitting to the side with the tether plugged in. 

Put it on the tracks for real, and it was all over the place.  As soon as you cracked the throttle, it was off to the races.  Not full speed like no serial data, just cruising the rails with no speed control at all.  To make matters worse, it would stop when I hit the reverse button.  However, trying to run it in reverse tripped the PH180 breaker!  Obviously, something bad happening!  Throughout the bad behavior, I had some commands getting through, I could blow the whistle and even turn the smoke on/off.

I finally tracked it down to the fact that one of the insulated tab FETs on the Cruise Commander was not so insulated!  The mounting screw was intermittently making contact inside the mounting hole with the tab and shorting one lead of the motor to the frame!  The reason it worked on the bench was the tender was not on the rails, just sitting to the side with the tender connected.  So, even with the short to the frame, it didn't have an effect.

To make it a little more complicated, it would measure fine on the bench, the short didn't always show up.

 

Like you say John: "Nothing is easier than the job someone else is doing"..............

I imagined this one to be easy, but I missed the boat.

gunrunnerjohn posted:

I was ready to put another one in when I figured it out.  Not the first thing that came to mind.  I had the thing out of the heatsink looking all over for something on the board at the edge.  Finally stuck the meter probes in the holes and got continuity from one of them, mystery solved!

I've always been very suspicious of those tab insulators.

FWIW Chuck, this is the first time I've ever had an issue with the insulated tab parts, and I've been using them for a very long time!  I've installed probably well over a hundred Cruise Commanders, and I've also used insulated tab parts in other projects, this is a first for me.  It is one that will be stored in the memory bank if I ever see this oddball behavior again!

I filed the hole in the FET a little larger and put heatshrink on the screw to insulate it, we back in revenue service.  

At first I thought for sure the FET(s) were shorted, but then it worked perfectly on the bench, even running it for an hour with a motor with a load in both directions.