Sanken transistor meltdown

[sjdemark]

Hello,
I have had a problem that has occured twice for me now in separate occasions. In these cases after a prolonged period (several days of running), one of the output transistors just starts arcing to the the screw that holds it to the heat sink. The first time this happened with a Sanken 2SA1216 I thought it might have been because I made the screw too tight. I then replaced it with another and used a mica washer between the screw and transistor. All has been working well with that amp since them.

However, I just had the same problem with another amp yesterday. Same problem, this time with a Sanken 2SA1106. But this time, I had used a nylon washer in between the screw and the transistor, so I don't think it was a problem of being too tight.

In both cases, the transistor was not shorted and actually still tested good. However, the transistor was now conductive between the outside of the case in the screw hole and the collector lead. Also, yes, I did replace the heat sink mica insulators and applied new grease each time.

Any thoughts on why this happened and how to prevent it?

[sjdemark]

Anyone???

I found a thread on this site where someone mentioned that sparking at plastic outputs could be caused by loose transistor screws. I still don't understand how that would cause the sparking... I can understand how the transistor might overheat, but I don't understand why it would arc. Any thoughts would be appreciated.

[AndrewT]

the high frequency signal on a TIG welder Starts the spark and arc without touching the tip to the weld.

Could you have a high frequency oscillation that is "jumping the gap" ?

[Mooly]

Only things that come to mind are that the transistors might be fake? or that the method of mounting is incorrect. It's surprising how "untight" the screws should be when mounting single hole devices. Are the screws spreading the force evenly ?

This is one of those things that you have to see firsthand...

Maybe post a picture of them.