How to blow an output transistor

[roger2]

I would greatly appreciate opinions on the following situation:

I recently took my bass amp head in for repair. It was blowing its circuit breaker when I turned it on. The technician fixed it, and reported to me that the amp had two output transistors "dead shorted" and that is what caused the circuit breaker to trip.

The amp is fine now. But I would like to understand WHY this happened.

The amp had worked fine the last time it was used (prior to the problem developing). I used it for band practice, shut it down and went home.

A couple weeks later I went to the band room (at someone else's house), picked up the amp and took it home. When I turned it on the circuit breaker tripped and the amp would not work.

The main point here is that nothing happend to the amp (or the output transistors) while I was using it - it worked just fine as it always had. All I did was move it to another location. The amp was NOT hot when it was moved. And the 120v AC at the new location was used for 4 other amps with no problems so I assume the power source was OK.

The situation brings to mind the following questions:

1) Is it possible for damage to occur to an output transistor when the amp is not powered up?

2) Can simple physical vibration caused by moving the amp cause 2 (of 10) of the amps output trasnsitors to become shorted? ( I would tend to think not, but that's why I seek opinions from more knowledgable people here on this board)

3) What are the most common ways to "blow" an output transistor AND when this occurs is it usually a "short" or an "open"? ( the repair tech did say that mine were shorted)


Thanks in advance for helping me solve this mystery...

[BrianDonegan]

I am by no means an expert, but here's my take:

1) Yes (other power source, hammer, etc), but I don't think that's what happened.

2) No (unless it was the hammer from Q1).

3) All of the output transistors I have seem blow have resulted in a short. Not that that is always the case, just every one I have seen.

The amp I am working on now had the same problem. It worked fine for years. Took it out one day, turned it on, and smoke came out. All outputs on one channel were shorted.

Heat and time are usually the causes. I think that apart from extreme usage, turning an amp on is the most stressful part of it's day (like most heart attacks occur in the early morning). There is an inrush of current and weak devices fail. Probably only one device shorted because of this, but the short killed the second device before the breaker tripped.

-b

[adason]

did you by any chance moved the stuff in cold weather?
if the gear is sitting in the truck in cold, once you brink it into the warm room, moisture gets precipitated inside
you need to warm the electronics before you turn it on
just the thought

i was working on a couple of hitachi amps recently which were blowing the fuses, both had output transistors shorted

[Tekko]

Same with a amp i had, it was working but going into mute after a minute so i turned it off and did some measuring on the driver ics, turned it on and god dc on one channel, tuned out to be a power stansistor that failed during the off period, changed it and now the amp works flawlessy, that output transistor probobly was the cause of the amp misbehaving from the start as it finally descided to fail.

[Giaime]

Oh, I have actually seen many output transistors that were open. It seems to happen with overdimensioned hi-fi amps when there is a single supply, usually in the order of 70-100VDC. I don't know the tecnical details of this.

These are the causes of breaking of outputs known (by me )

1) shorted outputs (but sometimes intermittent contacts to speakers had made damage)
2) the 120V - 230V problem (I've seen this! the power transformer resisted, but the transistors...)
3) bad solder joints to the outputs, drivers, bias pots... The strangest in these faliures that I've seen was a bad joint on the filter cap. This made the supply reach higher voltages and pass the voltage limit of the transistors: they opened and shorted.
4) impedance mismatch on speakers (on cheap devices) like the one who put 2ohm speakers to his cheap mini-hi-fi...
5) driver transistor faliure by heat or something else (notice that these are in many cases not mounted on a dissipator, but they DO work and get hot)

Think of strange other causes, if you have an amp head the first thing that comes into my mind is a open/intermittent connection/short in the speaker cable to the cabinet. You may have stopped playing, removed the cable, broke it in some strange ways (stepping on it, usually, or by your bandmates), returned home, put it on, fired it up, and it died.

You may think: "but if the cable was broken, why do it works now?" because the short could be intermittent, as many cases I have seen on instrument cables. But I never seen this on speaker cable.

[sam9]

N answer here but a couple of things to check:

A- is there any way the output could have shoeted? Not merely at the output connectors (binding posts) but out of sight some where that you can't see it. This could be something loose that moves a little and touches a ground/earth point.

B- Cooling. Heat can kill a transistor fast. If is uses a cooling fan, is the following scenario possible. Power is cut and the fan stops, the temperature of the heatsink and transistors go up, but the PS caps continue to supply current for a few seconds, SOA is breached and secondary breakdown occurs.

B- Alternative. This one really happened to me. Thermal protection is activated a couple of times but not fast enough. Damage to transistors is not fatal right away but it is cummulative. Finally you turn it on one more time and they "just can't take it any more" -- dead amp.

C- none of the above.

[Giaime]

I have to disagree. Damage on transistors is not cumulative on modern ones. Maybe with old Ge units, don't know... But with the moderns it is bad or good. No other choice. At least this is in my experience...

[maylar]

Possibly a voltage spike during turn-on. Or just random component failure... nothing lasts forever. Maybe the phase of one of Jupiter's moons is just right...

[sam9]

I have to disagree. Damage on transistors is not cumulative on modern ones. Maybe with old Ge units, don't know... But with the moderns it is bad or good. No other choice. At least this is in my experience...

Just repeating what the Sony repair guy told me after charging more to repair the #$%^*&^ than it was worth.

[BrianDonegan]

I learned in school (wow, long time ago now) that generally solid state devices have a failure curve inverse to that of mechanical devices, meaning that mechanical devices tend to work well and wear out over time, solid state tends to either fail early, or last a long long time.

This is of course a generalization. I think a big exception is power devices that get hot. Heat kills just about everything. It really makes solid state devices mechanical devices.

-b

BTW: I remember now seeing an dead transistor that was open. It was also physically open (big crack where the flames came out).