Acoustic 126 - hot as hell

[boomfred]

Hello there.
I do have a problem with an Acoustic 126, the combo bass amp.
Basically, it worked great until it blows a few weeks ago. 
The power board (170080) is faulty as I do get signal from the preamp board.
I thought I succeded in troubleshooting it (basically replacing the 40410 driver, a couple of resistors and diodes and the power transistors) but it only worked a few hours, went REALLY hot and went down again.  :'(
I’ve read here and there that original power transistors were « specials » and I’m afraid I only own « regular/not that old » ones.
Any idea anyone ?
I’ve heard of MJ15022 instead’ of 2n3055, any thoughts ?
Thank.

 

[Audio1Man]

You need the OLD RCA 2N3055 parts in the outputs. Replacing the parts with new fast parts could be done if you could compensate them. The old parts Ft=800kHz, while new parts Ft 4mHz to 25 mHz, If the amp was working and just got hot in a short time I would look for a BIAS problem, Try shorting Q3 and it will remove bias and you will get big crossover and should trop the big heating.

Duke

 

[iampoor1]

I have some old removed Motorola 2n3055s from the early 80s Message me if you need to buy some.  ;D

 

[boomfred]

l finally found some MJ15015G to replace the burning 2n3055.
It’s been working for a few hours now.
Thanks everyone.

 

[Bo Deadly]

I have seen 2N6488 used in place of 2N3055. At 5MHz it's a relatively slow part by today's standards. It is lower wattage but if you have room for heat sinks (it's a TO-220 package instead of the TO-3) it might be ok. The SOA chart doesn't look that much different.

 

[JohnRoberts]

+1 look for possible oscillation... easier to change compensation than devices.

JR

 

[JMFahey]

A couple details:
* those amps tend to be biased hot
* power transistors there do not use emitter resistors, which makes them less thermally stable, and *demands*  perfect matching for even current sharing.
Fine for a Factory buying them by the thousands straight from RCA or Motorola, not so much for an individual
Personally I add 0.22 or 0.33 ohm 5/7W emitter resistors.
* those selected Hometaxial 2N3055 are absolutely unavailable, MJ15022 work fine, never had unstability problems but scoping output driving real speakers (not mild load resistors) never hurts.
*IF unstable I may suggest a cure, but only if actually necessary, not "just in case".

 

[JohnRoberts]

A couple details:
* those amps tend to be biased hot
* power transistors there do not use emitter resistors, which makes them less thermally stable, and *demands*  perfect matching for even current sharing.
Fine for a Factory buying them by the thousands straight from RCA or Motorola, not so much for an individual
Personally I add 0.22 or 0.33 ohm 5/7W emitter resistors.
* those selected Hometaxial 2N3055 are absolutely unavailable, MJ15022 work fine, never had unstability problems but scoping output driving real speakers (not mild load resistors) never hurts.
*IF unstable I may suggest a cure, but only if actually necessary, not "just in case".

Even with matched or same batch power devices emitter degeneration (resistors) are good practice to force current sharing, and stabilize class A current bias.  Somewhere there are equations that predict stability wrt thermal resistance junction to air (i.e. the heatsink, but I have seen or used those equations for decades). IIRC they also make 0.1 ohm power resistors, that might help while altering the rest of the amp less.

The Vbe of typical power devices falls as their temperature rises. If the bases are driven from constant voltage, (as most are), when they heat up, they will draw more current, heating up even more. This positive feedback thermal runaway is an amp killer. This can even even happen from loose power transistors that exhibit too high thermal resistance. 

Emitter degeneration resistors work to reduce this positive feedback effect adding a linear term to exponential current vs base-emitter voltage relationship.  IIRC there were optimal emitter resistor values for thermal stability without wasting excess power.

JR

PS: I've seen my share of marginal heat sink designs. While product manager over power amps at Peavey I had engineering add an extra pair of output devices to one existing amp, that dramatically improved its thermal robustness. 

 

[JMFahey]

Yes.
To boot Acoustic heat sinks always were quite poor, usually just the bottom of the chassis , inside of the Tolexed case, of course, and exposed to free air through a window cut in the bottom.
So it was horizontal (the worst for natural convection) and sandwiched between head floor and cabinet roof, counting on head´s tiny legs (1/2" or less) to keep that separation.
In fact it was a miracle that most lasted quite a long time anyway.

As of transistors, I confirm they were RCA Hometaxial ones, and proper name should be 2N3055H to separate them from common Epitaxial ones, which should be labelled 2N3055E . Of course only Factories who bother to do so are those who make both types.

"E" type are nice transistors, fast and clean ... for home audio and such, (think NAD amplifiers or many 70s 80s European ones) NOT for rough life Guitar/Bass amps.
When RCA ran out , I bought a batch of, surprise surprise , India made 2N3055H "USHA" brand, labelled "UR" , others that worked were Yugoslavian made "Ei" brand, again "real" 2N3055H .
Not surprisingly, both were products of *military*  factories for their own use which of course made the robust version.

ST are excellent and you can get as many as you want for peanuts ... but too weak for Musical Instrument use.
Oh, you can use them with less that 70V (end to end) rails so +/-35V split supply ... tops .
And pulling 50W from a pair.

You could get almost 100W out of a well heat sinked RCA pair