transistor amp not working after shorting output

For some, the plastic case ones are fine, if the part may run warm a TO-126 case style is better. Your supply is 55 VDC, so I'd go with an 80 V device. The first transistor can be a 50 V device, so higher gain and lower noise (2N3904 or BC550 - watch pinout). MPSA06 instead of 2N3904 for the buffer (when added) and the transistor labelled 2N3904 should be an MPSA06 (TO-92) or BD139 (TO-126, might be better). The drivers could go to TIP31B or TIP31C (NPN) and TIP32B or TIP32C (PNP). The (C) version is rated for higher voltage and those would be what I would stock doing service.
 
IMG_20240110_193249.jpg
 
  • Like
Reactions: sg97
That would probably be fine. I'm not sure of physical positions, I would need those positions marked with the original numbers or circuit positions, also attach the schematic with those same designations to avoid confusion.

My gosh, the work done on this is awful. I would have fired that guy!
 
  • Like
Reactions: sg97 and Wakh
Hi sg97,
No, a TIP is a plastic case device. Not a valid replacement. Thermal transfer is not nearly what a TO-3 / TO-204 device is.

I don't know what older numbers there are out there. An MJ1502x is too fast and may oscillate. MJ21196 is the only option at Digikey, it may also oscillate. NOS (old stock, unused) is the other option. It cannot look new, or be sold as new manufacture. 2N5631 is one I used a lot of back in the day.
 
  • Like
Reactions: sg97
Mouser has MJ15001 in stock ATM. That and MJ15003 (out of stock) are derivatives of the original 2N5631. Finding legitimate 2N5631’s on the surplus market ended some time ago (good luck finding NOS). They actually have 2N3055s (OnSemi) in stock too, but going “bigger” would probably be a good idea.
 
15001/2 were originally used as drivers in the Leach amps, with 15003/4 outputs. 5631’s were just everywhere. Fairchild used to make them too, which is why they were attractive. Some say the Fairchild version was better (Dyna 400/416 used them, and I’ve run into many receivers of all makes that ran one pair on +/-50V). I think even RCA made them for a while. They also had a 15003, but I don’t recall seeing them in the field (Just their epitaxial 2N3773 after 1982).

Even modern 3055’s can suffer from hFE-in-the-toilet, and still be in spec. Although the 15015 is far worse in that department because of the inverse relationship between hFE and actual Vceo avalanche breakdown.