Help identifying transistor

[uvacom]

I have a TO-3 package transistor that I pulled out of an EHX Mike Matthews Dirt Road Special (schematic attached). It's supposed to be the PNP in the output stage, the NPN is a 2N3055 which appears to work fine. In testing this transistor (out of the circuit) with a multimeter, I measure a very low resistance when the common probe is on the collector and the positive probe is on the base, and a very high resistance in any other configuration. To me that suggests that this is an NPN with a burned out base-emitter junction.

Here is a picture: I think it might be a 2N5038 (which most sites list as an NPN, although I found a General Diode datasheet which lists a 2N5038 as a PNP in a TO3 package), but I'm just not sure. The markings are cryptic. Is it a Fairchild part? I could not find any information on these markings.



At any rate, it seems like a safe bet that I can just replace it with an MJ2955 but I am curious as to what this transistor is, and if it is in fact an NPN how it got in the circuit.

[Enzo]

I don't know what the part is, that S37004 number is a house number for whoever made the thing. SO specs won;t be published. Very doubtful it is a custom part, so I look to the listings of generics.

If the NPN is a 2N3055, then I can;t think of anything better to use as its comlement than the MJ2955 intended for such purposes.

Or replace the pair with MJ15003/15004, or even a pair of MJ15024/15025.

Don;t forget the check the driver transistors and all teh associated resistors in the power stage.

[uvacom]

I don't know what the part is, that S37004 number is a house number for whoever made the thing. SO specs won;t be published. Very doubtful it is a custom part, so I look to the listings of generics.

If the NPN is a 2N3055, then I can;t think of anything better to use as its comlement than the MJ2955 intended for such purposes.

Or replace the pair with MJ15003/15004, or even a pair of MJ15024/15025.

I thought about that, a house part, but it doesn't make too much sense to me that they would get a house P/N for the PNP but then leave the stock labeling on the NPN 2N3055. But who knows.

Good call on the MJ15xxx parts, I might as well replace them with something newer, I've read that the new epitaxial 3055/2955s don't have the same SOA specs as the old devices such as what's in here (likely made around '78 or so), so what the hell, right?

Don;t forget the check the driver transistors and all teh associated resistors in the power stage.

Yup, they're all fine except the 330 ohm resistor at Q5 is getting way too hot and even started to discolor after I left the amp on for a few minutes. From what I can tell that's because the B-E junction at the mystery PNP is dead so the 3055 is sourcing it's current through that resistor instead. I also simulated the circuit in pspice and was able to reproduce the symptoms by taking out the PNP and replacing the B-C connection with a short. I guess I won't really know until I pop in the new parts but it's not a terribly complex circuit.

[Enzo]

Cirrent. Well, you should be working on the amp without a load until it is stable and working.

The original parts may have had all house numbers, the 3055 might not be original. After all, 35 year old amps can have 30 year old repair parts in them. And for that matter, if during production they ran low on house marked parts, they may have decided to use a tub of generic ones rather than hold up the line for fancy printed ones. Who knows?

To me those MJ series parts are "universal replacement" power transistors. They wil work great in just about anything.

[uvacom]

Cool, well I replaxced the power transistors and whaddaya know, the amp works fine! But I have discovered that the stock celestion G12H30 has a blown voice coil. It's a "blackback" from the late 70s with a 1777 cone so apparently it's not collectable or anything, but it's still a bummer since that is a major reason I bought this amp.