Help needed on new output transistors

[reddish75]

Well my kenwood supreme 600 has thrown a pnp output transistor, so just replace it i here you say... not that easy i'm afraid the original sanken 2sa747a and 2sc1116a's are long out of production in which lies my problem.

The originals were around 14mhz, there are modern replacements that will do the job the on-semi MJ21193/94 but they are only 4mhz (yes i know this is minimum value.)

Step forward the stmicros 2ST5949 (NPN) and 2ST2121 (PNP) these are even better spec'd than the on-semi parts and are faster at 25mhz, in where lies my problem, these may well occilate so what can i do to remede this if it happens?, i posted this on audiokarma and EchoWars kindly responded with this 'I'd consider adding a bit of capacitance from the base to the collector of Qe6. But this is essentially hiding the problem, and you can end up with an amp that is actually slower than it might be if you used the On-Semi devices.'
Putting ferrite beads on the base's of the outputs is another option. Is there anymore options which dont carry any comprimise as much.

I posted on here as its more techincally oriented than audiokarma, the floor as they say is yours

Thanks
Chris

[ontoaba]

Using faster OPS wont make oscillations, because Cc is already at VAS stage.

note: higher ft is not always faster, but LAPT like A1216 - C2922 is fast you could try with if voltage is ok.

[reddish75]

Originally Posted by ontoaba

note: higher ft is not always faster, but LAPT like A1216 - C2922 is fast you could try with if voltage is ok.

I'm not looking for faster, just something with similar specs.

So with a constant current source at the vas means that the amp wont oscilate with faster than spec'd outputs?

Thanks

[Ouroboros]

Emitter-follower ouputs can sometimes oscillate (nothing to do with the VAS stage), but I see that there are already 10 Ohm stopper ressistors in the signal lines to the bases of the output transistors, so it is likely that everything will be ok.

[Mooly]

In practice I don't really see a problem with the MJ's

Did the failed output jump or was it pushed ?

I'll just mention that a common failure mode of many Japanese T0126 packaged devices (such as drivers etc) is intermitent failure with base-emitter going open circuit. Why this should be I don't know but years as a bench tech proved it so. There were ever any contributing factors involved.

[reddish75]

It completely commited suicide by itself in a corner with nobody watching.

With the MJ's i know they would work well and would'nt just run at 4mhz but they also wouldn't run at 14mhz like the originals thats why i would like to try the stmicro's, it's just a dilemma but availability of the ST parts just might swing in favour of the MJ's.

[Mooly]

I really don't think you will see any difference performance/distortion/stability wise with either of these vs the originals.