Class A SE power and driver transistor choice for stability and performance

[ionomolo]

Yes, i know this topic has already been discussed many times, but when it comes to transistors and integrated circuits i always fear following obsolete advise, so i'll ask again.

At this point i would choose FJL4315 for power and give the 2SC4793 the driver role. I've always been told that NPN's where better than PNP's, current transistors still follow this trend?

They would be driven from a very low Z and i belive this to be enough to ensure stability (the pole created by the output stage would start to act where the Vas gain has already fallen below unity). I'm right?

EDIT: I plan to parallel 4 output devices to make heatsinking easier, this would give a 800 pF output capacitance, is it just too much?

[Lumba Ogir]

Hi, nice to see you again.

I've always been told that NPN's where better than PNP's, current transistors still follow this trend?

Necessarily, yes.

I plan to parallel 4 output devices to make heatsinking easier, this would give a 800 pF output capacitance, is it just too much?

Obviously, you need more power this time. I would go for cascoding instead.

They would be driven from a very low Z and i belive this to be enough to ensure stability (the pole created by the output stage would start to act where the Vas gain has already fallen below unity). I'm right?

I`m not sure of that. Could you present the complete schema, please.

[ionomolo]

Originally posted by Lumba Ogir

I`m not sure of that. Could you present the complete schema, please.

It would be driven by an opamp. It may sound silly but i have one of these sounding right now and i like it. It uses a single onsemi transistors MJL4281A as power transistor, which has an output capacitance arround 600 pF as opposed to the 200 pF from the fairchild one. I belive that increasing the driven capacitance by one third will not make the whole thing explode.


I was looking for a good transistor choice to build one i can live with for decades.

EDIT: I'm also happy to hear from you again... but this takes too much time! It's hard to belive but i've promised myself that this time i would do something simple.

[Lumba Ogir]

Hi,
I did really like your previous design, which you came up with seemingly instantaneously. (It took me many years to approach that topology).

It's hard to belive but i've promised myself that this time i would do something simple. I belive that increasing the driven capacitance by one third will not make the whole thing explode.

OK, from now on I will not try to influence your intention here.
I don`t think it will be that dramatic, but you may need to make some change in compensation.