driving your power stage

[hacknet]

I've seen really good designs that use a MOSFET source-follower to drive the output stage from a high impedance tube gain stage, so I'd say it's not a bad idea, just a little heresy

i was thinking about that lately when i was designing a 2a3 in class A2.
what worries me is the input capacitance. i read somewhere that it is negated by the negative feedback from the source resistors. but after alittle simulations, the input capacitance especially with high voltage mosfets seem to be pretty scary. it seems like a monster load to drive and bandwidth is going to be another issue.

hope to hear more on the discussion of this topic!
thanks
JT

 

[Brian Beck]

i was thinking about that lately when i was designing a 2a3 in class A2.
what worries me is the input capacitance. i read somewhere that it is negated by the negative feedback from the source resistors. but after alittle simulations, the input capacitance especially with high voltage mosfets seem to be pretty scary. it seems like a monster load to drive and bandwidth is going to be another issue.

hope to hear more on the discussion of this topic!
thanks
JT

BROKEN RECORD WARNING: It's not just bandwidth that's a concern, it's also the non-linearity of the MOSFET's capacitance that's a potential problem. As you say, the negative feedback at the source in a source follower tends to reduce this problem, compared to the case of a MOSFET used as a gain stage or even as a CCS. This is because the Cgs is bootstrapped (greatly reduced) by the "following" action of the source, and also there is no Miller multiplication of Cdg. Depending on the output resistance of the preceding stage, there may still be an audible character due to the gate capacitance, however. This old post of mine briefly analyzes the potential problem with non-linear capacitances in another circuit example.

Non-linear capacitance modulation

 

[Bandersnatch]

The cascode will take care of the capacitance issue. With enhancement mode, you'll need batteries for the upper one. FOr depletion mode, you can get away without...
cheers,
Douglas

 

[hacknet]

BROKEN RECORD WARNING: It's not just bandwidth that's a concern, it's also the non-linearity of the MOSFET's capacitance that's a potential problem. As you say, the negative feedback at the source in a source follower tends to reduce this problem, compared to the case of a MOSFET used as a gain stage or even as a CCS. This is because the Cgs is bootstrapped (greatly reduced) by the "following" action of the source, and also there is no Miller multiplication of Cdg. Depending on the output resistance of the preceding stage, there may still be an audible character due to the gate capacitance, however. This old post of mine briefly analyzes the potential problem with non-linear capacitances in another circuit example.

Non-linear capacitance modulation

great reading material! will sieve through it!

The cascode will take care of the capacitance issue. With enhancement mode, you'll need batteries for the upper one. FOr depletion mode, you can get away without...
cheers,
Douglas

i was hoping to bias the e mosfet with the previous stage's dc. is that feasible?