simple hybrid idea

[nobody special]

Do you know how much gm you are expecting, and thus how much gain? AC balance might not be a problem...

Honestly, no. In my simulation of it I am able to drive the circuit above with 1V p-p and get full output. The output stage is basically taken straight from the RCA databook.
I could always match the differential mosfets and used matched output tubes... maybe then it would be alright?

[Tube_Dude]

I wonder what the input capacitance of your gain stage looks like?

Yes the input capacitance will be great...the gate source capacitance of the mosfet +the Miller capacitance ( the gate drain capacitance X the gain of the stage)...it will need some driving current in the high frequencys...

Mosfets are not my choice for input devices...

cheers

[EC8010]

At maximum output, one output valve has its grid at 0V (a -17.5V swing) and the other at 35V (a +17.5V swing). Thus, there is 35Vpk-pk between the outputs of your driver stage. If there is 1Vpk-pk going in, then you have a gain of 35, and since Av = gmRL, that suggests gm = 35mA/V for the MOSFETs. It probably would be worth matching the MOSFETs, as they are notorious for device variation.

[Tube_Dude]

Thus, there is 35Vpk-pk between the outputs of your driver stage. If there is 1Vpk-pk going in, then you have a gain of 35, and since Av = gmRL, that suggests gm = 35mA/V for the MOSFETs.

The gain can't be 35 because the degeneration source resistors...1k/100=10...and with the pot in series with the source resistors the gain will be even less...

[nobody special]

Originally posted by Tube_Dude


Yes the input capacitance will be great...the gate source capacitance of the mosfet +the Miller capacitance ( the gate drain capacitance X the gain of the stage)...it will need some driving current in the high frequencys...

Mosfets are not my choice for input devices...

cheers

I don't see the problem, unless you are using a tube preamp with a fairly high output impedance. The P-channel differential pair is used in all the Pass Aleph amps with no problems, so...???
I understand that a lot of you tube guys like to use all tubes, so it could be an issue.
I will most likely be using it with a zen balanced line stage, which will drive it with no problems.
So, then (I ask myself) why use tubes? It's different. I haven't seen anybody doing it this way. Maybe there's a reason for that, haha. I'm going to give it a go anyway.
Thanks for all the help



Steve

[nobody special]

Originally posted by Tube_Dude


The gain can't be 35 because the degeneration source resistors...1k/100=10...and with the pot in series with the source resistors the gain will be even less...

yeah, you're exactly right! I forgot to look at that when I added the balancing circuit.

[EC8010]

Originally posted by Tube_Dude
The gain can't be 35 because the degeneration source resistors...1k/100=10...and with the pot in series with the source resistors the gain will be even less...

Good point. The simulation that gave the 1V sensitivity must have been run before the 100R source resistors were included. Still, the degeneration will reduce the Miller capacitance, which is handy.

[nobody special]

how about this...
it gives a bias voltage difference of +/- 3.68, and the sensitivity is reduced in half to 2Vp-p for 37V grid-grid (Vpeak) out.

a good compromise?

I'm not sure how much of a voltage difference the output stage would need to see to compensate for imbalance.