Reasonable rail voltages for an Aleph-X do not readily lend themselves to getting decent performance out of tubes. Yes, I've seen 6922s run with 30V or so (e.g. Sunfire preamp), but the transfer curve will look like a snake and distortion will be much higher than if you had used a more reasonable rail voltage...say 100 to 150V.
Obviously, that's getting a little high for the MOSFETs in the output stage.
A much better way to approach this would be to run the front end tubes properly, then cap couple to the outputs. The only problem to be solved would be setting the bias in the output, but that's not an insurmountable difficulty.
This idea comes up about once or twice a year. A number of threads have been started for either Alephs or Aleph-Xs with tube front ends. I don't remember ever seeing one completed, but maybe I just missed that particular thread.
Note that an Aleph--with its proportionately higher rail voltages--would be a better candidate if you intend to directly connect to the outputs. If you decide to go with DC blocking caps, you can use a cathode follower. Even a 6922, a fairly robust tube in terms of current, is marginal for driving a gang of MOSFETs. You could do something like use one tube as the front end differential, then use another with both sections in parallel as a follower. The differential can easily drive two followers and you'd be able to get good current and a lower Zout.
Grey
Obviously, that's getting a little high for the MOSFETs in the output stage.
A much better way to approach this would be to run the front end tubes properly, then cap couple to the outputs. The only problem to be solved would be setting the bias in the output, but that's not an insurmountable difficulty.
This idea comes up about once or twice a year. A number of threads have been started for either Alephs or Aleph-Xs with tube front ends. I don't remember ever seeing one completed, but maybe I just missed that particular thread.
Note that an Aleph--with its proportionately higher rail voltages--would be a better candidate if you intend to directly connect to the outputs. If you decide to go with DC blocking caps, you can use a cathode follower. Even a 6922, a fairly robust tube in terms of current, is marginal for driving a gang of MOSFETs. You could do something like use one tube as the front end differential, then use another with both sections in parallel as a follower. The differential can easily drive two followers and you'd be able to get good current and a lower Zout.
Grey
Given that the grids will be at ground potential and the plates some 100V (or more) north of there, I'm not sure I see how you intend to make that come together without DC blocking caps.
Even assuming that you can, it'll be a dicey proposition. You absolutely must regulate both sets of rails and tie them together with a resistive divider. Without regulation, if your line voltage drifts, the tube rail will drift much, much further than the solid state rail. The bias for the output will be thrown off with cataclysmic results. The only reason the Aleph style amps bias so nicely is that the front end and output share a common reference at the negative rail. Anything that upsets that balance is going to make life difficult.
The problems can be solved, but you need to think each step through carefully before even trying to breadboard a test circuit.
Grey
Even assuming that you can, it'll be a dicey proposition. You absolutely must regulate both sets of rails and tie them together with a resistive divider. Without regulation, if your line voltage drifts, the tube rail will drift much, much further than the solid state rail. The bias for the output will be thrown off with cataclysmic results. The only reason the Aleph style amps bias so nicely is that the front end and output share a common reference at the negative rail. Anything that upsets that balance is going to make life difficult.
The problems can be solved, but you need to think each step through carefully before even trying to breadboard a test circuit.
Grey
I was thinking something along these lines. There's one more device in the signal path but I don't think it would change distortion much. Could even try source follower instead on output stage instead of common source. Not really Aleph-X-like anymore, but still would have the X feedback and the Aleph current sources.
An externally hosted image should be here but it was not working when we last tested it.
Hey Nixie, there's free, easy to use, schematic drafting software available at Express PCB. Program works like a charm, has most of the electronic components too. It offers a feature that allows you to create and save your own parts as well.
Newbies, like me, need clear drawings in order to comprehend and possibly, learn.
John
Newbies, like me, need clear drawings in order to comprehend and possibly, learn.
John