Hybrid amp -> tube for VAS stage

[rtarbell]

I'm currenlty designing a solid state amplifier with a VAS stage that looks like the attached image.

Note the common emitter and the following cascode devices are paralleled. I didn't want to do this, but according to simulations, leaving them as single devices will over-power them. My rails are +/- 140V (I know, that's a lot). Due to the dynamic biasing at the base of the cascode, the power dissipated by the common emitter and the cascode will be about equal (both have about equal current runnning through them, and both divide the output voltage evenly amongst themselves).

Here's my question:

I don't want to parallel the common emitter and cascode stage, is it possible to replace the common emitter and cascode devices with tubes? I was looking to replace them with pentodes (due to their higher output resistance which I need to preserve the high voltage gain of the VAS, and due to tubes having a naturally higher voltage and power rating than small signal transistors). I know that +/-140V is a lot for solid state amps, but it is somewhat small for tube biasing. Could this be done?

Thanks for your input!

[Eli Duttman]

I don't fully understand your SS circuitry. So, no comment in that dept.

There are tube types that are FINE at modest voltages. The 6GM8 is a LOW voltage twin triode that can't take 35 V. on its plates. The 6922 is a popular twin triode that's "typically" operated with 90 V. (less works) on its plates. Small signal pentodes, like the EF86 and 6AU6, work well at lowish anode voltages too.

Both pentodes and cascoded triodes should be passively loaded. Single triodes work VERY well with constant current source (CCS) loading. CCS loading allows the extraction of the tube's full mu (amplification factor), while also maximizing linearity.

How much gain is required in the VAS? Is a topology change acceptable to you? Is 100% DC coupling a must?

Use TDSL to get at tube data sheets.
TDSL

[rtarbell]

Thank you for the feedback!

I wanted to ask: I've looked on the internet and I've found several sites that state pentodes should not be loaded with a CCS, this I can totally agree with, but I don't exactly understand why. I know that a CCS will give a greater gain to the pentode, but I don't exactly see why this is a bad thing??

--> CCS loading is done in solid state amps to increase gain, and transistors don't have (many) problems with very high gains. In terms of the VAS stage with a standard amplifier topology, as long as there is global negative feedback in the amplifier, the quiescent bias point is set. Why doesn't the same hold true with pentodes?

Also let me ask: is it okay to load a common cathode pentode with a solid state cascode, and then load the cascode with a CCS? I would think it would be because the pentode sees a low impedance load, and thus the gain of the pentode itself is very small (the gain is produced by the cascode).

Thanks for your help!!

[Eli Duttman]

Pentodes and FETs are constant current devices and the Norton equivalent circuit is appropriate for them. If you CCS load, you find yourself in "head butting" situation where the gain device and the CSS are in conflict.