Bride of Zen cascoded

[knutn]

I am thinking of using the Bride of Zen as a combined preamplifier and voltage amplifier to drive my power buffer (with no voltage gain). Therefore, I have to place my 50k volume potentiometer before the BoZ. I am afraid the high frequency impact is too large with this potentiometer without cascoding.

So, has someone tried a cascoded version of the Boz?

 

[Zen Mod]

consiidering that you have 30V at output node , make that zener 18V , re-arange bias circuit for lower one

in any case, lower mosfet will work better with more Uds

 

[knutn]

I see, is there any "sweet spot" for the lower MOSFET?

 

[Zen Mod]

I would start from there ..... and up the current later

to do that , decrease both R load (1K) and Rs , to roughly maintain gain and voltage in output node

then fine tune voltage in output node with bias voltage

 

[juma]

...So, has someone tried a cascoded version of the Boz?

I did, and it didn't bring me much happiness either...
I suggest attached circuit:
Input buffer (Q1,Q2) gets you rid of the input cap and isolates the pot so you don't have to worry about it.
Amplifying MOSFET (Q3) works with local Schade feedback in triode mode, loaded with CCS (Q4,Q5).
Zout is about 150 R .
Gain (about 20dB) is set with R7/R5 ratio and half of the PSU voltage at Q3's drain is set with R6 (depending on Q3's Vgs/Id characteristic).
This circuit has much better PSRR than original. It will deliver more than 20V_peak of clean swing at the output and is linear up to almost 100kHz.
Of course, it can be further improved in many ways - first thing would be to use 2sk2013 instad of IRF610 for Q3.

 

[knutn]

The reason I wanted to use the BoZ was to make a 2 stage power amplifier without any feedback loops and compare it to my earlier much more complicated amplifiers.
The solution with a source follower input instead of the cascode can be an option, of course, though.

 

[flg]

This might be an option?

 

[knutn]

I have simulated Juma's circuit (without the JFETs), and I must say I am impressed. The topology is very similar to the Zen amplifier. So I think I will give it a try.

 

[juma]

Yes, Schade's "trioding" is great, but it comes at a price: the input resistance is determined by the value of R5 and it needs to be low - otherwise the frequency responce will be compromised in upper range and that's why the buffer (Q1, Q2) is needed.