• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

P channel mosfet concertina?

While doing the calculations needed for my next amplifier (8 x GU50 push pull @800-900v)
I found that a P channel mosfet as a concertina phase splitter solved many problems at once like dc coupling, ideal previous stage operating point and high gm.
Has anyone tried it or has any objections in using a P channel mosfet like the fqpf1p50, as a concertina phase splitter?
I already have the pcb for the stereo driver made so i will find out my self when things get more relaxed with the covid madness, until then i am curius to learn about other people experience on the subject...

datasheet here https://www.mouser.com/datasheet/2/149/FQPF1P50-189292.pdf
Schematic here (not fully tested yet)
 

Attachments

  • pp_stereo_drv.png
    pp_stereo_drv.png
    43.9 KB · Views: 350
Last edited:
It's a good idea, I think you are on the right track. You can also use a depletion mode fet which can be self biassed.

I am thinking of making a pcb too, as I am going to build a PP with MOSFET concertina but I have to include class A2 driver FETs too and I’ll want to add in an adjustable bias supply.

In your schematic, can you eliminate C4, or move it to before the source follower? If the source follower drives the grid if the valve you eliminate the risk of blocking distortion.
 
Last edited:
Hi.
Thank you for your comments!
Well if i remove C4 dc voltage will go in the grid but on the other hand since the AF voltage is rather low and the dc component is close to 100 v so i can't see how the non polarized capacitor can induce blocking distortion.
In the schematic the capacitor looks like a polarized one but in reality i use only non polarized polyester capacitors.
Lastly the mosfet buffer is added as a precaution in order to drive the medium 80-90pf miller capacitance of the tube, i have not decided if i will use it or bypass it ...
The bias supply is adjustable as i use smps all over and on top of that there are four pots (R52-54, R22,24) for trimming, from a floating 100Khz 18 supply i can produce any negative bias i want.
Needless to say that the power supply must be a completely physically separate unit and at least 1m away.

Chris
 
Last edited:

45

Member
Joined 2008
Blocking distortion is not related to capacitor type but to its charge and discharge because of the grid current flowing when the power tube's control grid is driven positive.
You can tame it with a suitable RC constant if the amp is not really meant for class A2 and it only occurs at max Pout. If the amp is class A2 or it's driven into positive field substantially, direct coupling to driver or transformer coupling are the options that eliminate blocking.
 
Hi, i know this but the circuit i posted is the driver only and C4 is actually connects the tone stack to the second pre amplifier stage, at no point the grid should go positive if all calculations are correct.

For the output tube stage i use a resistor to actually reduce the grid current if any, and i always avoid grid current all together in every stage.
The reason i want to use 8 GU50 (2p x2 push pull for each channel) is not for more power but for staying in class A or AB with normal listening level.
Chris
 
I decided to modify the schematic and make the mosfet buffer a dc coupled buffer by connecting the source to the negative ~60v bias and control the V2a and V2b bias with the mosfet bias.
Initially i added it because i wanted to drive V2 miller capacitance, the tone stack has a 100k volume control at the output and this coupled with the Miller capacitance would lower the frequency response at low output levels.

I will probably add the same circuit to the output of the mosfet concertina.
Chris
 
Last edited:
I have been doing much the same thing with N channel fets. This actually started many years ago when I stuck a pair of LND150's into a 12AX7 socket expecting to get dead parts, but quite surprised to get a working amp with no circuit changes.

Here is the schematic of a little low budget 4 watt guitar amp that I made several years ago. B+ is only 165 volts so every volt counts, leaving no headroom for direct coupling into the mosfet. I adjusted the resistor values to get 1/3 of the B+ across the fet, and 1/3 across each load resistor. The THD of the entire amp is under 3% at 4 watts, and around 1.5% at 1 watt, even with a $4 OPT.....until you slam it into clipping.

The first mosfet is used for a variable load impedance on the input pentode stage. Cranking the 1 meg gain pot (R9) wide open puts the tube's load impedance well over 1 meg resulting in stage gains over 1000 and unfortunately lots of microphonic effects.
 
^ Correct. P-channel Mosfets use holes as the majority carrier. Holes have about half the mobility of electrons, so the channels need to be twice as wide to have similar conductance. That tends to make the gate capacitance almost twice as large for P-channel vs N-channel devices, though there are other fabrication parameters at work.