The best of both worlds?


I've been really interested in the hybrid world, but I can't find many designs out there that do what I want.

I've made a circuit layout of a basic tube driven mosfet amp that I found somewhere on the net. E-mail me if you want a look? (please not it's not my design so I take no credit - only for the layout).

'As is' this circuit produces 50W into 8 ohms with a THD+noise of 0.64%, which doesn't seem too bad for a valve circuit considering most of this is second harmonic (softer on the ears).

If anyone thinks they may be able to improve upon the design (using bi-polar devices instead of MOSFETs in the output?) email me for the ciruit layout.

Thanks for looking folks!
Sorry folks I was tired (it's 00:35 here in the UK), here's the diagram!

Peter Daniel - this thread could get interesting! ;)


  • hybrid amp.jpg
    hybrid amp.jpg
    22 KB · Views: 573
The distortion is probably mostly due to the mosfets. I have no experience how PP mosfets with no NFB sound but these things are not triode valves... How is the distortion at low leves, say around 1W? Are the MOSFETs matched? I've listened to commercial amps with similar topology (counterpoint) and the feeling that the distortion was just too high never left me.

Here's the promised circuit with some values...

..also I'm running this thread in parallel on the tube side of the site and I'm taking all ideas into consideration and updating the design and reposting.

Who knows maybe after a few weeks (months?) we'll have a great circuit as a testament to the co-operation of the tube loves and the solid-state devotees.

Keep the ideas coming!!

:rolleyes: ;) :D


  • hybrid amplifier.jpg
    hybrid amplifier.jpg
    27.8 KB · Views: 446
Hi Annex666

I've got the distinct feeling that this circuit wouldn't work properly(or at least with a lot of distortion).

The bias circuit around the BJT is not supplied. I think there should be connections from either side of the bias setting potentiometer to the bias circuit. This way I could imagine it would work.

I assume you must use a larger bias current than would be needed for a BJT output stage, to get low THD with MOSFETS (OTOH BJTs couldn't be driven as easily !). But then I think it would work nicely.


...sorry guys, it was really late when I made that schematic, here's the latest version with the correct bias circuit also I've ommitted the Zener circuits as the MOSFETs have them internally.

If anyone would like to elaborate on the circuit, for example designing a low-distortion output stage, please feel free to post your ideas or email me so I can integrate them into the final design!



  • hybrid amplifier.jpg
    hybrid amplifier.jpg
    25.8 KB · Views: 381


2001-02-04 4:23 am
Several thoughts: You CANNOT omit the protection diodes. The integral diodes are for static only and will burn out at 100µA gate drive. You're better off adding 220pF to the gate of the BUZ900 rather than using unequal gate resistors. Since there is no feedback there is no power supply rejection ratio. The bias/DC offset circuit shown will inject humm and noise into the gates, add a Pi filter off the main supply. Omit the diode in series with the bias transistor and the 0R22 resistors, they were for a version made with HEXFETs. The bias transistor must NOT be near the heatsink. After setting the pots, measure and replace with fixed resistors. If the wiper goes open on the bias pot the bias will go up enough to make the amp boil water. If either of the offset trim pots go open the amp will put out 30VDC. Use a speaker relay with a 30 second on delay.