reliable tube/mosfet hybrid amp

Jump to Latest
JAZZ2250

JAZZ2250

Member
Joined 2002
#1
Hi, forum.

I was wondering why it's so hard to find DIY projects of tube/mosfet hybrid amp. Is it because hybrid stuffs cannot make better sound than all-tube or all-solid-state stuffs? What I would like to build is the one using tubes at input stage and solid-state (probably mosfet) at output stage. Any idea, experience or link to a website?

Thanks.
 
#2
Hi.
i just had the same idea some month ago. I build an amp with mosfet-power out and gain 1; sound very good; and a tube stage (srpp design) also with very good sound quality (as preamp in another amp); and: the result was very bad, until now iam not shure why, made some (30...) mods, but no top sound, as exspected...
alf
 
#4
As far a links go, TubeCad is a great site for Valves and
Hybrid Amps is http://www.tubecad.com.
In particular
http://www.tubecad.com/april_may2001/2001_04-05.PDF
http://www.tubecad.com/articles_2002/Power_Buffers/Power_Buffers.pdf

Also some other projects on the web,
http://digilander.iol.it/essentialaudio/index.htm
http://www.homestead.com/whaan/files/index.html
http://home.c2i.net/semfielec
http://web.tiscali.it/audiofanatic3/Schemi/Tipo/ibridi/finali_ibridi/finali_ibridi.html

There have been a couple of threads on this forum in the past, but not many listening reviews.
http://www.diyaudio.com/forums/showthread.php?threadid=5619&highlight=hybrid
http://www.diyaudio.com/forums/showthread.php?threadid=1405&highlight=hybrid

When time permits I have a 6922/BJT design I would like to try myself.

Regards,
Tim.
 
#5
Driving Mosfets

alfsch,
I was considering this some time ago,
I wanted to use a 12AX7 as the srpp stage, but after enquiring on this forum, was persuaded that a tube that has more output current capability is required,
apparently MOSFET input capacitance makes them hard to drive,
particularly at higher frequencies.

I'm sure someone out there can give a more technically correct explaination

Cheers,
Pete McK
 
#8
The link you posted is basically a unity gain buffer with MOSFETs with a tube voltage gain stage before it, is it not? I built a tube voltage gainstage, and am just finishing off my MOSFET power followers right now (they took me a while, spent a while not working on them), so soon I'll have some input on how it'll sound...
 
#10
Hybrid amp.

I have built several valve/mosfet hybrid amplifiers. Some with ECC82/81 tubes and others with 6SN7 voltage gain stage. The MOSFET's were Hitachi types.
The sound varied from OK to very good. Without negative feedback they have a high output impedance which affects speakers with a widely varying impedance. The output impedance can be reduced by increasing the number of output Mosfet's and/or by introducing negative feedback.

I think you should build a unit without negative feedback . That sounds good. Then introduce negative feedback and see what you gain and what you loose. It might be hard for someone to tell you how it will sound in your system.

In conclusion - build it. My amp presently sounds better than an Audiolab 8000S except in the very low bass end. On most music this does not show! Midrange on the Hybrid is much better.
Cheers.
 
#12
I used to own a counterpoint SA-100 which sounded quite good on the right set of speakers. But it was very fragile. I blew a set of output FETS, sent it to counterpoint to have repaired, got it back, plugged it in (without anything hooked to it), and blew the same fets again. I then traded it for a VTL amp.

It had power supply rail fuses too, but the fets were faster fuses than the fuses.

Good sound yes... Relable no... (not really fair; it used it for years with no problems before the failure)

So it is possible to make a good soundning hybrid, and I'd postulate that with a good quality bipolar output stage it would be quite reliable too. Or a fet stage with a better design would probably do the trick.

Sheldon
 
#13
What I heard from another forum is - Counterpoint' Hybrid amps blow the Fets - surprisingly large number of people do not seem to be aware of-
Is it a design fault and fets blow due to Oscillation ?
When we discus Hybrid amps Moscode Tube / Mosfet hybrid amps have cult following entirely due to their excellent design and reliability.
Bythe way I am looking for Moscode schematics - does anybody have them?
K.S

stokessd said:
I used to own a counterpoint SA-100 which sounded quite good on the right set of speakers. But it was very fragile. I blew a set of output FETS, sent it to counterpoint to have repaired, got it back, plugged it in (without anything hooked to it), and blew the same fets again. I then traded it for a VTL amp.

It had power supply rail fuses too, but the fets were faster fuses than the fuses.

Good sound yes... Relable no... (not really fair; it used it for years with no problems before the failure)

So it is possible to make a good soundning hybrid, and I'd postulate that with a good quality bipolar output stage it would be quite reliable too. Or a fet stage with a better design would probably do the trick.

Sheldon
Click to expand...
Moscode Tube / Mosfet
 
#14
kayes said:
What I heard from another forum is - Counterpoint' Hybrid amps blow the Fets - surprisingly large number of people do not seem to be aware of-
Is it a design fault and fets blow due to Oscillation ?
Click to expand...

I think it's more of a design flaw than a characteristic of MOSFETs. After all if FET's were fragile, then Nelson Pass' designs would be puddles of molten silicon. The Counterpoint amps don't have any small emitter resistors for load balancing. So if the output is shorted, the FET itself dissipates all the power by itself If I remember correctly, it's an avalanche type of failure too). According to Mike Elliot, the amps sound much better without them (If I recall correctly, there are provisions for them on the circuit board). I did find the amp to be a good sounding (a bit warm and soft; which was nice).

Sheldon
 
#15
I said I'd soon have some input, and I just finished off one of my amps (it's in pieces, needs the chasis to be assembled, will do tomorrow) but it really does sound nice, the tubes add their sound and the MOSFET power follower (Class-A) is indeed rather neutral (very nice sounding on it's own though) and blends well with the tubes. All and all I'm very happy with it, and will probably stick with that setup for a while to come. So MOSFET/Tube hybrids can indeed sound superb, give it a try, but with all things, a bad design will sound bad.
 
#16
A good tube/SS design is what I need. I'm so happy that you have made a success cause it will be my next project. I just didn't wanted to spend my time and money for what would never happen. BTW, could you provide some preliminary info before you prepare a full review of your work? I'm eager to know what you have made. Thanks.
 
#17
I made a simple source-follower circuit with IRFP044Ns, only one in the signal path, as well as one for the ccs (just slight variations on the power follower 99c design by Andrea Ciuffoli). Andrea reccomended a tube stage to drive it (he reccomended his headphone amp, but that sucker was a little on the expensive side to build). So I built a simple tube stage composing of a 6922 tube doing the voltage gain followed by a 6AS7G wired up as a cathode follower. The tube stage must have a low output impedence so as not to take off the high end due to the MOSFETs high input capacitance. I'm really happy with it, I'll probably tell you more when it's broken in, but for now I have to finish my second stereo unit, for these are meant to biamp with.
 
#18
NFB on the Hybrid

You should be careful about the negative feedback on the hybrid amp. The Nfb going to the cathode of the input tube. This and the inter coupling capacitor ( to the MOSFET stage) can cause a bass lift of several db in the low frequencies - below 10 hz. It shows up as increased bass and is not good as it causes very low frequency (damped) oscillation. If you control this with a low value inter coupling capacitor , the bass lift disappears and the low frequency instability seems to be controlled. It is easy to simulate this with a simple program like ECAP or if you are familiar , with Spice.
I tried this circuit and it sounds very good. I made many changes to it get rid of the oscillations. I am still not sure if it is unconditionally stable. Sounds good in any case.
cheers.
 
#20
Feedback

I tried this circuit with no negative feedback. I liked it. Mid range was nice - I cannot explain easily but it was good. Bass was a bit floppy as the output impedance was high -around 1 ohm or so. But if you crossover to a sub- or have a very well damped sub ( Q less than 0.6 ? ) it will probably be good. I plan to make an amp with no nfb with an active sub . I think this may be the ultimate answer. Without nfb I find that portions of the music that tend to 'shout' seem to be very well controlled. In addition the voice seems to be very well reproduced - more natural .
It is certainly a very good idea to try a Valve voltage gain stage with a unit MOSFET stage. You can try many configurations.
The no NFB one is good and maybe you can start with that. This is an area where 'low distortion figures' are to be largely ignored. They are probably just under 1 % or so ! The sound is worth it.
I am just hoping I get enough time to work on a no nfb design for my active Tannoy603 project.
Note that I tried the ECC series and also the 6SL7 and 6SN 7 series. The 6SN7 beats the rest hands down. For higher voltage output and more output stages you will have to use a power voltage stage. I think maybe you have to use the power pentodes in triode mode. I have bought several pentodes just for this. Haven't got time yet to try them out.
cheers.
 
You must log in or register to reply here.
Top Bottom