Thank's for the link..I will take a look..
Regards
M. Gregg
Most hybrid amps seem to just stick tubes on one end and SS on the other end of the amp. Cut and paste essentially.
My idea of what a hybrid amp should be, would be to combine tube and SS into a single stage of the amp for at least for the output stage. For example, using the SS part as output of a Darlington stage with a tube driver. The idea is to use the SS part for the current gain, and the tube (triode mode) for voltage gain, for driving it, and still provide the triode feedback function for the whole stage.
So for Darlington, the cathode drives the gate or base connection, and the plate connects to the collector or drain, so it includes local triode feedback control from the output.
Another combination would be to use a tube with a SS cascode above it.
With a triode on the bottom, this could include feedback to the cascode gate or base terminal.
Then there is the Sziklai pair with a tube input and PNP or P type follower for the plate to output. Once again the tube (triode) plate sees the output (almost) and provides gain control feedback.
Some other ideas around are to put a damper diode in series with the source terminal of a Mosfet, so as to modify its gm to emulate a tube. Or for bipolars, put the damper diode in series with the base terminal.
For pentodes, one can put a SS follower on the plate with a feedback network from the source or emitter back to the screen grid.
These schemes all end up making hybrid devices that still act similar to tubes. Then any standard amp circuit can be built up from them.
My idea of what a hybrid amp should be, would be to combine tube and SS into a single stage of the amp for at least for the output stage. For example, using the SS part as output of a Darlington stage with a tube driver. The idea is to use the SS part for the current gain, and the tube (triode mode) for voltage gain, for driving it, and still provide the triode feedback function for the whole stage.
So for Darlington, the cathode drives the gate or base connection, and the plate connects to the collector or drain, so it includes local triode feedback control from the output.
Another combination would be to use a tube with a SS cascode above it.
With a triode on the bottom, this could include feedback to the cascode gate or base terminal.
Then there is the Sziklai pair with a tube input and PNP or P type follower for the plate to output. Once again the tube (triode) plate sees the output (almost) and provides gain control feedback.
Some other ideas around are to put a damper diode in series with the source terminal of a Mosfet, so as to modify its gm to emulate a tube. Or for bipolars, put the damper diode in series with the base terminal.
For pentodes, one can put a SS follower on the plate with a feedback network from the source or emitter back to the screen grid.
These schemes all end up making hybrid devices that still act similar to tubes. Then any standard amp circuit can be built up from them.
Member
Joined 2009
Paid Member
Very interesting so far, some interesting ideas from Smoking Amp too. A few points I noticed from reading around so far:
1) there have been a lot of good reports (not just this thread) of combining a tube pre-amp with voltage amplification, feeding a SS power amp. Sometimes with global feedback, sometimes only with local feedback. In fact tube + SS seems to be the dominant hybrid topology with some nice amplifiers such as: Moskido, Vilfort Audio, Moscode, Tenor mono, Thorens TEM 3200, Wavebourn Tower, Apex, Croft and Counterpoint etc.
2) There are very few examples of SS front end with tube and OPT at the output end. The ARC is one commercial example of good repute.
3) Many of the hybrids using SS outputs are push-pull and so there is a need for a good phase splitter and for well matched output devices. An all tube output has the benefit of an OPT for recombining the output halves to create the complete waveform. Perhaps it takes a lot more to create a good push-pull hybrid with SS output than a simpler SE output (e.g. wavebourn tower).
4) MOSFETs seem to dominate hybrid designs. Yet BJTs remain very popular in all-SS amplifiers. And the few hybrids using a circlotron topology seem rank very highly (e.g. Thorens)
1) there have been a lot of good reports (not just this thread) of combining a tube pre-amp with voltage amplification, feeding a SS power amp. Sometimes with global feedback, sometimes only with local feedback. In fact tube + SS seems to be the dominant hybrid topology with some nice amplifiers such as: Moskido, Vilfort Audio, Moscode, Tenor mono, Thorens TEM 3200, Wavebourn Tower, Apex, Croft and Counterpoint etc.
2) There are very few examples of SS front end with tube and OPT at the output end. The ARC is one commercial example of good repute.
3) Many of the hybrids using SS outputs are push-pull and so there is a need for a good phase splitter and for well matched output devices. An all tube output has the benefit of an OPT for recombining the output halves to create the complete waveform. Perhaps it takes a lot more to create a good push-pull hybrid with SS output than a simpler SE output (e.g. wavebourn tower).
4) MOSFETs seem to dominate hybrid designs. Yet BJTs remain very popular in all-SS amplifiers. And the few hybrids using a circlotron topology seem rank very highly (e.g. Thorens)
Last edited:
Member
Joined 2009
Paid Member
I have found that tube preamp plus SS amp is not nearly as nice as having tubes in both locations; in fact, I think that the amp contributes more to the "tube sound" than the preamp does. By tube sound I mean sonic holography, and ambient information retrieval.
My Midlife Crisis amps use a MOSFET as a plate load for the 6E5P driver and to supply grid current to the 833C output tube. Would that be considered "hybrid"? I continue to be amazed at the sound of the system with these amps in place; far, far superior to the preamp/SS combo I had before, especially in the area of ambient cues and natural decay of notes. I wouldn't go back.
My Midlife Crisis amps use a MOSFET as a plate load for the 6E5P driver and to supply grid current to the 833C output tube. Would that be considered "hybrid"? I continue to be amazed at the sound of the system with these amps in place; far, far superior to the preamp/SS combo I had before, especially in the area of ambient cues and natural decay of notes. I wouldn't go back.
If you substitute "Solid State" for "class D" then I certainly agree. People in our DIY Club have been trying different combinations of Triodes and exotic SS output devices (including a few not available to the public). My impression is that the amps retain a bit of the characteristics of both the triode and the solid state devices. My best characterization of the sound depends on my mood...a sort of half-empty or half-full thing.
The amps that we've been playing with are SE Triodes feeding (fairly exotic) SS devices which are also being run in class A. So far, the sound has (to my ears) been a sort of averaging between the sonic attributes of both devices. The amps have increased power to drive the speaker, as a result of the SS output devices, yet retain a fair (but not total) amount of the characteristic Triode "sound" or sweetness, which I quite enjoy most of the time, but can be somewhat overwhelming at other times. The Hybrid Amps also retain the SS advantage of being able to handle Bass Drivers with high resonance peaks, which usually spells Sonic Death to a SE triode driven system.
This is certainly not the only way to create a (perhaps better) all-around amplifier, but it's not bad.
Best Regards,
TerryO
Last edited:
The reference was the same class-d amp with his opamp preamp active.
Last edited:
Is there a switch to .. umm.. switch between OP-AMP and Tube input quickly? Or did the tube went in after some work to remove the opamp? I personally find it difficult to compare sound signatures if they're not instantaneously switched. Having a switch and ensuring both produce the same output level makes sure that you can cross off "builder's expectancy" from the list of possible comparison biases.
Member
Joined 2009
Paid Member
I've found that if a switch isn't available that I can still get a good impression of differences providing I listen to a couple of songs long enough for my brain to get accustomed to the sound of amplifier A and then when I switch over to amplifier B my brain will register that something is different and start to tell me how. One thing I've found is that it's not as easy to hear an 'upgrade' but when I swap out the better sounding amp for a worse sounding amp my brain very quickly tells me something is 'wrong'.
Member
Joined 2009
Paid Member
I think the answer for me is yes. I once heard a SS amp vs a tube amp on YouTube. Many laughed (on this forum) at the prospect of comparing them on Youtube but they did sound different to me. I could not describe how they were different, it was rather subtle, but subjectively at some unconscious level I clearly preferred the sound of one of them over the other.
I build the amp 4 times, with different changes in power supply(SMPS or linear)and opampstage/tubestage. And did mutiple blind listen sessions with a total of 30+ persons. And almost everybody could hear the difference between the versions. Not everybody likes the same setup.
I can not tell with is best. Only the one i like most.
I did a little experimenting with this topology, using a power opamp to drive a 1626 in SE A2:
Class-A2 and a power opamp driver
I did also try P-P, using two opamps, one configured as inverting and the other non-inverting. It worked as expected. It would be interesting with a small dual pentode like a 6360 or 832 and low plate voltage, pushing some grid current.
This showed promise, but like so many things, I never had time to really perfect it...
Pete
Member
Joined 2009
Paid Member
Hi Pete,
Looks interesting. What drew my interest to the SS input tube output topology was the gushing reviews written about the ARC amplifiers these past few months. And these amps use negative feedback all over the place (including cathode feedback from the OPT) with some clever cross-coupling they also claim no tube matching required - so no romantic notions of tradition here.
Looks interesting. What drew my interest to the SS input tube output topology was the gushing reviews written about the ARC amplifiers these past few months. And these amps use negative feedback all over the place (including cathode feedback from the OPT) with some clever cross-coupling they also claim no tube matching required - so no romantic notions of tradition here.
Last edited:
If I would use the amp only for clean acoustic guitar sound, I'd go all SS with a compressor/Limiter to keep it from ever clipping (and a great digital reverb).
If you ever want to have it distort, for effect, I tried many different topologies and did quite a bit of research and eventually concluded that no single distortion mechanism by itself sounds good to my ear. It's the way several distortion mechanisms combine that makes the kind of bluesy distortion I like (as opposed to heavy metal severe distortion). The guitar amps I finally built for myself that I love the sound of are all tube, no feedback anywhere, flat cone speakers (not ribbed - some ribbing is OK - so they breakup sooner but do it in a much better sounding way - Webber offers the choice at reasonable prices). The interaction of the output transformer with the output tubes and speaker is a big part of good distortion, and is next to impossible to model and do with transistors. Transistors are too non-linear to ever have no NFB, but tubes are linear enough, unless you're trying for the utmost in fidelity. If the drive to a PP tube output stage is 1dB off balance, the 2nd harmonic distortion product will dominate. Tubes have such loose and inconsistent tolerances that this is virtually inevitable, unless high negative feedback screws it up by correcting it. The 2nd harmonic distortion product (same note an octave higher) is always a good thing unless it's so excessive that I.M. distortion is too high. Some audible I.M actually sounds good (ZZ Top - Well Dressed Man).
My schematics can be seen on my hobby website:
Bob's Website
If you ever want to have it distort, for effect, I tried many different topologies and did quite a bit of research and eventually concluded that no single distortion mechanism by itself sounds good to my ear. It's the way several distortion mechanisms combine that makes the kind of bluesy distortion I like (as opposed to heavy metal severe distortion). The guitar amps I finally built for myself that I love the sound of are all tube, no feedback anywhere, flat cone speakers (not ribbed - some ribbing is OK - so they breakup sooner but do it in a much better sounding way - Webber offers the choice at reasonable prices). The interaction of the output transformer with the output tubes and speaker is a big part of good distortion, and is next to impossible to model and do with transistors. Transistors are too non-linear to ever have no NFB, but tubes are linear enough, unless you're trying for the utmost in fidelity. If the drive to a PP tube output stage is 1dB off balance, the 2nd harmonic distortion product will dominate. Tubes have such loose and inconsistent tolerances that this is virtually inevitable, unless high negative feedback screws it up by correcting it. The 2nd harmonic distortion product (same note an octave higher) is always a good thing unless it's so excessive that I.M. distortion is too high. Some audible I.M actually sounds good (ZZ Top - Well Dressed Man).
My schematics can be seen on my hobby website:
Bob's Website
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.