Seems that McIntosh made a number of tube amps years ago with a very similar topology. One difference though is that they don't use the cross coupling capacitors. The only reason I can think of for this is that the transformer has to have a 1:1 ratio between the cathode winding and the anode winding if you want to cross couple then using caps. Maybe theirs didn't? Another amp, the Quad 22 used what looks to be similar arrangement to the McIntosh but I think the cathode windings are way smaller than the anode ones and they don't contribute much to power output; their main purpose is for negative feedback.
Part of McIntosh MC-240 below.
Part of McIntosh MC-240 below.
Attachments
Circlotron,
have you seen the Atma-Sphere Circlotron tube amps???
www.atma-sphere.com
Ralph Karsten has been building Circlotron OTL tube amps for over 20 years, hold many patents. and these amps sound GREAT!
Fully balanced, super stable! In fact, one of his demos, is to go over to the amp, while its playing, put on a kitchen oven mit, grad an output tube and yank!
No pops, no noise, no smoke, nothing, just sweet sound as if nothing ever happened! AWESOME!
Ralph lives not far from me, i am working a deal to get a super rare Proto-type amp of his from a mutual friend of ours!
Now if i only lived a few miles away from Nelson Pass!
Zero
have you seen the Atma-Sphere Circlotron tube amps???
www.atma-sphere.com
Ralph Karsten has been building Circlotron OTL tube amps for over 20 years, hold many patents. and these amps sound GREAT!
Fully balanced, super stable! In fact, one of his demos, is to go over to the amp, while its playing, put on a kitchen oven mit, grad an output tube and yank!
No pops, no noise, no smoke, nothing, just sweet sound as if nothing ever happened! AWESOME!
Ralph lives not far from me, i am working a deal to get a super rare Proto-type amp of his from a mutual friend of ours!
Now if i only lived a few miles away from Nelson Pass!
Zero
Re: Overcoming transformer blues.
Hi Circlotron,
I read and re-read your thread on the Tetrahedron topology and think I have newbie's rudimentary understanding of the "see-saw" operation(although, the finer details are beyond my capacity).
Anyway, guessing from the pictures of your working amp and the schematic, I assume that the small tranfo is a phase-splitting input transformer that feeds the green-wired "output" trafo and the big white one is your power tranfo(?). Or am I missing something?
Or, does your amp require balanced input?
And how sensitive is the design to power supply noise, filtering, etc.?
Anyway, very cool and even cheaper to test out. My brother's an EE and maybe I can get him to help me understand the finer points of this design. Man, it looks like a testbench version could be made for the price of a decent dinner out.
Circlotron said:
Hi Circlotron,
I read and re-read your thread on the Tetrahedron topology and think I have newbie's rudimentary understanding of the "see-saw" operation(although, the finer details are beyond my capacity).
Anyway, guessing from the pictures of your working amp and the schematic, I assume that the small tranfo is a phase-splitting input transformer that feeds the green-wired "output" trafo and the big white one is your power tranfo(?). Or am I missing something?
Or, does your amp require balanced input?
And how sensitive is the design to power supply noise, filtering, etc.?
Anyway, very cool and even cheaper to test out. My brother's an EE and maybe I can get him to help me understand the finer points of this design. Man, it looks like a testbench version could be made for the price of a decent dinner out.
Re: Re: Overcoming transformer blues.
Amp sensitivity to noise is very low. Absolutely =no= turn-on thump into the speaker. Well, provided the cross-coupling caps are equal value of course.
I'll have to draw up a pic to illustrate the see-saw action a little better. Rest assured though it really does work.
Yep, exactly.darkmoebius said:
Amp sensitivity to noise is very low. Absolutely =no= turn-on thump into the speaker. Well, provided the cross-coupling caps are equal value of course.
I'll have to draw up a pic to illustrate the see-saw action a little better. Rest assured though it really does work.
Power output test
This experiment amp has been lying idle for quite a while and the power supply and chassis eventually got swiped for another project. I never did get around to doing a power output test so today I rigged up the main part of the cct again and applied a regulated 35vdc supply. I used a matched pair of IRF540's and set the bias so that crossover distortion just disappeared. Cranked up the signal till the peaks just started to flatten a little then backed off. distortion was in the range of 2%-3%, mostly 3rd harmonic. I had a 3 ohm and a 6 ohm load (not ideal, but that's what I had)
Results @ 1kHz
6 0hms = 33.2v peak = 88W rms.
3 ohms = 32v peak = 170W rms.
The heatsink was just that slab of aluminium visible in the earlier pictures so after about 60 seconds of this kind of power output it got to about 60 deg C. Peak dissipation in the mosfets would have been about 90w halfway up the sinewave.
I'll have to see just what power I can get into a 4 ohm reactive load now. Bump the DC rail up to 50v too.
This experiment amp has been lying idle for quite a while and the power supply and chassis eventually got swiped for another project.
I never did get around to doing a power output test so today I rigged up the main part of the cct again and applied a regulated 35vdc supply. I used a matched pair of IRF540's and set the bias so that crossover distortion just disappeared. Cranked up the signal till the peaks just started to flatten a little then backed off. distortion was in the range of 2%-3%, mostly 3rd harmonic. I had a 3 ohm and a 6 ohm load (not ideal, but that's what I had)Results @ 1kHz
6 0hms = 33.2v peak = 88W rms.
3 ohms = 32v peak = 170W rms.
The heatsink was just that slab of aluminium visible in the earlier pictures so after about 60 seconds of this kind of power output it got to about 60 deg C. Peak dissipation in the mosfets would have been about 90w halfway up the sinewave.
I'll have to see just what power I can get into a 4 ohm reactive load now. Bump the DC rail up to 50v too.
McIntosh was able to wind their output transformers' primaries "bifilar", that is, right next to each other. This removed the need for extra coupling capacitors across the ends of the windings. Your circuit, being the McIntosh topo without impedance matching, could be done the same. An interesting variation was described by Norman Crowhurst in 1960 using two separate output transformers, the extra coupling caps (of course), paralleled secondaries and the McIntosh bootstrapped drive. For valves, of course, in that era.
All good fortune,
Chris
ps: your original post was never answered. Yes, you want the same load impedance no matter what local feedback you're planning.
All good fortune,
Chris
ps: your original post was never answered. Yes, you want the same load impedance no matter what local feedback you're planning.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.