Tetrahedron output stage topology.

[Circlotron]

For a long time I have been fascinated with the "circlotron" output stage topology because you can use N-channel devices exclusively and it is totally balanced, unlike quasi complementary symmetry. The downside though is you have to have two floating dc supplies for every output stage. Imagine a 5.1 HT setup! Anyway, a while back I saw this setup in Glass Audio and have been bursting to give it a try. The big advantage is that you can use a single supply for all stages just like in other amp topologies. I did make a small one a while back but wound up a small ferrite transformer and because of the lack of inductance the LF response was not that great, but it did work... So, on the weekend I got busy and pulled apart and rewound an old 120va tranny with 43 turns quadfilar of 1.25 mm wire. I just used what I had and what would fit. Another really good thing is that the transformer doesn't have to have any great audio qualities, it's major job is to transfer dc to the two electrolytic caps that are going oppositely up and down at an audio rate. The tranny is sort of like a double ended see-sawing (teeter-tottering for you Septic Tanks ) common mode choke. You can make it out of any old junk, just so long as the windings are identical voltage. at the moment I am shooting for 200w into 8 ohms down to 50Hz with a *single* 40v rail. Will post further pics as the project proceeds.

BTW, I refer to it as a Tetrahedron topology because you can redraw it as a perfect 3D tetrahedron with one tranny winding as the top edge and the other tranny winding as the bottom edge. CT's to power and ground. Each end of the bottom edge has a fet source and a cap negative on it. They split to opposite ends of the top edge so each end of the top edge has a drain and a cap positive. Draw it on paper. Looks cool. Great to dream about.

[mrfeedback]

Graham, good work - any running examples yet ?.
Are these for your new subs ?.

You mentioned that you wife does not like thunder - yesterday morning it frightened the living beejeezus out of me and my lady too.

......Sunday morning sleeping in soundly after a nice evening was rudely interrupted by a lighning strike, I swear directly over the top of her roof at precisely 6.21 AM.
As a reflex action both of us instantly flew out of bed and then looked at each other, still half asleep but wide awake at the same time.
You need a bit more than 200W to go near to that.

Eric / - nothing can surprise me after that awakening.

[traderbam]

Intriguing.
It looks a little to me like having two SE amps in bridge configuration. What is the drive circuit for the gates? It seems like they need to be in anti-phase.
BAM

[stelios]

my little project

Inductance loaded current amplifier

skip all the way to the bottom of the page for an accurate schematic..

I've built the circuit but have not connected the transformers yet.
I'm currently having problems with heat conduction as the transistors don't seem to couple well to the heatsink no matter what I do.
Actually, there seems not to be any coupling at all... so it's shelved for the moment. I'm using silicone pads and I think I will replace them with sth else.

I'm very interested to see how much bass you'll manage to get by winding your own toroid as this what I envisage to do myself eventually. One thing I want to know is this : do power toroids have ferrite cores ? or are they tape wound ? or iron-powder ? Is there a default core for this application ? I don't want to take them apart to find out (yet)..

regards

Stelios

[Circlotron]

Stelios, thanks for telling me about 'your little project', I'm printing a hard copy of it right now so I can have a closer squiz at it. AFAIK mains frequency toroids would all have strip wound iron/steel cores because this is the only stuff that will make enough inductance with a reasonable amount of turns for use at low frequencies.

I actually used an EI type transformer because that was what was available, and it is relatively easy to wind, although that few turns on a toroid doesn't present too much of a problem. I would have liked to use one if I had one. Yesterday I ran it up on 50Hz ac on the bench and observed the magnetising current versus voltage and decided that at 50Hz one complete centre tapped winding of 43 + 43 turns is good for about 25Vrms. Proportionally higher at higher frequencies and vice versa. 25v = 78 watts into 8 ohms, double that into 4 ohms, and would need close to 40vdc supply rail. Hopefully I'll get something going by the end of the week.

BAM, the gates are indeed driven in antiphase and biased with respect to earth. The circuit has a voltage gain of 2.

mrfeedback, probably someone at the antipode of your house has a 7 Hz Tesla coil cranked up and focussed on you.

[Nelson Pass]

Well Circlotron, you just can't stay away from these
weird topologies, can you?

You get my prize for most interesting circuit of the week.

[djk]

What a novel idea, a tube amplifer made with transistors.

If you would like an old pile of iron from Bogen and Dukane amplifiers, let me know.

[stelios]

I have to say that I find this type of circuits really perplexing.
Any links where I can find out more about its operation ?
I can't understand the purpose of the capacitors...



Circlotron:

I have also written a small spice code for R. Burfoot's circuit but unfortunately I can't get it to work. You're welcome to it though if you want. (I use Winspice).


regards

stelio

[Circlotron]

There is a patent covering the tube version dating back to the forties or something. I'll have to look for it again.

Who is R. Burfoot, the original designer?


Stelios, think of it like this (works for me anyway). Two see-saws one above the other. A reasonable space between the two. The upper see-saw with the pivot point going to + volts. The lower see-saw with the pivot point going to earth. Left fet reaches from left end of lower ss to left end of upper ss. Right fet reaches from right end of lower ss to right end of upper ss. Cap1 from bottom left to top right, and cap2 from bottom right to top left so they cross like an "X". Fets pull see-saws so they alternately touch together at left end, or touch together at right end. Touching actually means clipping. Caps are like struts that lock the motion of the see-saws together so they always move accurately in unison. It parallels the operation of the transformer and makes it behave as though it had zero leakage inductance between it's 4 windings. All 4 windings are always driven, unlike a conventional class AB cct where for a lot of the cycle only one half of the primary winding gets driven, making for poorer prim to sec coupling and inefficient use of the copper. Actually this feature is only an issue if you had the speaker driven off a secondary winding like in a high impedance tube cct; for this one you can run the speaker from a source to source connection. I think the ultimate in aesthetic symmetry would be to use a dual voice coil loudspeaker and drive one coil from the sources and the other from the drains. Another thing, seeing each drain looks into the other source, we have a current sink hooked onto a voltage source. Very happy arrangement. Niceness definitely radiates out of this circuit.

[traderbam]

"Niceness definitely radiates out of this circuit. "

You have definitely LOST THE PLOT, mate!