Transformers

[gbono]

I'm starting a project and have been researching sources for PT and OPT. Anyone familar with torroids and do they really give an advantage over EI trannys? I also do not see any safety/regulatory ratings on products from the cheaper transformer suppliers. What's up with that?

Lot's of good info on torroids here:http://www.plitron.com/

[Enzo]

COmpanies that make toroids really like them.

Toroid PTs work just fine, though their power on surge is higher.

Toroid OTs work, but can't handle even small DC currents. In other words a SE toroid OT wouldn;t be good.

[gbono]

Plitron's site has several torodial designs for SE applications but they are all under 15W. What's the isue with DC current and a toroid?

[Enzo]

Have you read all of Plitron's tutorials and ap-plication notes? I think they go into limiting DC through their toroids. Better knowledge than mine exists on this board for toroid technical stuff. SOmething about the DC current saturates the core.

Plitron used to sell a book - can't recall the title, and the author's name was something Scandinavian I think - about designing tube amps with toroids. I have it somewhere, and I found it very interesting when i read it.

[Albert Kreuzer]

"Modern High-End Valve Amplifiers
based on toroidal output transformers"
by Ir. Menno van der Veen

The name is Dutch BTW

Cheers,
Albert

[Enzo]

That's it. Dutch? Sorry, no disrespect to the Dutch. I was a little too far north I see.

[Steve Conner]

As far as I know, Plitron's toroidal SE OTs have an airgap cut clean through the core with a diamond saw. This allows them to tolerate DC, but it nullifies all the advantages of the toroid over a regular EI core. The only reason Plitron do it is because they specialize in toroidal transformers: they want donuts with everything, as it were.

P-P toroid OTs have no gap and are very sensitive to DC imbalance. I've seen tube amp designs that had servo bias circuits with op-amps and suchlike, to get rid of the DC offset.

[Robert M. Martinelli]

Originally Posted by Steve Conner

As far as I know, Plitron's toroidal SE OTs have an airgap cut clean through the core with a diamond saw. This allows them to tolerate DC, but it nullifies all the advantages of the toroid over a regular EI core.

P-P toroid OTs have no gap and are very sensitive to DC imbalance.

Hi Steve,

Toroidal transformers are more difficult and more expensive to wind and build, I don't get why they cut that airgap....The main advantage of toroidal transformers vs. conventional ones is they don't have any loss in the magnetic circuit, so I don't understand why they produce ( with greater effort and expense ) toroidal transformers and then do something that produces losses and puts them on par with conventional ones....If they do it only to render them able to tolerate DC, I find it pointless, as I wouldn't pay more to have the same performance of a "normal" transformer, I'd just throw a "normal" transformer in and save some money for my next build....( the same thing many people around the world do, after all )

Am I missing something, and, if so, can you enlighten me? Is this another "marketing guys" trick?

Regards
Bob

[Enzo]

One tremendous advantage of toroids is not electrical. They are not tall. You can build a tube amp in a 2-space rack. Small ones even in a 1-space.

And you cannot make a SE amp without DC in the transformer, so there is that to consider.

[R.G.]

I designed transformers for a living for about a hundred years from 1917-1979.

The biggest thing to remember about transformers is that ALL TRANSFORMER DESIGN IS A COMPROMISE, FROM START TO FINISH.

Ferromagnetic (iron-family based) materials have hugely better magnetic properties than air or vacuum, so their use is mandatory for all low frequency transformers (below about 100kHz).

Iron also is conductive, so the same magnetic fields that make the transformer work make currents flow around in the iron. This amounts to induction heating, and can burn out the core. So iron cores are always subdivided up into flat leaves of iron which make the currents flow a long way to reduce heating. A toroid is the simplest example of a core made from thin strips of transformer iron. Every air gap is from about five to twenty or more thousand times worse at conducting a magnetic field than iron, so if your transformer has a core that is the equivalent of, say, eight inches in magnetic length, an air gap of 800 MICRO-INCHES might halve the magnetic "conduction" of the core. This is again a place where a toroid with no clear air gap is a winner.

"E" shaped cores have two air gaps per lamination no matter what you do; so do cut C cores and other cores. So toroids are inherently more efficient magnetically than pieced cores. Another way of saying that same sentence is that toroids are inherently easier to saturate than pieced cores. They're the same sentence, just looked at a different way.

For the same power level, temperature rise, and frequency range, toroids are about half the weight/mass of an E-I core. They saturate much more easily and draw much higher inrush currents because of the "perfection" of their magnetic cores. Toroids and DC excitation are mutually antagonistic. So a single-ended OT made from a toroid is almost a contradiction in terms. A toroid that's been cut in two places is known as a cut-C core or a C-C core. They have air gaps, don't saturate as easily, and lose the no-gaps advantages.

Because push-pull OTs count on the cancellation of DC currents in their primary windings, toroidal OTs are super-sensitive to DC balance on their output tubes, and need either some kind of servo mechanism to enforce DC cancellation or the introduction of air gaps (along with doubling size and weight) to withstand the DC imbalance.

I also do not see any safety/regulatory ratings on products from the cheaper transformer suppliers. What's up with that?

For AC line connected transformers, that's taking a huge chance. They're just not paying the $10K - $100K for type approval from safety labs to get certification. In the USA, there is no federal law requiring them to do so.

[gbono]

R.G.

Thanks for the lesson on transformer design - nice explainations. One other benefit of a toroid is that they do not appear to generate acoustical noise - 60 Hz buzz. I assume this is due to the "tight" core. Can you explain why a toroid appears to radiate less emi than a EI transformer?

Type approval isn't a requirement in the US except many distributors, Costco, Best Buy, etc, will not sell your product without it. Type approval appears to be a must have in the EU and Asia.

[Robert M. Martinelli]

Originally Posted by R.G.

For the same power level, temperature rise, and frequency range, toroids are about half the weight/mass of an E-I core. They saturate much more easily and draw much higher inrush currents because of the "perfection" of their magnetic cores. Toroids and DC excitation are mutually antagonistic. So a single-ended OT made from a toroid is almost a contradiction in terms. A toroid that's been cut in two places is known as a cut-C core or a C-C core. They have air gaps, don't saturate as easily, and lose the no-gaps advantages.

Thanks R.G.
So, the picture I had in mind was correct, after all.....the "cut core" thing is crap, as it makes the use of a toroid "useless" from an "electrical" standpoint, nullifying the advantages....( except for their lesser height, as Enzo wisely noted ).

I'll keep on using toroidal PTs and stick to "conventional" OTs.

Best regards
Bob

[Steve Conner]

Hi all,

Toroids radiate less EMI because they are more "perfect". The closed core, due to its high permeability and lack of air gaps, keeps almost all of the magnetic flux locked inside it. Maybe 99.95% of it.

If properly wound, the windings are also symmetrical and the stray flux from them would cancel even if there was no core, an advantage that can't be had with an EI geometry. This takes a little subtlety though, to avoid the "single turn effect". Essentially you have to wind half of it and then wind back the way you came. I once made my own toroid transformer to supply 50A at 12V, not knowing about this, and it leaked enough flux to send CRT monitors crazy from 3 feet away.

I also had similar thoughts about toroidal OTs. I'm not scared of the servo bias circuit in itself, but I sometimes wake up at night in a cold sweat, worrying about the extra low-frequency pole it introduces into the NFB loop. Well, maybe not that bad and of course it doesn't matter if you're using triode output tubes without any NFB.

Norman Koren's TENA is a nice example of an amp with toroid OTs and servo bias. He punted the NFB loop issue by just not using overall NFB.
http://www.normankoren.com/Audio/TENA.html