Odd/Crazy output transformer design?

[smoking-amp]

Another variant would be to use a thin bobbin of square or rectangular (large dimension) shape and just stack a zillion small snug fitting U-I laminations over it. If one could find "long" U lams with an extra wide bottom to the U, one could skip the I lams, easier to assemble. Or use a bunch of small diam. cut cores fitted over the bobbin like a string of beads with ty-wraps. Could even try the common ferrite cut cores, but it takes like 4 or 5 times as much core area using ferrite as compared to steel. Make it the same overall size as the amplifier chassis and it can fit in that typically unused space around the inside corner/edge of the chassis. Then some set screws tapped into the corners of the chassis could be used to tighten up the lamination stacks to maximize Mu and minimize core vibration.

[Tyimo]

Smoking-amp!

The ultimate output transformer is also quite easy too make using a toroid winder machine with progressive wind capability. No interleaving required, no magic or mumbo-jumbo OT voodoo, simple as making a peanut butter sandwich. You can set the machine up (typically programmed control), and return when its all done. No labor. Just a single progressive wind layer for each winding across the full core with good insulation between windings.

Could you explain to me how to do that progressive wind?
For example on a PP OPT?

Greets:
Tyimo

[LinuksGuru]

Originally Posted by Tyimo

Could you explain to me how to do that progressive wind?
For example on a PP OPT?

Progressive winding used to decrease layer to layer capacitance. For example: 10 turns forward, 9 turns backward (see picture attached). Another benefit of this kind of winding that it is possible to distribute very smoothly large number of turns in single section.

However, properly designed audio frequency toroids don't need it. In other words, its possible to avoid this increase of complexity without sacrifice of product quality.

[LinuksGuru]

There is also another ill-fated idea for audio output transformers - R-core.

R-Cores are invented by Kitamura Kiden for power transformers. They built from continuous strip of metal (without gaps), have small circular cross-section area (relative to length/width) optimized for 50/60/400 Hz, and not 20 Hz. Moreover, R-cores have 2-bobbin design, which is not good for audio.
O-Cores. Made by the same non-cut technology as R-cores but shaped as doughnut (O-core looks like toroid but is not toroid !). Suppliers for these cores are scarce, and application. for audio-frequency transformers is very questionable.

[Tyimo]

Progressive winding used to decrease layer to layer capacitance. For example: 10 turns forward, 9 turns backward (see picture attached). Another benefit of this kind of winding that it is possible to distribute very smoothly large number of turns in single section.

Thanks LinuksGuru!
Now I understand it.

However, properly designed audio frequency toroids don't need it.

Could you tell me what do you mean on "properly designed audio frequency toroids?"
For example designed and calculated for 20Hz?

Tyimo

[LinuksGuru]

Originally Posted by Tyimo

Thanks LinuksGuru!
Now I understand it.

Could you tell me what do you mean on "properly designed audio frequency toroids?" For example designed and calculated for 20Hz?
Tyimo

1) Lowest frequency 20 Hz;
2) Leakage inductance <= 20 mH;
3) Ability to sustain +/- 7mA imbalance of output tubes idle current without LF drop at max and low signal level/lowest frequency (20 Hz);
4) Software simulation of transformer behavior and winding layout;
5) Rigorous testing of finished unit in real amplifier circuit (not just generator/oscilloscope).

[Tyimo]

Thanks!
What about fully symetric winding for the primeries? Like bifilar winding?
Tyimo

[LinuksGuru]

Originally Posted by Tyimo

Thanks!
What about fully symmetric winding for the primaries? Like bifilar winding?
Tyimo

Do you mean balanced or bifilar windings? This is not the same.

Bifilar one made with 2 wires simultaneously, usually in CFB/unity coupled transformers to minimize phase shift and thus, avoiding possible oscillation.

Balanced made on 2 bobbins, or on one bobbin separated in 2 equal segments.

None is required for "standard" audio output transformer with single center-tapped primary and single secondary. Although 2 parts of primary will have slightly different DC resistance (approximately 10% or so, usually 15 - 30 Ohms), this is nothing compared to full primary impedance (several KOhms) and can be neglected.

[Matt.B.H.]

Still working on the Pi wound output trannys. Its just taking longer than expected due to house moves twice

R-Cores could be used in output trannys. the windings would need to be balanced either side the same as with a single c-core. Obviously they would only work for PP and would not tollerate any curent imbalance. Or they could be used single ended either parafeed or with a winding with DC on, servo controlled to track in opposition the DC in the primary.

[Tyimo]

Do you mean balanced or bifilar windings?

I mean bifilar windings. (made with 2 wires simultaneously)