PCB ground planes

[abbey road d enfer]

My MC preamp PCB has besides top and bottom poured ground planes, with controlled continuity around obstructions, backed up by a multiple "ground" exposed vias, connecting fat aluminum spacers to a 2mm aluminum plate, thus unlikely to produce any significant ground potential differences.

Brute force reduction of "ground" connections is probably valid for a circuit of small size with a separate PSU.
No one in their right mind would do that for a unit combining a PSU, a power amp and a preamp, or for a mixing desk, or receiver/amp.

 

[Cqwet Dbdfte]

Yep.
Some high end cell band transceivers use single PCB with complex aluminum castings on both sides. Passive intermods and other noise have to be below -150dBc. 60Hz not much of an issue.
Remoting audio PSU is not a bad idea.

 

[CJ]

Check Vacuum Tube Valley Issue 8 for a great article on OTL tube amps
.
Output transformers that I have check usually are good out past 500 kc with minimal interleaving.

Back to original topic, somewhat, the Langevin AM 16 takes advantage of circuit board trace capacitance by running the top and bottom traces over the top of each other.

 

[abbey road d enfer]

Definitionally every impedance conversion is a transformation, and yes cores are used.
https://www.highfrequencyelectronics.com/Feb04/HFE0204_Sevick.pdf

I wasn't familiar with the terminology.
It turns out that the principle involved works when the inductors have an impedance much higher that the nominal circuit impedance, which requires a ferromagnetic core.
The typical unun xfmr with two equal windings provides a 2:1 voltage ratio (4:1 Z ratio). Then it isn't really different than a basic autoformer.
16:1 impedance ratio can be achieved with 4 equal windings, which is not too different than an autoformer with (N+3N) turns.
Autoformers have a slight advantage over normal xformers, because the current in the low-Z side is smaller than in a normal xfmr. It does not make a big difference in practice.
Now with a 4:1 voltage ratio xfmr, the primary impedance of 128 ohms (16x8ohms) is not an easy load for tubes.
And I don't think that the principle of transmission line can be implied here, with the shortest audio wavelength being about 16 miles.

EDIT: had mixed impedance ratios with voltages ratios.

 

[abbey road d enfer]

Back to original topic, somewhat, the Langevin AM 16 takes advantage of circuit board trace capacitance by running the top and bottom traces over the top of each other.

Being fully balanced from head to tail, it only makes sense.

 

[Cqwet Dbdfte]

An autotransformer is in effect sort of a transmission line transformer, 16:1 would have the square of the ratio as impedance ratio, so an 8ohm load would present ~2k on a tube stage, not an unworkable proposition, if two of these on a core, a push pull design would be another 4x plate to plate, and balance out DC imbalance.
Still a DC servo would be needed, not really super easy.

 

[abbey road d enfer]

An autotransformer is in effect sort of a transmission line transformer, 16:1 would have the square of the ratio as impedance ratio, so an 8ohm load would present ~2k on a tube stage, not an unworkable proposition,

A quadfilar unun results in only 16:1 impedance ratio, 4:1 voltage ratio, so my calculation was correct.
I did not make it clear enough in post#124. Corrected it since.
A 16:1 voltage ratio unun would just be some kind of autoformer, with the same constraints.
Of course it would provide a correct load for a tube stage.