Litz quadfilar vs 'standard' 2623

This inquiry may belong in 'The Chamber', but I don't know of anyone having compared the Litzwire 2623's yet with the standard bifilar, (trifilar?) non-vintage ones. I ask because I'm gearing up for a large purchase of output tx's  (40+) and was wondering if anyone has anything to say about them.  Jeff over at CAPI says users have commented positively, hearing increased highs and lows from 'Putting on the Litz' (couldn't resist).  If you have an impression of them, I welcome your input!

Thank you!

I have not compared, but those transformers should have a response over 50khz (100khz ?)  Hard to believe someone could hear that.
But of course I can be wrong.

No of course we cant hear 100khz , but we may still be able to percieve a difference in the sound ,Rupert Neve certainly thinks so and makes most of his modern gear with flat freq well out beyond 100khz . One definate benefit of wide band width audio is that any high frequency phase deviations will be well outside the audio range .Its nice to have the extra wide flat response ,but of course it very much depends on source material ,say in the case of mic'd electric guitar even a poor response at 20khz will make very little difference .

The main point of response to 100kHz or so is to make sure that 20kHz is actually flat. If you have 5-10 stages in a row, such as in a mixing console or a chain of several devices, the small rolloffs can add up. For one amplifier at the end of the chain, such as home listening, you can get away with a lot. Still, there's no reason to destroy the HF response if you don't have to.

In the context of a modern recording chain, where there is a limited number of xfmrs, the difference between Litz and non-Litz (all other factors being equal) should be negligible, hardly measurable, and non-significant in practice. There is one drawback to the use of Litz wire, not meantioning the cost, in particular, due to the increased insulation, the actual effective wire area is smaller, which increases DCR, that, in turn, is detrimental to LF response. The claim of offering extended both LF and HF response seems dubious to me.

Thanks for your thoughts gentlemen.  I'll get a bit of both I guess and report back some time from now. 

I suppose a tell is that no one wants to officially go on the record as claiming (showing) how they sound better, prolly because as you say abby the difference is negligible.

Litz does seem to make a difference at radio frequencies when your building antenna's which are a kind of transformer  ,but Im not sure if theres any benefits to it in an audio transformer , CJ's probably the best man to ask.As Abbey said its going to have higher resistance than the equivalent magnet wire diameter , your still going to need the same number of turns to make your transformer and maybe extra insulation ,so it sounds like a compromise to use it .

there are three ways to wind this xfmr, one is with Litz wire, which is very expensive by the way, the second is with twisted wire, 2 and 1/2 twists per inch, and the third is to simply lay down four pieces of wire as a flat ribbon.

since all three will have the same amount of turns,  then the inductance at the low end should be the same.  DCR is pretty low to begin with due to the lack of turns, so small deviations should not mean much.

as far as the high end,  those transformers should be flat to 1 Mega Hertz.

how do they do this? capacitance. somebody figured out how much capacitance you need from the side by side wire to make this happen without a resonant peak somewhere. 

but there is a trade off. as the transformer goes from being an inductor to a capacitor, phase shift happens.

so at the freqs around 100 K Hz and beyond, you will see up to 180 or more degrees of phase shift (remember that in an inductor current lags voltage by 180 degrees and in a capacitor current leads voltage by 180 degrees)

so no free lunch, but who can hear a 180 degree shift at 1 Mega Hertz? 

all three winding methods seem to have about the same frequency response,  the twisted method can be done in a warehouse with a twist drill,  just run 4 wires  down to a tie off point, walk back to the front, hook the four wires to a hook and twist away. may put some  Chubby Checker on the juke box while your at it. 

Few years ago I experimented  with "twisted 4 wires" and it  was OK. My main problem was to find 30+m space so the wires can be twisted properly It was winter with lot of snow. Next time I will do that in summer time!

Thanks CJ! Nice lesson.

As a TOTAL aside, I read today that humans had no name for the color 'blue' until roughly 1000 years ago. Even then, it was rarely used in literature until fairly recently.  Apparently the sky was white. The sea? It was 'wine colored'. We could not 'see' that which we could not name.  I suppose the same will go for HF response; Once our brains are augmented with tech, 100kHz may be part of our songs before long...  c כ

Edit: Ref link:    https://www.sciencealert.com/humans-didn-t-see-the-colour-blue-until-modern-times-evidence-science

Blue might have been about the rarest pigment going uptill a certain point in history ,prussian blue's accidental discovery meant even the average muck savage could have it ,royalty still like their blue though.

Then theres the ten to one rule with frequency, if you want flat at 20khz let 200khz be your -3db

Tubetec said:

Then theres the ten to one rule with frequency, if you want flat at 20khz let 200khz be your -3db

Click to expand...

Actually, 200kHz BW results in 0.1dB flat at 30kHz.
That is valid for a system with one sigle dominant pole, giving a 1st-order response. There are many examples of systems that have a higher-order response, where the "flat" response is much closer to the -3dB point.
That's the case with many broadcast equipment.
Not mentioning such extreme cases as reconstruction filters in D/A converters, where the 0.01dB point is less than 0.9 times the -3dB point.