What differentiates good audio xmfrs from bad ones?

Frequency response delivering full power is what I think. Some trick winding process is likely to cost more.

Greg

Some things:

Quality of the core material. Care and bulk of wiring. Frequency range to be covered. Power requirements.

After which, like anything else, all about materials wound on the thighs of virgins on Walpurgis Night.

Peter Wieck Melrose Park, PA

The capacitance of the windings will limit the high frequency response. Perhaps the best will have the secondary wire in parallel with the primary wire. Never had one to take apart.

Paul

The core of a transformer has laminations, because a solid metal lump would have electrical conduction losses (eddy currents) to cause large heat losses (and reduce output efficiency). A power transformer has thick laminations (60 Hz losses are small anyhow, but at 60 kHz, that would be inefficient), and a switchmode transformer is made of high-electrical-resistivity ferrite (at 60 kHz, the size can be small, though such a low-magnetization material wouldn't be great for lower frequency power).

Audio transformers are most efficient when made from very thin laminations of soft iron, which is a more expensive construction than is used for either

The parasistic elements. (copper, iron...).

Thanks for the edifying reply. Some time back I wanted to use the mains power as a frequency standard when adjusting a generator. I was worried about connecting my 'scope directly to 125 volts so I used a 12 volt out step down xmfr. The wave form was really distorted. So I guess this was a perfect example of the xmfr being made in a way that was unsuitable for audio. Thanks Again, Eric

That can happen because of saturation and remanence: the addition of a load on the low voltage side (like, a light bulb) will help greatly. The best small-signal audio transformers are potted in protective shells, in order to prevent strain on the magnetic laminations (which makes the magnetic core do those nonlinear things).

Thin laminations of silicon steel is the norm.

.... Phil

.... Phil

From a guy with a Ph.D who did a bunch of magnetism studies. When unloaded, power transformers have lots of undesirable signal-handling features.

Has he got a web site - I could do with a good laugh.

Give up - you have NO clue about audio or mains transformers.

... Phil

I had no idea! But I should have. So if I had loaded the xmfr down some the wave form would have more closely resembled a pure sine wave. Because of the tube amp I bought I have got the bug a little to learn about audio stuff in general and tube stuff in particular. I ordered a B&K Precision 4100 function generator last night and when it gets here in a week or so I'll be hooking up the TEK 465B to an audio xmfr, and some others, to see how the waveform gets distorted with frequency, voltage, and waveform. Thanks, Eric

Core saturation is not a function of load "TO A FIRST ORDER APPROXIMATION".

Only the primary voltage matters.

But to a second order, if the transformer core is near saturation unloaded, when you load the secondary, the primary current and the primary resistance will in effect LOWER the primary voltage and reduce the core saturation.

So considering the second order effect of primary winding resistance, yeah, loading a transformer _may_ reduce the distortion a little.

mark

The source impedance of a pentode or beam tube output stage is pretty high, triodes are somewhat better.

Only those amplifiers with large NFB ratios avoid the issue.

..... Phil

Has he got a web site - I could do with a good laugh.

Maybe he has a minor in AP. AudioPhoolery.

yep good point...

so it is very possible for a tube amp to have a bit less transformer saturation distortion when loaded compared to unloaded.

i.e. the distortion can go DOWN when loaded.

NO !

Saturation is a value coming from the magnetic core, yes.

But H (magnetic strength) is directly proportional to the current, not to the voltage, at constant frequency. The load is very important. This leeds the core to the satruration point if too high.

A free transformer (without load) rarely comes to saturation.

Eddy's current (we call it Foucault's current in France) are proportional to I^^2, like copper losses.

The applied AC voltage and frequency are what matter.

Others here have it right cos they know that they are talking about - you don't.

Piss off you bloody idiot.

.... Phil

Or, you know, you COULD just look this s*it up and learn how it works.