Hi. I've never posted to this forum before, and perhaps I'm making a mistake by replying to a long dead thread, and doing so by bringing back a topic that the thread drifted away from even longer ago. You'll all think I'm an idiot, for sure. But, although I enjoyed reading the entire thread quite a bit, I wasn't sure if anyone earnestly addressed the original subject. Maybe that's because some of the answers are kind of common knowlege. Whatever. I'm going to throw in my two cents anyway.
chessrock said:
This is just sort of a general question/topic . . .
I've been noticing a trend lately in a lot of the audio gear that I like. It's a sonic thing. Anyway, I'm kinda' digging the sound of stuff that has transformers. I don't know why that is . . . and I realize it's a very general statement because there are obviously going to be a lot of different factors that make something sound good, etc. etc.
I think that one reason that transformer balanced stuff, especially newer stuff on the market (that's appeared since electronic balancing became a reasonable possiblility), tends to sound good is just because of the simple fact that transformers are so expensive. An audio quality transformer is generally going to be the most expensive single component in a device, and by a huge margin. If a designer is serious enough to include one or two, that means it's not a device that's designed with budget as the first priority in mind. So the superior sound quality isn't necessarily due to the transformer, but the presence of a transformer is more like a symptom, in this case of a rigorous, no compromise design. That said, certainly there a lot of very high quality designs that don't use transformers, such as Sytek's very nice instrumentation type front end, it's just that usually if there's a transformer in there, you can sort of reason that other corners were probably not cut.
I'm just going to throw this out there, but to anyone out there with an opinion . . . how much do think the Xformer contributes to something's sound? What are your favorite Xformers and how do they differ from one another?
When I listen to something with a Jensen in it, for example, it tends to sound pretty transparent -- like why do they even bother? 
With other stuff, I can definitely hear something going on -- and it isn't always a good thing, either. Sometimes, I think I can hear it actually doing something kinda' nasty. For anyone who might know: Does a lot of the sound have to do with what kind of transformer is being used; or does it have to do with how the Xformer is implemented in it's design? I'm assuming it's probably a little of both.
I know that some people these days are really into a transformer "sound", and I 'm willing to acknowlege that there is such a thing. Still, I think the best transformers shouldn't and indeed don't have a sound. A good transformer is able to pass a signal with less coloration than the usual coupling alternative in low impedance chains -- large microfarad value electrolytic capacitors.
When they do have a sound, correct me if I'm wrong, but I think it breaks down in two basic ways.
1) High frequency resonance. We habitually think of frequency response curves as sagging downwards at both ends of the rating, falling at 20 Hz and falling at 20kHz or whatever, but almost all transformers' upper rated limits are actually bounded by a RISE in frequency response rather than a fall (the rise is a resonant peak that's quickly followed by a fall). This HF resonance, as we all know, causes transformers to create a ringing when hit with transients. Depending on the frequency and Q of the resonance this can seem to either add grit or buzz or a pleasant sense of air. Resonance problems are oftem eliminated with low pass or band reject network across the transformer's secondary, or sometimes simply by the Miller capacitance of a triode in the first gain stage.
2) Low frequency distortion. If you send an AC signal through a transformer and measure it's peak or even RMS value with a simple volt meter, you'll usually find that low frequencies are represented much better than the transformer's response rating. That is, the transformer claims it's only good down to 50 Hz, but at 10 Hz you've got peak to peak output that's nearly identical to 100 Hz. This is because a transformer's low frequency cutoff rating is usually bounded not by absolute level, but by distortion. Somebody said that transformers are much less likely to clip than amplifiers. True, transformers aren't usually going to exactly clip, but they do saturate nevertheless, and also exhibit distortion due to hysterisis, or the inertial magnetic qualities of the iron. Both of these types of distortions are more significant at low frequencies. Both modes of distortion are harmonic, which means that the transformer adds higher frequencies of sound to the signal in sympathy with the low frequency content. I think this is probably why people often say transformers can add low end punch. It's easier to hear those added higher frequency components than the very low stuff, but your ear still associates them with the lower frequency fundamentals...the bass seems bigger without actually increasing in absolute level. And you don't necessarily hear it as distortion because the low midrange all the way to the top isn't distorted at all.
Input transformers often have high winding ratios -- stepping up the voltage of the signal a lot -- and are often designed with minimizing size as a priority. As a result, I think they tend to create the problems mentioned above more often than other types. High ratio inputs designed for tubes often exhibit MUCH worse high frequency ringing than other types. For example if you compare an Altec/Peerless 4722 with a 50,000 ohm secondary to an Altec/Peerless 15095 with a 15,000 ohm secondary you see a much bigger rise, even though they're made by the same company and (I presume) wound on the same core. The 4722 goes for more money on the used market than the 15095 because it's "TUBE", but I'd rather use a 15095 as a tube input, because what it sacrifices in gain, it makes up for in clarity and balance. That's the same reason Jensen doesn't even make a transformer with a typical 50,000 ohm or higher "tube" secondary, but instead recomends the JE-115 with a 15,000 secondary.
Probably my favorite mic input transformer was made by Haufe in west germany. Haufe made the input transformers for the Telefunken V72 and V76. They have a highish ratio of around 1:12 (300 to 50,000), but have just about the smoothest, most precise but also extended high end of any high ratio step-up. I'm not just saying that because of the way the Telefunken modules sound either, by the way. I have a bag of about thirty of these Haufe inputs sitting on my shelf, and have been using them in various circuits. They're amazing. There are a few other extremely well designed and built german transformers with similar remarkable qualities (including the tiny little Beyers that Ampex used in their tube preamps) but the Haufes are the best of them IMHO. Although they're really only designed for mic levels (you'd have to pad line levels before the primary for best performance), I'd take them any day of the week before a UTC A-10, or the more commonplace (and excellent, of course) Jensen JE-115.
Well, I've been writing for pretty close to an hour. What an ass. I guess I'll stop now.
The sample reate does not alter our perception. It may fuck up the sound, but it doens't alter our perception of it.
