transformers/output impedance

160562 .... you are wandering in the same technical mind (pun intended) field that I found myself. What Kijanki said is spot on. You may find it helpful to retrace the steps I have taken to avoid the land "minds" I stepped on by pulling some of my posts.

One interesting post in particular dealt with "back impedance matching." IOW, output trannies work both ways. The output tranny steps down output tube voltage and impedance characteristics to the match speaker load. At the same time, a speaker's impedance characteristics, which vary as a function of frequency, are stepped up on the primary side of the tranny. This is the load the output tube "see."

As Ralph (Atmasphere) and Al (Almarg) have explained, output tubes operate best (i.e., low distortion and power delivery) if presented with a load (impedance) that falls within the tubes "sweet spot." Since speaker loads vary as a function of frequency, the output tubes may operate outside their sweet spot. If so, power delivery and distortion may suffer to some degree.

The optimum solution: if using a tube amp, (i) select speakers with a ruler flat impedance curve that corresponds to the output taps (e.g., 4, 8 or 16 ohms; preferably 4 or 8), (ii) )zero phase angle characteristics and (iii) high sensitivity. Good luck with finding a speaker possessing all of these characteristics at the same time ... and that sounds good.

Bottom line: life is about compromise and trade-offs. Ce la vie.

Thanks Kijanki. As I have come to appreciate, there are so many design considerations and trade-offs going on, at some point, amp design is about acoustic philosphy and art. Ergo why Ralph (Atmasphere) opted for the OTL amp topology.

As a consumer, though difficult, IMO, it is incumbent on me to at least understand enough of the theory so I can make as an informed a choice as possible when matching components. Admittedly, it has been a verticle learning curve, marked with many missteps and dead-ends.

Cheers,

Gotcha Kinjanki. I'd love to hear one of Ralph's OTL amps, but part of my verticle learning curve of bad choices is that I bought speakers that were designed and voiced to be driven by a low output impedance, high power, high current SS amp. I'm dumb-arsed lucky that my ARC Ref 150 has a low'ish output impedance, especially off the 4 ohm tap, and has a large power supply (1040 joules), so it can deliver some serious juice when needed.

Unfortunately, I am dubious (albeit not certain) that one of Ralph's amp would be a good match for my speakers. If Ralph thinks differently, he should of course chime in and correct my misconception.

Cheers,

Ok Ralph .. you offered. Please do. :)

Perhaps you might also touch on the what is meant in plain English when ARC says its amp's "[o]utput stage coupling is a combination of “ultralinear” and Audio Research’s patented 'partially cathode-coupled' topology ...." From what little I can glean from the description, it looks like some variant of local feedback. That's just a guess.

P.S. to my last post. Was thinking to myself that if zero negative feedback is so disadvantageous, why would top flight companies like ARC use it???

This is just a guess ... so here goes. Perhaps simply stated ... NF is used to reduce the output impedance and increase DF for **marketing** purposes.

As Ralph and Al have explained many times already, the world of tube friendly, true Power Paradigm speakers is not as large as Voltage Paradigm friendly speakers. So .. using NF to reduce output impedance and increase DF increases the playing field of speakers that may be compatible. That's my thesis.

Perhaps a good case in point is the match with my amp and speakers. Technically speaking, my speakers are poster-child "SS-amp" friendly speakers because they have rough impedance and negative phase angle functions. Yet my ARC amp manages to muscle its way through the tough spots with a modicum of aplomb.

I am somewhat dubious that a true Power Paradigm amp having a "high'ish" output impedance would do as good a job. Not because of design, build or sonic deficiencies. But just because of the "high'ish" output impedance. At the very minimum, all other factors being equal, there will be a tipping point where acoustic coloration will outweigh reduced NF induced distortion (e.g., TIM and odd-ordered harmonics).

I don't know how most Power Paradigm amps address power supply issues. I suspect that access to lots of joules can only be a good thing when the amp is being tasked to deliver current when called upon.

This post is just a bunch of guesses. So please don't come down hard on poor little me. Perhaps Ralph who is in the business can speak to the bona fides of my surmises.

Cheers,

Al and Ralph, I reread Ralph's explanations in the other thread. I understand a little bit better why using the term "output impedance" is a bit of a misnomer. Maybe the industry could sex-it-up by referring to the attribute as "synthetic" or "apparent" or "hypothetical" output impedance.

I think the point that Ralph was trying to make is that using a meter to measure an amp's impedance/resistance at the output taps is **not** what the term "output impedance" is speaking to. Instead, it's referring to the amp's operational behavior in delivering current/power to a speaker load **as though** it's output impedance was a specified number.

Where an amp's "apparent" output impedance is close to zero, one could expect that the amp would double down current/power if speaker load is halved, but only if it is acting like a true constant voltage source, and **only if the amp is performing within its operational limits.**

And ... as Ralph said, using NF only permits the amp to perform in such a fashion. Ergo why the use of the term "output impedance" without an additional adjective is a bit of a misnomer, almost misleading.

I guess in a goofy kind of way, if a designer kept adding NF to the circuit, at some point, the NF could be greater than the current/power being delivered by the main output circuit. This may be what Ralph meant when he said that NF could also be viewed as a voltage source.

Well, maybe I still don't get all of it, but maybe a little bit more than I did before. This still makes amp/speaker matching a tricky business.