There was a passing fancy for so-called "damping factor controls" about 50 years ago. I think some of the Marantz amps had them as well. Most other amps from the era and earlier aimed for a low (for the time!) source impedance. The only amp I'm aware of that actually aimed for constant power wasn't commercial, it was published as the White Powertron. Interesting design and the first audio use I'm aware of for the cross coupled phase splitter.
OT, but where could I find an output impedance vs. frequency curve for your OTL circlotron amps? If I ever end up with suitable speakers, it's a design I'm quite interested in trying.
I think that manual control was very good for amp owner to optimally suit one single amp to various type of load(speakers) .🙂
Last edited:
well whilst doing the paradigm thing lets look at the typical 60s valve amp I have. 3 sets of output taps (and 3 different feedback settings to match those). So you can match the output devices to deliver 35W into 16, 8 or 4 Ohms. Any variance in that impedance of course changes the FR. That to me is a constant power type of view, driven by the limitations of valve technology.
Or do you disagree?
Or do you disagree?
EV used them as well as Heath. The fancier amps were usually the ones that had some sort of damping control; Fisher call theirs the 'Z-Matic'.
On our amps impedance vs frequency looks very much the same as there is no output coupling cap or transformer. A suitable speaker, if you want to try a smaller amp, would be 8 ohms in the bass and maybe 90 db or more (depending on room size). The impedance curve does not have to be flat as long as the dips are reasonable or if the designer used a tube amp in the design of the speaker. 16 ohm speakers work quite well even with our smallest amp (30w/channel).
I do. First off the idea of constant power is that the amp will make the same power regardless of load. The reason for doing this, especially in the 21st century, is to eliminate colorations caused by distortion (which the ear converts to tonality). You can't get flat frequency response no matter how hard you try; so instead get rid of the tonal colorations since the ear gives distortion a hefty weight in the scale of actual FR errors as opposed to tonality induced by distortion. IOW your viewpoint of the topic as expressed appears to me somewhat limited, if not outright myopic.
Your amplifier employs a feedback scheme to help it work with various impedances and in so doing is trying to be a voltage source, and not a very good one based on your comments. Something to understand is that in the 1960s the industry had not yet made the transition to the voltage paradigm and was still wallowing about. By the early 1970s though the rules were pretty well in place. The last thing to change was how speakers were rated- the power paradigm measurement is of course 1 watt/1 meter; the voltage paradigm is 2.83V /1 meter.
So the earlier tube amps of the 1950s and 60s tended to be inconsistent with respect to how much feedback was applied- and thus some were true voltage sources and some were not (and some had it sorted out in the 1950s and others were still scratching their heads in the late 1960s). Its helpful to understand also that no-one at the time thought "Oh, we have to get rid of this old power rule and go with the new voltage rule"; it was much more gradual as the industry sought a standard for insuring plug and play frequency response. I'm old enough to have been along for the ride such as it was and have serviced hundreds of amplifier designs so I've seen this up close and personal.
We all know that amplifier/speaker matching is an on-going conversation; well guess what- in this conversation we are discussing its roots.
There are a lot of designers that don't think the power rules are obsolete yet, which is why there are still SETs and horn loudspeakers very much in evidence in high end audio; there is still the disparity between amplifier specs on paper and how an amplifier sounds. The amp specs emerge almost entirely from the voltage rules; obviously since you can't tell how the amp sounds from the specs they don't have a lot of meaning. This is because some pretty basic human hearing/perceptual rules (some of which we have known since the mid-1960s) routinely get ignored by the audio industry in its pursuit of the almighty dollar.
All tube amps have significant output impedance and are therefore affected by varying impedance in the speakers. Look at any set of measurements of amplifiers into a simulated loudspeaker. For 50 years ago it was pretty damned good and about 1/10 the distortion of some of its competition.
By changing the taps you change the impedance that the output stage sees which changes the damping factor by definition. So why is the long dead 'z-matic' any different?
As for myopic there are plenty of studies showing we are more sensitive to FR than distortion. So forgive me if I am going to need a bit more persuading. If as you say there are some known perceptual rules you can give us some references other than marketing copy. I am very open to new ideas, just not david icke style wierdness.
By changing the taps you change the impedance that the output stage sees which changes the damping factor by definition. So why is the long dead 'z-matic' any different?
As for myopic there are plenty of studies showing we are more sensitive to FR than distortion. So forgive me if I am going to need a bit more persuading. If as you say there are some known perceptual rules you can give us some references other than marketing copy. I am very open to new ideas, just not david icke style wierdness.
Hi Billshurv !
Read article -Critical damping- from my post #60 , I hope you will find the right answers to your questions .
Read article -Critical damping- from my post #60 , I hope you will find the right answers to your questions .
Nope. I just see someone working for a speaker company in the days before T-S parameters and acoustic suspension existed trying to come up with a way of making the driver behave in an open baffle. Thanks to the last 60 years we don't have to do that anymore.
Sorry if I am missing some long lost audio grail but seems to me this is a 1950s solution to a 1950s problem. And if you like vintage it's possibly the way to go.
If I am being thick and missing something obvious due to my poor vision then please enlighten me.
Sorry if I am missing some long lost audio grail but seems to me this is a 1950s solution to a 1950s problem. And if you like vintage it's possibly the way to go.
If I am being thick and missing something obvious due to my poor vision then please enlighten me.
I don't quite understand what you mean here. Are you saying that the impedance is constant from 20Hz-20kHz? And what sort of resistance is it?
But I cannot control the amp if I cannot amplify the nasties and feed them in opposite phase to the amp. One way or another, they HAVE to go through the amp.
And that is precisely the reason for the spectral growth mentioned before.
Jan
Last edited:
I don't follow you. Output impedance is a specific definition which has nothing to do with the technology that causes it. If you have a black box, there's a simple way to measure Zout.
It is not necessary to know what is in the box, a tube or ss amp, a SE, no FB or high feedback, whatever.
Of course the NUMBERS will depend on what's in the box, but not the definition and not the measurement method.
A definition is that, a definition.
Jan
Last edited:
There can't be two different definitions of output impedance. At least one of them must be wrong. If an output stage (or any other circuit) has a sufficiently linear response that output impedance is a meaningful concept then there is just one output impedance at a given frequency. This is not a matter of which 'paradigm' we are using.atmasphere said:
How you calculate the impedance may vary with different circuit architectures, but the result must be the same. Likewise, there are several ways to measure output impedance but they must give the same result. If not, 'output impedance' is undefined - presumably because the stage is too non-linear.
No. If you want a low output impedance then it is absolutely vital that you do "amplify external artifacts". It is these very "artifacts" which help define the speaker impedance. Part of the purpose of the feedback is to control the speaker via electrical damping. In some cases this can be more important than simply linearising the amplifier.Fast Eddie D said:
Last edited:
Billshurv
I think that many solution but not quit all from 50` and even from earlier ages is quit valid even today , same as in case of vintage stuff many modern solution have come,stays for a while and then goes on the junk yard of history, only the best solution which can bring the listener close as possible to live sound can stays forever.
As RTV professional repair man(TV mechanicus) with experience of over 25year I succesfully repaired thousand and thousand of broken TV sets , but hundreds and hundreds of modern SS based `hi-fi` amps , with few very rare exception most of them sounds like crap to me .
Yes I like many vintage stuff ,I collect vintage gear and parts whatever is possible , but I have No Any Dogma against Any modern stuff , whatever which new solution is hiding inside of that modern gear , again only and only than if that modern gear sounds good to my ears .
About your enlightment , sorry but I can`t help you .
Guys, if only you could see yourself through thy eyes of well intentioned observer! He would probably say: "There is no real knowledge in this audio electronics area, only a plethora of opinions."
It is the knowledge I am seeking. Given no speaker is a purely resistive load, and no speaker is particularly linear, how does 'constant power' work? I can't see how it can. Current drive I can understand if the speaker is suitable for that, but no one can explain, other than to tell me I am unenlightened/myopic.
More interesting is whether it makes any (positive) difference if the nasties are multiplied in sequential steps instead (or only for the voltage gain part).
Fact is that none of those concepts have stood the test of time, not the low cost integrateds, nor any of the costly power amp specimens as e.g. Audiolabor Stark, Pioneer M7, etcetera.
If you read what Siegfried Linkwitz did with KEF in the 70-ies you realise that these people just started to find out what the hell was going on in speakers.
Jan
True, but if you talk about several ways to create a particular output impedance, that is pretty simple to characterize as to what comes closes to what you want. All simple engineering to do as well.
Where it gets hard is to specify wich Zout you want for 'best' sound.
Jan
The nasties I refer to is trash coming in through the output wires, like back emf and RF. You do NOT want to amplify that!
An amplifier with good damping and impedance characteristics will control this. You can certainly design an amplifier with conventional global feedback that is not susceptible to these artifacts. Furthermore, the Zobel network on the output is supposed to shunt RF.
The feedback is to control the amplifier, period. It must correct distortions produced by the stages of the amplifier and their interaction.
The spectral growth is most definitely not caused by external artifacts. It is a function of the distortion of each individual stage, compounded by the next stage, and skewed by the negative feedback. It is an unfortunate thing that we must design around and strive to mitigate.