Sound Quality Vs. Measurements

[dvv]

Originally Posted by nigel pearson

It might be . I recently built a zero feedback amplifier which has 0.03 % THD at 0.63W . It was better than that down to 50 uW . It used the opposite distortion of a pentode and triode . Pre distrotion I suppose . If all triode it had considerably more distortion ( 5 % full power ) . The amplifier made the Old DIN4550 just . If I remember correctly 0.1% THD at 1.6 W . It doesn't seem much to shout about , however for a zero feedback SE design it is very good . 2 valves and 0.56V in for 8V out .

I am not comfortable that it is a low distortion design . It looks like low distortion . The sound is rather good . I notice multi-tracked voices like never before , especailly if the same voice . I have been offered some 211 valves to see if a few more watts can be had . Bass is not the greatest . I hope to try it with Quad ESL 63's soon as they do not require high damping factor . I certainly feel the all triode version was inferior . Along the way I tried Schade feedback , I have doubts about it . I would liken the effect of my design to that of Ultra Linear feedback although this is not feedback . Have I become a valve convert ? No , if anything the other way . I certainly have enjoyed doing it .

Nige, the prototypes I am working on for myself (still two, a single ended input version and a fully complementary version, I still can't make up my mind which is it going to be), both have a THD of less than 0.15% at 1 kHz, full nominal power (28.3 Vrms) into 8 Ohms, and less than 0.35% 20...20,000 Hz, both with no global NFB. By 50 kHz, this is 0.7%, but the power output is down to 21.5 Vrms.

Now, some will say fine, but that's the simulator - true, these are the simulator provided figures, but I do believe in them because that simulator has never lied to me, or let me down. But I agree, the proof of the pudding is a live model which can actually do that.

However, in my view, more important than low THD figures are stability (without which we have nothing), a good harmonic decay behavior and as symmetrical as possible slew rate (i.e. as similar as possible rise and fall times, 10-90% and 90-10%).

Just my 2 cents' worth.

[nigel pearson]

Hi DVV . I suspect it has some feedback mechanisms ? Super spec .

As I understand it any transition resistance device will have a curved voltage transfer characteristic . A tiode is much like a transistor if enhanced to a pentode form . The triode has internal feedback is the simplistic way to explain it's linearity .

My amp has cathode resistors which do cause some degeneration . They are to limit current . If I was to use fixed bias I could eliminate that doubt . I am fairly certain I would still have the low measured distortion from an amp with no obvious negative feedback .

The big deal is that triode- triode is high distortion as usually seen . Pentode-triode is low distortion . I can even make mostly second harmonic doing that if accepting a higher overall distortion . Negative feedback is doing the same thing . The difference being that the negative feedback is slightly time delayed and not the curve that the device has . I have to say on balance global negative feedback is our best friend as it gives low output impedance as a bonus . I have found valves and global feedback seldom doing nice things . Having built this amp I feel I completed my training I started 38 years ago . That was plotting a pentode curve in my exams . Never touched a valve since except to do a few repairs . I now feel qualified to talk when they are discussed . Not super qualified I admit .

As far as I can see no amplifier can have low distortion and zero feedback . Most people mean zero global feedback , that's different . My amp might be as close as one can get . With a 211 I might get 5 watts 0.1% THD by this route and 20 W before too nasty . Class D may well be different .

[dvv]

Originally Posted by nigel pearson

Hi DVV . I suspect it has some feedback mechanisms ? Super spec .

Of course it does, for a kickoff, its input stage(s) use degeneration to cut the gain down to 6.8:1 or so, 220 Ohms emitter resistors, 1.5k collector. I don't see how one can make an amplifier with literally zero feedback, no local NFB, no global NFB.

As far as I can see no amplifier can have low distortion and zero feedback . Most people mean zero global feedback , that's different . My amp might be as close as one can get . With a 211 I might get 5 watts 0.1% THD by this route and 20 W before too nasty . Class D may well be different .

That's what I said a few posts above. To me, the idea is a perversion of the generally sane idea that it's good to reduce global NFB and increase local, however, to really work well, even that has to be carefully balanced and judged. My own proposed figure of 26 dB global is a tentative value only, a very general guideline, not a rule carved in stone. Quite simply, my experience is such that this figure is about right, but how you reduce distortion to the point where that is enough global NFB is up to you.

Then again, as you know, I have my own topologies, the ones I like and am comfortable with - change them for others and it's back to the drawing board. Give me weird, Sony-like topology, and I need at least a month to even understand it to the point where I can intervene. For HK, I need five minutes to determine which version is it from about three general topologies they used or use, and I can do things with it, for better or for worse.

As for the THD figures you quoted at low power levels, I must say mine are usually completely beyond reliable measurement, i.e. below 0.001%. That's because I like to run my output stages relatively "hot", with a bias current of about 120...140 mA per transistor. Remember I use never less than 3 pairs, and often 4 pairs, so my sub-1W power is well within pure class A operation and is supported by at least some global NFB, yielding vanishing distortion figures.

BUT, also remember this - I make such things for myself only, they are NOT commercial products, so literallly ALL of the production run constraints simply do not exist. Most folks here, who work for the industry at large, are far from being that lucky, and I would say that I believe they are often enough forced by price constraints to use less to much less of what they'd like to use.

I use 7 lb heat sinks by default, and if there's need, I'll just throw in another heatsink per side for good measure. I'll turn a stereo amp into a monoblock in a jiffy if I feel like it. I'll throw in as many capacitors as I think it should have, not thinking about costs for even a second. To me, it's all fun and games, but to them, it's professional work, it's business, an entirely different ball game.

And make no mistake, I respect their achievements more than I respect my own, precisely because I know most of the constraints they had to deal with to get there. It's much easier to be good with no contraints.

[AndrewT]

Originally Posted by dvv

.......................... a classic bipolar differential pair should have an overall gain of no more than 11:1, and no less than 5:1, and the in-between values would show up the bipoars at their very best, .............

I have seen this in the past and some kindly Members have given formulae to determine this.
But here I have sinned. I have not saved, nor remembered the details.

Could you give your method of determining the component values with an example calculation?

[nigel pearson]

DVV . Have you ever tried using one of you amps with a gain > 100 ? It would be to use a passive preamp on tape decks like Nakamichi . I did it once and loved the outcome . A friend found it too noisy with 100 dB sensitivity speakers ( what was that about ! ) . Distortion was still very very low as would be yours . You have a damping factor of 6 so do not need even 26 dB of feedback ( 10 dB ? ) . I must make that amp again , I gave my last one to a lad who worked for me . I wanted to make a volume control for it that eventually reduced the feedback as volume increased . The control would be part conventional and part feedback relaxation . At no time would the amplifier produce unacceptable levels of distortion ( < 0.05% THD with nice harmonics ) . Damping factor would stay at > 10 . It would have an override so as to use an active preamp . I would liken the experience to Guitar amps . Yes there are a few minor issues when not listening to music . When the music plays that is forgotten .

I can imagine most amps have an optimum feedback arrangement . Like the very good suggestion yesterday of an op amp with gain 3 sounding worse I am sure most amps have a bad and good point that is closely spaced .

[dvv]

Originally Posted by AndrewT

I have seen this in the past and some kindly Members have given formulae to determine this.
But here I have sinned. I have not saved, nor remembered the details.

Could you give your method of determining the component values with an example calculation?

Och man, I already confessed to a crime the same as yours, that of not saving good stuff, and I am now haggis meat myself.

Actually, we have a form of haggis down here in Serbia, as well. Similar, but not the same. But similar enough to help me feel like a distant cousin when visiting Scotland in 2009. LOVED IT!

[bcarso]

Originally Posted by nigel pearson

As far as I can see no amplifier can have low distortion and zero feedback .

Translinear circuit - Wikipedia, the free encyclopedia

[dvv]

Originally Posted by nigel pearson

DVV . Have you ever tried using one of you amps with a gain > 100 ? It would be to use a passive preamp on tape decks like Nakamichi . I did it once and loved the outcome . A friend found it too noisy with 100 dB sensitivity speakers ( what was that about ! ) . Distortion was still very very low as would be yours . You have a damping factor of 6 so do not need even 26 dB of feedback ( 10 dB ? ) . I must make that amp again , I gave my last one to a lad who worked for me . I wanted to make a volume control for it that eventually reduced the feedback as volume increased . The control would be part conventional and part feedback relaxation . At no time would the amplifier produce unacceptable levels of distortion ( < 0.05% THD with nice harmonics ) . Damping factor would stay at > 10 . It would have an override so as to use an active preamp . I would liken the experience to Guitar amps . Yes there are a few minor issues when not listening to music . When the music plays that is forgotten .

I can imagine most amps have an optimum feedback arrangement . Like the very good suggestion yesterday of an op amp with gain 3 sounding worse I am sure most amps have a bad and good point that is closely spaced .

I have seen the concept, Nige, even used it on a few occasions. However, it has not stuck to me. I suppose mostly because I am used to building preamps with 50W power devices at their output, delivering a stunning vibrant bass line if required that no passive device I ever heard (a total of 6) ever really captured my imagination.

A good friend makes his living by making passive preamps and selling them all around Europe for something like € 2,500 (app. $ 3,200), and he's been trying to get me hooked for years, but sorry compadre, no go. To my ears, passive devices somehow take some of the brute energy out of the signal, and I just plain hate that.

It could also be a placebo effect for all I know, but whatever the reason, they don't do it for me.

[dvv]

@Nigel

I forgot ... While it's of course possible to give an amp a closed loop gain of >100, I NEVER do that.

If you look over the range of what is considered to be High End today (Krell, Levinson, Accuphase, etc), you will be able to see that most of them operate with a closed loop gain of 25...30:1, rarely more. Obviously, the greater your input signal, the less overall gain you need to get it up to what you need. Therefore, their input sensitivities for nominal power output range from approximatels 1V to 2.5V.

I find that to be quite resonable for two reasons. First, literally ANY active premp worthy of that name should be able to deliver that kind of voltage no problemo, and at least twice more over and above that, including all op amp preamps; if memory serves, the lowest maximum output voltage I have ever seen was 6.5 V, and most will do better than that. Second, the nominal power amp output voltage is, in many cases, way over what is actually required - with a gain factor of 30 and an input sensitivity of 2.5 V we are talking powers of 700+ Watts per channel into 8 Ohms, or in our buddy Wayne's case, 5,600 Watts into 1 Ohm. While entirely possible, that kind of power does not come cheaply, you do need a pretty hefty power supply and a fine line of output devices to get there.

I suspect John may have a comment or two on this matter, this is right up his alley.

[dvv]

Originally Posted by bcarso

Translinear circuit - Wikipedia, the free encyclopedia

Brad, I did take a look, but I won't pretend I understood everything perfectly. My reasoning is: if one has one transistor driving another and another, and we assume each transistor has a gain of only 50:1, by the end of that route the signal would have been multiplied (50x50x50) 125,000 times.

If not, then we are using some form of degeneration, or local feedback, along the way, no matter how exotic it may be.

If you are better acquainted with this principle, please correct me if I'm wrong.