John Curl's Blowtorch preamplifier part II

[scott wurcer]

Originally Posted by Bob Cordell

Hi Chris,

Anatoliy is correct about the sidebands being of unequal amplitude if both am and fm at the same modulating frequency are present. I believe I also mentioned this several posts ago. Because the upper and lower sideband phase relationships are different for am and small-deviation fm, the sidebands will augment each other on one side and fight each other on the other side.

Bob

Bob there are times when pairs of sidebands counter-rotate (opposite phase). In that case one could make an argument for maintaining equal magnitude. One should be able these days to use free FFT software and a good sound card to separate the real and imaginary parts of at least a couple of pairs. I actually don't have a uA741 handy.

[Wavebourn]

Originally Posted by scott wurcer

Bob there are times when pairs of sidebands counter-rotate (opposite phase). In that case one could make an argument for maintaining equal magnitude. One should be able these days to use free FFT software and a good sound card to separate the real and imaginary parts of at least a couple of pairs. I actually don't have a uA741 handy.

Thank you Scott; I completely forgot about symmetrical complementary designs that may hide such nasty pictures from observers. In order to force them to reveal their nastiness DC shifted test signals may be a good solution.

[john curl]

Well, I hope that a week's vacation might put us back on track.
I realize that some do not think that this is the right direction to take, i.e. the high open loop bandwidth or no global bandwidth approach. I can understand that, but it is the direction that I have found most promising.
While the Blowtorch preamp and the designs that Charles Hansen makes are of the 'no global feedback' approach, It might be interesting to look at lowish feedback designs, something like 20-40dB feedback with a minimum open loop bandwidth of 10KHz. This would be a compromise, of course, but much easier to make, and measuring pretty good, especially with regard to power amps, that do tend to distort badly at high levels, even if they sound just fine at normal listening levels. IC's might be included too, as most here cannot get the really good parts, such as Toshiba fets, as easily, anymore.
I am more interested in video IC amps than audio ones. Anyone have any input on this?

[Bob Cordell]

Originally Posted by john curl

Well, I hope that a week's vacation might put us back on track.
I realize that some do not think that this is the right direction to take, i.e. the high open loop bandwidth or no global bandwidth approach. I can understand that, but it is the direction that I have found most promising.
While the Blowtorch preamp and the designs that Charles Hansen makes are of the 'no global feedback' approach, It might be interesting to look at lowish feedback designs, something like 20-40dB feedback with a minimum open loop bandwidth of 10KHz. This would be a compromise, of course, but much easier to make, and measuring pretty good, especially with regard to power amps, that do tend to distort badly at high levels, even if they sound just fine at normal listening levels. IC's might be included too, as most here cannot get the really good parts, such as Toshiba fets, as easily, anymore.
I am more interested in video IC amps than audio ones. Anyone have any input on this?

Hi John,

Welcome back!

This is a fine direction to take, as it remains a point of controversy in the audio industry apart from what you or I think.

I think there are a couple of things we should bear in mind in the discussion.

First, it is important that we are all on the same page when it comes to the definition of high feedback and low feedback. For example, it is eminently possible to design an amplifier with 40 dB of feedback having a 2 MHz gain crossover and yet having a 20 kHz bandwidth. This is no sweat, especially if one is allowed to use local feedback to limit the VAS gain so that the open-loop bandwidth gets to 20 kHz. Would this fall into your category of high feedback?

On the other hand, we could have virtually the same amplifier with 40 dB of NFB at 20 kHz and a 2 MHz gain crossover, but with 80 dB of NFB at 200 Hz. This would probably fall into your high-feedback/low open-loop bandwidth category, but surely would measure just as well as the earlier example in terms of slew rate, TIM, THD-20 and most anything else.

I'm sure we will diagree on the matter of PIM between the two examples above. I'm guessing you would argue that the latter amplifier has more PIM due to its lower open loop bandwidth in spite of the fact that they both have the same 2 MHz closed loop bandwidth. I'm guessing you will also argue that relevant PIM cannot be measured by a procedure that we agree on (i.e., of late, you seem not to think much of the Otals PIM test).

We also need to distinguish between readily measurable differences and differences between hi/low feedback/open loop bandwidth designs that cannot be accounted for by measurements. This is where I think the greatest interest lies.

There would seem to be very little interest in you and I disagreeing on how much TIM is made by two amplifiers having the same slew rate and closed loop bandwidth, but different open loop bandwidth because that is generally simply measurable.

Just some thoughts,

Bob

p.s. - will you be at RMAF?

[john curl]

Hi Bob, I will be at the show.
For me, it is the SUBJECTIVE impression that makes me want to achieve high open loop performance. That is the easiest thing to compare.
I agree that TIM will not necessarily be changed by open loop bandwidth, with modern designs, but we have solved the TIM problem, except for the very cheapest (and most popular) IC's in most affordable audio equipment, and that won't change until the better IC's get significantly cheaper.
I had hoped that further PIM research would be taken up by others, but the confusion and the 'may I say politics?' of the last few weeks, have not been very productive, at least in my estimation. Again, I would like to target for affordable designs, by both amateur and professional alike, that sound better than their typical measurements generally predict. I personally think that open loop bandwidth is the right direction to go toward. In fact, I am pretty sure of it. Please, let this thread exhaust this direction, rather than go into some sort of yes, no, debate, much like health care is debated in the USA at this time.

[Wavebourn]

According to my subjective experience;

the difference between amps with 30 dB global feedback from 20 Hz to 20 KHz and amps with 30 dB feedback on 20 KHz and 80 dB on 20 Hz is audible such a way so the former when sounds in other room fools imagination as if the band plays in the other room; the later sounds like a nice sounding equipment sounds there. The later one may be greatly improved by shorter feedback paths added that equalizes all stages before global feedback is applied.
Edit: the key is impact of impedance (mis)matching between stages.

[john curl]

Subjectively, I think you are right, Wavebourn.

[Wavebourn]

Originally Posted by john curl

Subjectively, I think you are right, Wavebourn.

I think similarly: subjectively you are right, John.

[Bob Cordell]

Originally Posted by john curl

Hi Bob, I will be at the show.
For me, it is the SUBJECTIVE impression that makes me want to achieve high open loop performance. That is the easiest thing to compare.
I agree that TIM will not necessarily be changed by open loop bandwidth, with modern designs, but we have solved the TIM problem, except for the very cheapest (and most popular) IC's in most affordable audio equipment, and that won't change until the better IC's get significantly cheaper.
I had hoped that further PIM research would be taken up by others, but the confusion and the 'may I say politics?' of the last few weeks, have not been very productive, at least in my estimation. Again, I would like to target for affordable designs, by both amateur and professional alike, that sound better than their typical measurements generally predict. I personally think that open loop bandwidth is the right direction to go toward. In fact, I am pretty sure of it. Please, let this thread exhaust this direction, rather than go into some sort of yes, no, debate, much like health care is debated in the USA at this time.

Hi John,

I'll look forward to seeing you at the show.

This is fine to not dwell on TIM and PIM.

I also agree with you on the healthcare debate, which is depressing and dismal. We may disagree on the details, but we probably agree that some kind of change is needed and that the current process for that change is frustrating.

One thing I'll throw in on the open loop bandwidth issue: I believe that those who design for wide open loop bandwidth often by necessity and nature pay more attention to open loop linearity than many do who use a lot of NFB. Of course, there are MANY exceptions to this, including myself and many others here who take open loop linearity very seriously.

To me, one of the core questions is the following: construct two amplifiers identically, with the same NFB gain crossover. Let them both have 30 dB NFB at 20 kHz. Let them both have very good open loop linearity. Let them both be highly stable with good gain and phase margin. By some quality means, limit the low-frequency gain of the second one so that the open loop bandwidth is 20 kHz in the second amplifier.

Do they sound different?

Which one sounds better?

Speculation on why they sound different if they do.

Cheers,
Bob

[Edmond Stuart]

Originally Posted by Bob Cordell

Hi John,
[snip]
To me, one of the core questions is the following: construct two amplifiers identically, with the same NFB gain crossover. Let them both have 30 dB NFB at 20 kHz. Let them both have very good open loop linearity. Let them both be highly stable with good gain and phase margin. By some quality means, limit the low-frequency gain of the second one so that the open loop bandwidth is 20 kHz in the second amplifier.

Do they sound different?

Perhaps

Which one sounds better?

Perhaps the first one.

Speculation on why they sound different if they do.
Cheers,
Bob

As the first one makes more sense. Limiting the OL gain at LF is plain B******T.

Cheers,
Edmond.