John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
I would like to see the math for the non-harmonically related frequencies or how three orders of magnitude more PM in a speaker than an average amplifier does not matter.

Unfortunately Scott you can't see the math of perception. But according to many researches (I'm not name-dropping) :) we perceive and compare by similarity. If distortions mimic what we expect to hear in the nature we ignore (filter out) them; we do that all the time recognizing sounds passed through the air, through walls, reflected by different non-linear surfaces and so on, and the more they differ the more of processing is needed to reconstruct and recognize sounds. It leads to a listening fatigue that can't be measured precisely, but can be used as an indicator.

Vintage speaker are more venerable, they cost more than modern ones. Why? Because on the run for lower THD numbers designers made speakers that produce more alien to the nature distortions, but anyway they are mostly mechanical devices, so their higher THD and PM levels add less of unnatural errors than amps do.
 
AX tech editor
Joined 2002
Paid Member
They convinced me, already. But then I am more willing to learn new things, and accept that some people actually know more than me about it. I do have a call into Mitch Cotter. He is tied up on his project, but I did leave a message asking for further details, just 1/2 hour ago. It may be until tomorrow, before I hear from him. I have tried over the last 18 hours, already.
Personally, I don't need to convince anyone here about anything. It is your gain or loss to follow up. It is now obvious to me, but it is because I participated in presenting the original graph, and did dozens of separate measurements, myself, in Matti's lab, getting the same graphs. I KNOW the test procedure, and its correctness.
What Mitch Cotter contributed, was NOTING the FM distortion as it would appear on the same graph. He showed Matti, Matti verified it, himself, (didn't tell me), thanked Mitch, then wrote his first paper on PIM, ALL math, no measurements. I just met up with Mitch a few months ago, and we talk almost every day. Does this help?

Not really. I'm stil being asked to believe your and Mitch's blue eyes. If you understand it so well, and are so strongly convinced that it is correct, why not explain what's going on?

jd
 
AX tech editor
Joined 2002
Paid Member
Unfortunately Scott you can't see the math of perception. But according to many researches (I'm not name-dropping) :) we perceive and compare by similarity. If distortions mimic what we expect to hear in the nature we ignore (filter out) them; we do that all the time recognizing sounds passed through the air, through walls, reflected by different non-linear surfaces and so on, and the more they differ the more of processing is needed to reconstruct and recognize sounds. It leads to a listening fatigue that can't be measured precisely, but can be used as an indicator.

Vintage speaker are more venerable, they cost more than modern ones. Why? Because on the run for lower THD numbers designers made speakers that produce more alien to the nature distortions, but anyway they are mostly mechanical devices, so their higher THD and PM levels add less of unnatural errors than amps do.

Scott didn't ask for any math on perception. He said: "I would like to see the math for the non-harmonically related frequencies or ...". He asked for the math explaining how non-harmonically related freq lines can show up in the spectrum. Or so I understood.

jd
 
Last edited:
John,

You asserted that Barrie Gilbert's paper proved your point on PIM. That is not so, even though Barrie did not make any mistakes. I frankly think that you are very much in the habit of putting your own spin on what others say.

Cheers,
Bob

I just re-read, in part, Barrie's paper. The distortions are all related to the input frequency(s), the math BTW is not very complex. He shows a Taylor expansion of the first and third order on an input stage generates thirds, nothing remarkable.
 
You guys amaze me. However, I am right on this, just wait and see. However, I don't have the time or energy to redo everything and spoon feed those who don't want to change their paradigm.
My job in life is making better audio products. I already know how to do so, without 'proving' it to anyone. It is just NOW that we have separate PIM measurements for IC's by one associate of mine, and a more complete understanding of the magnitude and position of PIM compared to TIM, by Mitch Cotter.
I am giving the ANSWER to successful audio design from MY perspective and understanding. Many here may have a separate opinion, but you better be right, if you want to compete with the 'Big Boys' like Charles, Nelson, or me, all 'Engineering Technicians' by Anatech's personal definition. We may choose to disagree on this,as well, but so what?
 
Last edited:

Is it good measurement, or bad one?

(1:10 probe, 3K load resistor)

Tower-IV-TA-3.gif
 
Unfortunately Scott you can't see the math of perception. But according to many researches (I'm not name-dropping) :) we perceive and compare by similarity. If distortions mimic what we expect to hear in the nature we ignore (filter out) them; we do that all the time recognizing sounds passed through the air, through walls, reflected by different non-linear surfaces and so on, and the more they differ the more of processing is needed to reconstruct and recognize sounds. It leads to a listening fatigue that can't be measured precisely, but can be used as an indicator.

Vintage speaker are more venerable, they cost more than modern ones. Why? Because on the run for lower THD numbers designers made speakers that produce more alien to the nature distortions, but anyway they are mostly mechanical devices, so their higher THD and PM levels add less of unnatural errors than amps do.

I largely agree with what you have said here.

First, I am very reluctant to compare distortions created by speakers with distortion created by electronics. That a loudspeaker may create several percent distortion of whatever kind should not make us complacent about a solid state amplifier that creates 0.1% distortion. I'm comfortable thinking of them as two completely different things. Similarly, I'm at peace with the situation where a vacuum tube amplifier with 0.5% measured distortion sounds better than a solid state amplifier with 0.1% distortion. This doesn't mean we throw the baby out with the bathwater and completely dismiss distortion measurements, especially ones that are sensitive to crossover distortion and 7th harmonic level, for example.

Second, I totally agree that "unnatural" distortions may be much more audible than the more normal distortions. John asserted that PIM produces much more relatively audible "unnatural" (non N/M) distortions than other distortion mechanisms. If PIM really does produce such types of "unnatural" distortions, then we have to look much more closely at PIM. Perhaps we then need to set a criteria for such distortions as needing to be, perhaps, 10 or 20 dB lower than would otherwise concern us. This issue is not where I disagree with John on PIM.

My concern is twofold. First, it is very unclear how PIM can create such in-harmonic distortions under any reasonable conditions at any reasonable levels.

Secondly, if a lab experiment exposes such in-harmonic distortions, how does one conclude that they are a result of the PIM mechanism as opposed to something else?

Cheers,
Bob
 
Status
Not open for further replies.