John Curl's Blowtorch preamplifier

Status
Not open for further replies.
Wavebourn, I have been at this for a long time. I have even learned from others, as wire and cables is not my forte. You apparently have difficulty in learning new things, and become upset when a potential error is pointed out to you about your own design.
Now, IF you have anything useful to say about wire differences, I would certainly like to hear them, but as my associates and I know, from direct experience, there are differences that are audible, even at line level. There even could be more important added factors, when we are working at MC phono cartridge levels. Once again, if you have anything constructive to contribute, then please do. If not, I would prefer that you not stir the fire. :geezer:
 
John;
I am still your friend, despite of your judgmental comments about my designs and my design philosophy. As I said you many times, I see we still have a lot to do with errors made by our lovely topologies containing active elements, before blaming on something outside of our creatures.
I see many flaws in the core of all of your designs you are proud of; however I said many times I see them nice, and it is the truth. But they are nice like some paintings that are nice without serving certain needs. You call me to learn, but you refuse to listen to anything different from your already well established beliefs.

Talk to you later, my dear friend!
 
Maybe the case is that Anatoliy is engaged in PA, rather than home hifi?

I can hardly believe that someone who has a decent system at home, with well placed speakers, that are able to yield good localization, has never heard the difference after signal cables change, regardless the reason is residual resistivity, EMI pickup or anything else. The diffference is quite important.
 
Well, PMA, that is how I find it, as well. I wish that it was not that way, regarding wires.
Actually, I am pretty much finished discussing wires, as there has been little collaborating input from others here, and lots of flack instead. I was thinking about common mode chokes next, if we can get up any interest.
 
PMA said:
I can hardly believe that someone who has a decent system at home, with well placed speakers, that are able to yield good localization, has never heard the difference after signal cables change, regardless the reason is residual resistivity, EMI pickup or anything else. The diffference is quite important.

I agree, I've experimented a bit with different cables and did not find two that sound the same. RFI alone can't be the reason for differences because I've tried cables of exactly the same topology but different copper purity / treatment and there were definite differences in SQ, mainly in detail, ambience and the focus of instruments in the soundstage.

I believe we listen to music differently and that is the reason why some hear clear differences and some not. I've compared a modified CDP against a standard version a while ago. With some people the reaction was "WOW" and a few didn't hear any difference at all.
 
One small, but important (to me) input I found by simply Googling' Van den Hul cable distortion'. Of course, anyone here could do it, but this is what I found.
It was an interview of VDH by 'Hi Fi News' in Dec. 1990.
VDH stated that he measured with a 500Hz single-cycle pulse at -85dBV, and found differences between wires. You can find the whole article if you look.
That is a very difficult level to measure with FFT spectrum analysis.
 
syn08 said:


Or not :-( I have a 40" plasma upstairs and I can certainly tell from the analyzer screen when people are watching TV in the family room. A 19" LCD computer monitor 6ft. away from the setup is also clearly visible. All between 60 and 70KHz and at around -120dB. Can reliable measure -130dB only at 3AM when everybody (including neighbours) are sleeping and lights are off, TV's are off, etc...

I was once about to think I got a RRR effect, but then there was only a CFL on in the cold cellar. That's how I missed a Nobel prize :rofl:

I wonder if it's an electric field broadcast, or if it's the fast transition speed. At least with tubers, the deflection coils tend to limit the slew rates.

The Nobel...oh well, c'est lavie..


john curl said:
I had to remove the small TV from my lab because of this problem, and I can also pick up the neighbors' TV's almost all the time. Since it is always at 15.75KHz, I can mostly ignore it. However, I like to remove it to make cleaner pictures.

Why not get your equipment to be less sensitive to outside interference? You would learn a lot if you could take T. Van Doren's EMC course.



metalman said:
picoamps ... hah, I was performing measurements down to femtoamp resolution!

Used to run a pair of Impact 4 probe setups, measured 4 inch wafers with IGfets onboard among other things.....when I datalogged Ig, it'd give me fempto and attoamp numbers. Lowest rez was 100 attoamps, but of course the numbers were nonsense. At least I understood the difference between resolution and accuracy..some here are not so lucky..

john curl said:
I really don't think that anyone has actually measured harmonic or IM distortion at 100nV or less levels. I have seen no proof.

So then why do you report VDH as finding 50 nV stuff? Are you being selective here? Did he show you proof.... or not??

john curl said:
Resistance measurement creates a DC bias that probably overwhelms anything that might have been there before.
Who knows..my vote is non-familiarity with the equipment and the test limitations. With your setup, I would suspect test limitations..you do not know how the shield resistance, the cable impedance, the connection contact resistance, nor the effective dielectric constant of the cables you test impact the results.

One could argue that the composite result is all that counts, however most people do not pop a cold one, plant their derrier, and listen to distortion analyzer waveforms...correlation to results below the noise floor of the equipment to actual music results is rather tenuous...

Cheers, John
 
jneutron said:

I wonder if it's an electric field broadcast, or if it's the fast transition speed. At least with tubers, the deflection coils tend to limit the slew rates.

Most likely the SMPSs. Almost each piece of consumer electronic (from laptop to TV) comes nowadays with a SMPS, built in or external. Inaudible EMI (to me) but certainly measurable.

I was always wondering (and tempted to ask the GEB) about the environmental conditions for their listening tests. Do they listen at 3AM when the neighbour's CFLs are off and the home electronics auto shuts down? Or perhaps they do the listening tests in the wild, away from any EMI? How can the GEB claim they hear unmeasurable differences between cables while they don't seem to care (in terms of providing some sort of reference/standard/specification) about such a simple, measurable effect? And, under these circumstances, what value does have for others a listening test report like "cable A is better than cable B"?
 
KBK said:
Now these are the sort of anomalous things that catch my attention. That make me sit up and pay attention. What was the conclusion?

For the tunnel stuff, they will not be cooling the machine down to 4.5 Kelvin (2.5 kilometers long) until Jaunary, results will probably start in march..they are trying to actively cancel the vibration of the multiton cold mass.

The closed cryostat problem was detected during a dark matter test here...an independent test to confirm (or refute) results of a test in Italy (I think). It was some laser beam/polarization/gradient magfield/dark matter whooseywhatzit experiment. Don't know what the results were, they've not published yet. Signals were supposedly down in the -140 or -180 db range...fraid I don't have the details..I just do the magnets..


KBK said:
Explain this exact problem, John. I may/may not have a fix, but I'm not sure I correctly understand what it is you are saying.

Any power resistor of any reasonable dimension, will produce an external magnetic field (even dale NI's or caddock ceramics). Any attempt to measure the voltage across that resistor requires connecting two wires to it. If you examine the loop formed by the wires, the resistor, and the measuring device (scope, dvm), you will find that the loop formed will trap some of the magnetic field created by the resistor.

If that magnetic field changes with time, that loop will generate a voltage. Magnetic field is abbreviated as B. The rate of change of that B is called B dot. The loop error voltage is proportion to negative B dot, as defined by Faraday's law of induction.

If one were to keep the same physical resistor size, but lower it's resistance, one would find the B dot error remains constant, but the IR drop will go down, this means the B dot error will become more significant for lower impedance systems. This is typically neglected in audio, but is is very important when one considers interchannel delay constraints in the 2 to 5 uSec regime.

My first load resistor to eliminate the B dot error was 40 watts, 4 ohms, was in theory 60 picohenries inductance (but measured at 250 pH due to physical constraints), and carried zero B dot error.

It is not a practical construct for the normal person. So I've designed some better, scaleable resistors that lend themselves to a more sane assembly process by sacrificing a tad of the low inductance design..and they can be fed by a matched impedance flatline, so that they will only suffer prop delay issues, which are no big deal..

Cheer, John
 
syn08 said:
I was always wondering (and tempted to ask the GEB) about the environmental conditions for their listening tests. Do they listen at 3AM when the neighbour's CFLs are off and the home electronics auto shuts down? Or perhaps they do the listening tests in the wild, away from any EMI? How can the GEB claim they hear unmeasurable differences between cables while they don't seem to care (in terms of providing some sort of reference/standard/specification) about such a simple, measurable effect? And, under these circumstances, what value does have for others a listening test report like "cable A is better than cable B"?

My take is this:

1. They hear a difference.
2. They then test using current equipment to find an electrical reason for the difference they hear.
3. They come up with null results within the reasonable range of measurements.
4. They squeak whatever differences they can find out of the measurement equipment, without regard to the resolution or accuracy of the equipment down at the level they work at. (John Curl very quickly found his analyzer incapable of the resolution or accuracy he required, so modified his equipment to better it...something the normal person cannot do.) Now, ascribes what he measures to what he hears..

I believe the answer to be found in accurate interchannel measurements....within the specific listening environment, with all the line cords, chassis, IC's in situ. Alas, what equipment can we use to correlate two channel, arbitrary time domain signals, to discern both interchannel level differences and interchannel timing differences, to determine how a human will perceive the apparent synthesized location of a sound spacially??? So far, all I can think of is ears..

Lab tests are useful to find specific geometric issues with wires, but are not useful for defining the system tolerance on the geometrics...things like external loop susceptibility and broadcast.

Cheers, John
 
B dot error of power resistors

I forgot to mention a very important point.

The B dot error of a low impedance resistor produces the exact same waveform pair that M. Hawksford provided in his '85 skin effect paper. The truncated sine stimulus drive, and the overshoot/RL decay to zero is exactly the waveform one gets when viewing the voltage across a low impedance resistor during a truncated sine.

I had this problem when I was viewing the current across a 250 micro-ohm resistor in series with a 3 inch diameter diode, and the waveform looked exactly like MH's. Since it couldn't be TRR (three orders of magnitude too slow), I looked at the CVR.

When I dropped the scope wire down the geometric center axis of the CVR, the error went to zero. Much more elegant solution than the other possibility...calculation of error voltage from slew measurements...messy..

MH used a shorted steel wire pair, so the impedance was very low..hence the B dot error he reported as "residual".

Cheers, John
 
I do not know who do you speak about, but I care about EMI a lot and try to design and interconnect the way resulting in minimum measured EMI disturbing signals. Have comparisons and correlations of "sound quality" with EMI level.

In my experience, the cables that seem to sound very "airy" are especially suspicious. Almost allways a lot of "air" means a lot of EMI pickup (no shield or only 1 end connected), and also the cables, "airy" with simple music, fail with complex orchestral music, sounding harsh.

Regarding SMPS, try to replace with good linear PSU with low primary-secondary capacitance and then tell me, if there was no difference.
 
PMA said:
I do not know who do you speak about

Hey, who said I do??:D

PMA said:
.... but I care about EMI a lot and try to design and interconnect the way resulting in minimum measured EMI disturbing signals. Have comparisons and correlations of "sound quality" with EMI level.

In my experience, the cables that seem to sound very "airy" are especially suspicious. Almost allways a lot of "air" means a lot of EMI pickup (no shield or only 1 end connected), and also the cables, "airy" with simple music, fail with complex orchestral music, sounding harsh.

The best transmitter is the best receiver.

ElectroMagnetic Compatibility (EMC) is all about control of currents.

NEC is all about putting currents wherever they can do the most harm to our music.:D :bawling:

The RCA phono jack is an instrument of the devil.

Correlation is not causation. It may very well be that EMI correlates to what one perceives..don't know.

Cheers, John
 
PMA said:


Regarding SMPS, try to replace with good linear PSU with low primary-secondary capacitance and then tell me, if there was no difference.

Yeah, starting right now:

- 3 x Laptop power supply
- 5 x desktop PC
- 5 x PC monitor (2 19" LCD, 3 19" tube)
- 7 x TV 19" to 40"
- 7 x Satellite set top boxes
- 8 x AAA, AA, cell phone, Blackberry chargers

I'll let you know when I'm done. I'm also thinking about breaking in the neighbours houses to do the same.
 
syn08, without any jokes, I have no problems with measured EMI in my audio chain until I connect cheap, 200,- USD universal SACD/DVD/CD player with SMPS directly into my system. The direct interconnection and cable shield capacitively coupled currents are the problem. Not PCs and other stuff connected to mains.

I seldom speak about theoretical issues only.

And sorry, I do not connect my audio system to TV, settopboxes and similar stuff. Yes, when I look at DVD, but then I do not speak about high-end sound. If you mean a commercial sh_t like this, then yes, nothing makes difference, but then I do not understand why to be engaged in high-end oriented thread, when no high-end is listened to, just sh_t.
 
jneutron said:

1. They hear a difference.
2. They then test using current equipment to find an electrical reason for the difference they hear.
3. They come up with null results within the reasonable range of measurements.
4. They squeak whatever differences they can find out of the measurement equipment, without regard to the resolution or accuracy of the equipment down at the level they work at.

Or

1. They need to hear something, to make a difference (for their own overinflated ego, justify expenses when wife finds out about the cost, business/marketing reasons, score tonight, etc...).

2. End of story. Anybody else is stone deaf.

Obviously, the burden to prove such, beyond any reasonable doubt, is on them. They usually do diddly squat about.
 
Status
Not open for further replies.