John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
That said I still find it of interest, insulting the guy helps who exactly?

You must be new here , the insults havent yet started ..... :)

Scott , Sy and John have a special , err , relationship! Sy 's the enforcer of tides and greetings to all things presented by John for boutique application and try as they might , John's monologue will continue with no IC in sight , to Scott's chagrin ...

Seen .... :)

PS: Sy maybe steaming over the A21's, err long story ..... ;)
 
Last edited:
therefore, more 'intelligent' ways of measuring need to be devised ... ;)
If you try to measure the "subjective performance", you will surely fail. In your posts, you always explain what you feel with your tests, however.. you never quantify your "measurements". But it isn't the real problem...the trouble with your comments is that you are taking yourself as the "reference line" for all measurements, but you are unable to see that everything that you consider marvelous, can probably be taken as garbage by another one.
So... what's the truly "subjective reference line"???
The answer is: there isn't such thing as "truly reference line" when we talk about "subjective measurements". However, engineers need to design a loudspeaker or an amplifier or whatever, in order to allow all the people to listen to the music.
What can be done about it? It is quite simple: measure THD, measure the radiation beam, measure the reflection time, etc. You need to MEASURE everything that can be used as a quantification factor
So... "subjective measurements" are crap.. at least from an engineer point of view, unless he is trying to modify the acoustic behaviour from a room (they are crap anyway, but at least can be taken into account with a grain of salt).
 
So... what's the truly "subjective reference line"???
The answer is: there isn't such thing as "truly reference line" when we talk about "subjective measurements". However, engineers need to design a loudspeaker or an amplifier or whatever, in order to allow all the people to listen to the music.
What can be done about it? It is quite simple: measure THD, measure the radiation beam, measure the reflection time, etc. You need to MEASURE everything that can be used as a quantification factor
So... "subjective measurements" are crap.. at least from an engineer point of view, unless he is trying to modify the acoustic behaviour from a room (they are crap anyway, but at least can be taken into account with a grain of salt).
I appreciate your points, and agree it's very hard ... THD, or just straight distortion, for me at least, is the most important indicator, and that's what I focus on. Since this results from the sum total of all the elements in the system working the only truly valid way of verifying performance is to 'measure' what's emerging from the speakers, and compare that against what was recorded, the source waveform.

That of course has huge challenges, and currently no mechanism or software processing is going to have a chance of unraveling meaningful differences - items like Diffmaker are still far too primitive ...

But, I hope to work towards resolving that situation: acquire a far better recording setup, particularly one that handle high SPLs without overloading, so that the direct driver output can be captured with minimal reflected energy intruding; then calibrate the loop; and finally do some waveform processing and start to digest what the differences can tell me.

So far my admittedly very primitive recordings of what my gear does tells me that my ears aren't lying, the system produces the results that my brain thinks it is - therefore I'm motivated to go the next step ...
 
Depends on what you want to learn. This thread mostly deals with circuit design from 30-40 years ago and the application to fashion audio. Modern designers with the goals of performance and reliability would (and do) make use of technology that wasn't available then, but there's lot of other threads about that. Take your choice.

SY - Please prove that there have been any improvements in SQ in the products of what you call "Modern designers" when compared to the 30 - 40 years old designs.

I think you know that you cannot....
 
Now where were we? If we would want to make this particular example, we would have to look on the design sheet for the 2SK147 and pick some numbers off the graph.
First, to get 13 ma on the input stage, you would need approximately a 17ma Idss. You can do your own calculation of this.
To understand the effective gain, you have the equation 300/5+1/Gm
In this example, that is 300/5+18 or 300/23 or a gain of 13.
The noise ideally would be 2/3 (1/Gm) +5 or 17 ohms. It will be slightly worse than this because no jfet is perfect, and without 1/f noise.
Now, this is not an IDEAL circuit, because it is NOT push pull, and it will have LOTS of 2'nd harmonic distortion. Also, you have to hand pick the current source load to match what you want the second stage current to be.
A better solution is to make a COMPLEMENTARY gain stage just like the first one, but with complementary parts, remove the current source, and let the two complementary gain stages drive each other. Then you get a voltage gain of about 26 , 3dB lower noise, AND much less THD, especially if you convert the 5 ohm resistor to a 10 ohm, 10 turn wirewound pot that can be carefully adjusted. Now the noise will be equivalent to about 9 ohms which is about 0.4nV/rt Hz. (ideally) QED (This is not a perfect circuit, and please correct me if I made any mistakes) '-) The REAL circuit has a bit more to it.)
 

Attachments

  • 2SK147.jpg
    2SK147.jpg
    586.4 KB · Views: 365
  • ddd.jpg
    ddd.jpg
    200.2 KB · Views: 366
  • scp1 simplified.jpg
    scp1 simplified.jpg
    198.9 KB · Views: 368
SY - Please prove that there have been any improvements in SQ in the products of what you call "Modern designers" when compared to the 30 - 40 years old designs.

There haven't been. Boxes of gain could and were made to be sonically transparent decades ago- it's a solved problem. The issue these days in fashion audio is differentiation of parity product (from a pure ears-only POV). From a measurement POV, one can now make quieter and lower distortion electronics using fewer parts, which won't sound any different than other, older boxes of gain, but are simpler and less expensive to design and build.

Of course, many designers in fashion audio are still making effects boxes, and there are hobbyists who like to spend time rationalizing that, but that's a different story.
 
There haven't been. Boxes of gain could and were made to be sonically transparent decades ago- it's a solved problem. The issue these days in fashion audio is differentiation of parity product (from a pure ears-only POV). From a measurement POV, one can now make quieter and lower distortion electronics using fewer parts, which won't sound any different than other, older boxes of gain, but are simpler and less expensive to design and build.

Of course, many designers in fashion audio are still making effects boxes, and there are hobbyists who like to spend time rationalizing that, but that's a different story.

As expected: Absolutely no evidence, and even additional fogging..
Please prove the last statements and document the evidence.
 
not many academics will bother with spending limited time, grant money on null results where null results are expected - hard to get published too

there are occasional amateur blind, controlled tests reported on the web - can be dismissed as poorly designed, executed – just anecdotes – but they do exist and I would place them a little above completely uncontrolled, sighted anecdotes to the contrary

those who believe they hear amplifier differences are the ones in the position to show positive results and are really the only ones who can provide “proof” by providing scientifically acceptable blinded, controlled listening tests with statistics “rejecting the null hypothesis”

you wouldn't think “night and day” differences would be so elusive


the action professionally is in loudspeakers since they do make clearly audible differences even in blind listening – and at least the biggest companies have funded labs for extensive controlled subjective tests of speakers – the principle researchers when questioned really don't care much for amplifier “controversies” - they have bigger fish

Geddes and his wife have published distortion audibility research – I believe his current position is that it is worth testing for very low level “crossover” type problems but that smooth low order distortions that rise even to % at max output aren't readily audible
he has tested, found a <$200 Costco Pioneer home theater amp that measured just fine by his criteria – was used to demo his 5 figure $ loudspeaker system (for surround)

a question I have is do all tube amps pass Geddes test – many are push-pull, lots use cheap iron with measurable low level hysteresis
LOW LEVEL AUDIO SIGNAL TRANSFER THROUGH
TRANSFORMERS CONFLICTS WITH PERMEABILY
BEHAVIOR INSIDE THEIR CORES
MENNO VAN DER VEEN
 
Don't forget that $10k challenge guy who's run a few thousand trials. And the peer-reviewed work published in JAES by people like Lipshitz, Vanderkooy, and several others decades ago. Or the work cited by Clark. And, and, and. Like I said, this has been referenced here numerous times, and if someone with a track record of making wild claims is too lazy to look it up, I'm not going to do it for him.
 
There haven't been. Boxes of gain could and were made to be sonically transparent decades ago- it's a solved problem. The issue these days in fashion audio is differentiation of parity product (from a pure ears-only POV). From a measurement POV, one can now make quieter and lower distortion electronics using fewer parts, which won't sound any different than other, older boxes of gain, but are simpler and less expensive to design and build.

Of course, many designers in fashion audio are still making effects boxes, and there are hobbyists who like to spend time rationalizing that, but that's a different story.

Where have you hidden the evidence?
Hearsay are obviously valid evidence for you, but not for others.
So prove and present evidence for your claims as what you do now is claiming that a statement becomes a truth if the statement are repeated enough times even if it might not be true at all.
 
SY - Please prove that there have been any improvements in SQ in the products of what you call "Modern designers" when compared to the 30 - 40 years old designs.

I think you know that you cannot....

I would have to say yes there has, i do find newer amplfiers today sonically knock the socks of the older generation amplfiers..
 
Last edited:
Status
Not open for further replies.