why do you apparently want to sound ignorant or deliberately obtuse?
which is exactly how you come off when making these assertions to skilled technologists with decades of electronic design, testing debugging experience
you know better - thermal distortions have time constants - just repeat your test signal at the appropriate time scale - it can still have a perfectly fine periodic representation for anything that is repeatable
even hysteresis is most often characterized by cycling the system through the range of internal stored states
My guess would be the opposite- he didn't do the test. He's not hiding some sort of horrible flaw in the preamp. It probably would test just fine.
Now that strikes me as curious, since if this test is so important and is a key to understanding the difference between a "good" box of gain and a "bad" box of gain, one would think that a nice comparison of (say) the Blowtorch versus a so-called "mid-fi" (John's term) box of gain would provide a demonstration of the superiority of John's approach. But since he's the only one here who has the DUT in his possession, I think we're going to continue to have the superiority asserted without evidence.
Some people want faith and mystery, others want data and transparency, and we are all free to choose which we are and whom we listen to.
My experience with thermal transient distortions is based on measurements of currents near to zero during short circuit tests of very high voltage circuit breakers (SF6 VHV and vacuum HV). These measurements vere done in time interval of -10us to +10us near current zero. The short circuit test current was up to 63kA and the measured current was in 2A/div. That means about 1/10000 of current amplitude. The aim was to measure phenomena of near-to zero current (after switch-off), arc voltage before switch off and ITRV (initial transient recovery voltage) after switch off. The squirrel cage shunt (or coaxial shunt) amplifier had to be designed especially as free of thermal transients. Please do not teach me on the subject. No periodic measurement was able to reveal amplifier thermal transient behavior properly.
Wow, once again. I feel that I have been gone over by my wife's divorce lawyer! '-)
Oh well, I can apologize for the ambiguity in my past statement. Sometimes 'jokes' don't carry over with different cultures, but here is the situation:
I did once use the Hirata box to test one of my 'studio boards', a professional series of discrete op amps that I first designed in 1977. These boards ultimately went into two 8 channel studio consoles, and three 30 ips master tape recorders.
An interesting (for this discussion) 'addition' that I made to the board was an AC BALANCE CONTROL. This adjustment was added to distortion trim each board, to remove any residual 2'nd harmonic distortion remaining after everything else was matched, due to a slight imbalance in the circuit topology, itself. I do not have a clean copy of this circuitry, but I will put up what I have shortly. In any case, ADDING the AC Balance control caused HIRATA DISTORTION. Removing it, eliminated Hirata Distortion from the measurement to below residual. That was the INSIGHT that I learned from the Hirata test.
Oh well, I can apologize for the ambiguity in my past statement. Sometimes 'jokes' don't carry over with different cultures, but here is the situation:
I did once use the Hirata box to test one of my 'studio boards', a professional series of discrete op amps that I first designed in 1977. These boards ultimately went into two 8 channel studio consoles, and three 30 ips master tape recorders.
An interesting (for this discussion) 'addition' that I made to the board was an AC BALANCE CONTROL. This adjustment was added to distortion trim each board, to remove any residual 2'nd harmonic distortion remaining after everything else was matched, due to a slight imbalance in the circuit topology, itself. I do not have a clean copy of this circuitry, but I will put up what I have shortly. In any case, ADDING the AC Balance control caused HIRATA DISTORTION. Removing it, eliminated Hirata Distortion from the measurement to below residual. That was the INSIGHT that I learned from the Hirata test.
To test me ?
I had made no comments about Ron's publication. I have read-it and thought: 'Nothing new under the sun'.
Once again, i had no attitude, because no comments from me. Just carbon copy of your words, then embarrassed explanations from your side, looking more like a pupil caught in the act than a teacher giving good example.
We are on Internet. If oral talks fly away, written words remain and everybody can make his own opinion.
You had pontificated a lot about Ron's test to explain-us, heretics, how good you are, and how bad are all the others. And that this test would show black on white how your <50 copies legendary preamp was sounding so superior to anything else in the known universe.
Last edited:
Ron Quan used to work with me when he was an undergrad in engineering at UCB. We have remained colleagues ever since. This is our connection. Ron and I do share in believing in 'golden ears', not necessarily our own, as we both knew an individual who was outstanding at hearing audio differences. He worked at a radio station, but could he hear differences! He was our 'inspiration' for future testing, to find a test that could approximate what the human ear hears.
Ron has recently retired from day to day engineering, so he has gone to testing for audio differences, namely PIM, or FM distortion, in his 'retirement'. Ron has met with an extraordinarily high amount for resistance to his testing and findings, so I tend to go out and support him. That is all there is.
I do wish that it was easier to measure these things.
Ron has recently retired from day to day engineering, so he has gone to testing for audio differences, namely PIM, or FM distortion, in his 'retirement'. Ron has met with an extraordinarily high amount for resistance to his testing and findings, so I tend to go out and support him. That is all there is.
I do wish that it was easier to measure these things.
That i had learned, reading this thread, resuming most of the John Curl's incomes, as the requisite for a 'high end' system:
- It has to be designed by the one and unique John Curl with secret selected and paired parts and secret (but known) schematics.
- It had to use special and secret cables, burned during weeks and oriented according to a secret procedure. Not to omit the use of special and secret power cords too.
- A decent chassis have to be machined from plain aluminum bloc.
- It has to be outrageously expensive.
- John Curl is working on a different (and much higher) level than all of us.
Near everything else is the story of his wonderful life, very useful for us, poor humans in charge to write his biography.
Did i missed something ?
- It has to be designed by the one and unique John Curl with secret selected and paired parts and secret (but known) schematics.
- It had to use special and secret cables, burned during weeks and oriented according to a secret procedure. Not to omit the use of special and secret power cords too.
- A decent chassis have to be machined from plain aluminum bloc.
- It has to be outrageously expensive.
- John Curl is working on a different (and much higher) level than all of us.
Near everything else is the story of his wonderful life, very useful for us, poor humans in charge to write his biography.
Did i missed something ?
Again, you miss my meaning, Esperado. My statements are not specifically directed at YOU, Esperado. You are among perhaps a dozen or more engineers who might be interested in this test. However, if nobody reads the papers written by Hirata, then there is little to be gotten from some test numbers, or even getting a further understanding of the test, itself. That is why I said 'Test on Friday' to get some here to read about the test, themselves, before we went into it further.
They in fact look a lot like squared off raised cosines. If you take say all frequencies from 5k to 10K and exclude all else you get sort these shaped waveforms. We did this with tweeters 32 yr. ago looking for spectral leakage. Linkwitz has similar waveforms on his website. This at some level is equivalent to multi-tone testing.
I don't see it unless you created some common mode effect. I have much better luck keeping all trims DC coupled and if necessary include a second trim to bring back DC balance. I have posted several circuits here.
I have also showed how with very slight DC feedback to the null pins you can give your beloved high OLBW amps a DC gain of 120dB and OLBW of 10Hz and they will show no difference in any of these tests.
Last edited:
Scott, I don't know PERSONALLY why I got Hirata distortion with adding the AC balance control parts.
This MIGHT be a USEFUL TEST DEMONSTRATION of the Hirata box, IF I ever get some POSITIVE interest in the distortion and what it measures. At this point, I don't feel inclined to put out the effort to do so. I ALREADY know the results, you are the ones 'in the dark' so to speak.
This MIGHT be a USEFUL TEST DEMONSTRATION of the Hirata box, IF I ever get some POSITIVE interest in the distortion and what it measures. At this point, I don't feel inclined to put out the effort to do so. I ALREADY know the results, you are the ones 'in the dark' so to speak.
Last edited:
Thanks for the input, PMA. Of course, you are correct. That is why the non-linear distortion generation of tantalum caps was originally missed. Virtually all the early audio designs had a SINGLE power supply, which put DC voltage on both the input and output tantalum caps. This is how we designed audio in the '60's.
Last edited:
- Status
- Not open for further replies.