not Nagging you, "inoculating" others
I didn't search out a 38 yr old circuit to "throw in your face", quite the reverse, you brought it up
you cite it as a "success" and therefore supporting evidence for your following Otala's theories - including the "flat loop gain" prescription
you have made specific claims about the JC-3 inner feedback R and loop gain, the effect on distortion, IMD that are not objectively true
all "recently" - compared to the age of the circuit - in this thread
whether the claims of audibility, superiority, (only heard sighted) reflect any objective signal amplification circuit reality - they can't logically be supported by invoking simply wrong technical circuit operation claims
you have done it again and again – and I'll try to be there showing where the technical claims are wrong
then fanboys attack the technical argument with black bag rhetoric rather than extending, refining or showing up the flaws in the technical argument - and you are at best silent, never showing evidence in public of the grad EE courses at Berkeley you occaisonally bring out to beat people daring to ask reasonalble questions over the head with your "authority"
despite your admission past evidence suggests it is too late to expect you to abandon repeating the reasoning flaws, favorite stories – but just maybe some reading these forums will remember the technical arguments if they really want to learn circuit design, feedback, signal theory – at least they get some exposure to the real depth of “conventional engineering” vs the audiophile charactateur
I didn't search out a 38 yr old circuit to "throw in your face", quite the reverse, you brought it up
you cite it as a "success" and therefore supporting evidence for your following Otala's theories - including the "flat loop gain" prescription
you have made specific claims about the JC-3 inner feedback R and loop gain, the effect on distortion, IMD that are not objectively true
all "recently" - compared to the age of the circuit - in this thread
whether the claims of audibility, superiority, (only heard sighted) reflect any objective signal amplification circuit reality - they can't logically be supported by invoking simply wrong technical circuit operation claims
you have done it again and again – and I'll try to be there showing where the technical claims are wrong
then fanboys attack the technical argument with black bag rhetoric rather than extending, refining or showing up the flaws in the technical argument - and you are at best silent, never showing evidence in public of the grad EE courses at Berkeley you occaisonally bring out to beat people daring to ask reasonalble questions over the head with your "authority"
despite your admission past evidence suggests it is too late to expect you to abandon repeating the reasoning flaws, favorite stories – but just maybe some reading these forums will remember the technical arguments if they really want to learn circuit design, feedback, signal theory – at least they get some exposure to the real depth of “conventional engineering” vs the audiophile charactateur
Last edited:
Thx. I'll look over there. Note that I do invest in time and equipment so that I can learn and develope.... even in retirement. The audio portion of electronics is just a hobby to support my music habit.
Many would be surprised to know i dont listen over and over to anything I make for myself or publish and endlessly tweak it to "make what sounds good". I design it like all other engineers do - with theory, tests and measurements and then build it and then listen to music. I check the end result to make sure the distortion is low and the noise is low enough.
Take that little headphone amp I just published. Cost to copy it would be about $1.50 per channel in Asia. It only uses 8 cheap transistors which arent even great compliments. The THD is below -110-115DB it seems (into 30 Ohms). And, it sounds fine. I have not changed anything in it - just built it and listened to it as built.
I havent designed from scratch anything for my music listening in many years. Yet, its easy to get good numbers and even without a lot of gnfb or circuit tricks. No sim needed. And dc coupled without servo. I am surprised that all amps by now arent as good or even better and as cheap. Only IC's are as good or better - for a price.
Maybe we should be teaching how to do this sort of thing for DIY'ers? Thats why I came to this sight -- to show how easy and cheap it can be -- not High-End dollar stuff.
I am actually fine with anything as long as I have music to listen to all day long. Except MP3 lossy codecs.... those are just bad sounding so I use them when travelling or background. But I also own three iPOD models, too. Guess I flunk the Audiophile test.
BUT, there are some - however loose - correlations to look at.
-Thx, RNM
Last edited:
John,
Perhaps I have missed out. I try to use a 200 khz bandwidth and would love to see 160 db of range on the entire sample including the random (as can be) bits. (As life is finite I agree there are restrictions.)
Your insistence that noise and music can't be treated as periodic and continuous signals.
OK, then that's where we disagree, otherwise I may as well burn my dissertation and take an ax to the FTIR spectrometer in the next room.
Ed you dwell on the theory restricted to the EE101 literal unqualified definition (either the time or frequency domain is "infinite in extent") as if that means for practical problems "it is only a theory". Time and frequency are the two domains and like the Heisenberg uncertainty principle we can use windowing to control the ratio of the uncertainty in the two with the product a constant. You present only the two limiting cases. If you want -160dB you are up against some serious hardware problems, but if you simply want that numerically with double precision IEEE math (I think it has a 52 bit mantissa) just get out a good text book and do the numbers.
No. Frequency and time are conjugate variables. One representation is exactly equivalent to the other.
When I do theory I do use the unqualified definition. That helps me gain insight, just as does building and measuring real systems. (It is an interesting twist here that a practicing engineer in the audio field is being called too theoretical!)
What I would like to see in an analyzer, what is real and even more important what I am willing to pay for are quite different issues.
I do have a bit of a chuckle that virtually all the digital audio systems I see today are "24" bit. Even the low cost ones. To me there is a large difference between 24 bit arithmetic and a 24 bit accurate converter. (I don't want to get into the specific qualifications of true 24 bits, but 1/2 LSB accurate at a little more than double measurement bandwidth would be a start. The last time I complained about 24 bits someone pointed out a chip that was accurate to 24 bits... in a few seconds.)
The other chuckle is that more hairs are split by native English speakers than those for whom it is a second language. (Although there are certainly some language gaps also.)
Now for a current project I am doing analog filtering and complex multiplications, turns out filter tilt is an issue to get decent summing.
Peak level is only a time domain concept, but the information survives the transformation. Peak level for a real musical note is a complex problem, as you take longer and longer time records each note will occupy more and more and smaller bins around an "average" frequency. Which is to say someone plucking a G string every second or so will make sidebands on that note that will contain the time domain information transformed into the frequency domain. I have never tried to separate out a single note from a recording, but I suppose you could make a stab at it by taking all the bins around, say 440 and its harmonics, and doing an inverse transform. Wouldn't make much sense except with a single instrument.
Last edited:
- Status
- Not open for further replies.