I find it interesting that 12" of wire has a directionality, but you folks never talk about resistor directionality. Capacitors are a different matter since the outside foil is connected to one end.
We audio designers talk about a lot of things, we just don't discuss them here. Cryoing, laser welding, wire type and purity, for example. I just like to express MY 'experience, strength and hope' here sometimes for interested parties to understand what has worked for ME.
Of course, the feedback resistor was just the FIRST thing that we changed, then we went to REMOVING all the bypass caps and replacing them with a minimal set. Later we changed all the rectifiers to high speed soft recovery types, etc.
Of course, the feedback resistor was just the FIRST thing that we changed, then we went to REMOVING all the bypass caps and replacing them with a minimal set. Later we changed all the rectifiers to high speed soft recovery types, etc.
Cheap good common purpose metal-film resistors for audio frequencies. Stable with age, non-magnetic. Low distortion, low noise. Relatively small (compared to VS and especially BLP), so when used of the same size can dissipate more power.
The MLT resistors are not that bad, for definite groups they have +-50ppm (enclose a picture from the datasheet). But, again, resistive material with almost zero tempco could be of marginal quality for audio applications.
Attachments
Vlad, I think you are on to something important. Near ZERO tempco may be attained with compromises in resistor construction that compromises the audio quality in some way. Any ideas?
Yes, parts keep getting better. Just for fun here is the data on one of the not so good Vishays. Note it even has ninth order distortion!
This particular resistor is still used in some DIY audio projects. Not sure if any commercial offerings use it as I really haven't looked inside many.
SY,
Voltage coefficient was sort of the label for all the stuff for which the source of the problem was not clear. The method of measuring tempco was to determine the resistance at two different temperatures, it did not allow for non-linear tempco curves. The other issue is that there are thermo-electric effects in any termination. If the contacts on each end are not very close to identical then even uniform heating will cause a change. There are some smaller effects, but it seems that thermo and tempco predict almost all of the behavior.
Jan,
-140 db for resistors are actually not as good as some of the circuitry that is available! I think it is possible to get complete amplifiers with distortion below this!
There are two issues that are currently not well understood. One is the influence of noise and the other is the response to waveforms that are not accurately modeled by Fourier series.
ES
Attachments
Last edited:
Actually Scott it turns out resistors are directional! Every time I drop one it always travels downward! 🙂
ES
(Thanks for the straight line!)
I actually hadn't though of trying the directional tests on resistors, might have a chance next week though.
The major difference between MLT (C2-23, C2-33H) and C2-10 is in constructon of resistive layer. C2-10 in order to work on higher frequencies don't have spiral cuts, their resistance depends on thickness of metal layer. And they are more expensive. The cost that has no value for audio. C2-22 is the same device rated for twice lower power. But it is still the same, when I use routinely MLT-0,5 for 50-100 miliwatt of power instead of 500, and MLT-2 for 100-500 miliwatt.
Last edited:
When I listen to sounds it is kind of miracle, how without proper thorough analysis I can hear instantly things that are not seen on screens of analyzers. Can you analyze this sounds? Tey sing all around, and it is real stereo stage, with reflections, you can close eyes and hear where they sing, where are trees, where is creek... Amazing! 🙂
http://wavebourn.com/Frogs.wav
what is your point Wavebourn?
cameras aren't any good at scene interpretation either - yet optical distortion measurements, resolution, dynamic range, spectrophotometry, colorimetry are all used to to design better cameras/film/photo print systems
I suspect your clip would be one of the more difficult to AB/X under heavy perceptual compression despite having tossed >80% of the Shannon-Hartley Channel Capacity "Information" - the difference would be easily detected by waveform spectrum analysis despite not being obvious "by ear"
cameras aren't any good at scene interpretation either - yet optical distortion measurements, resolution, dynamic range, spectrophotometry, colorimetry are all used to to design better cameras/film/photo print systems
I suspect your clip would be one of the more difficult to AB/X under heavy perceptual compression despite having tossed >80% of the Shannon-Hartley Channel Capacity "Information" - the difference would be easily detected by waveform spectrum analysis despite not being obvious "by ear"
Last edited:
Well everyone, I must be either a 'fool' or an 'idiot' for trying to tell interested parties what I found wrong with the HCA-3500.
After all, it was MY design, at least it had my name and my circuit topology incorporated in it. It was based on my recommended changes to the HCA-2200 Mk2 to make a simpler and more powerful amplifier, and it measured well on my test bench and that of 'Stereophile', so it should be (if I am not a 'fool') perfect, at least in my opinion. Yet it failed with the reviewers and the marketplace. Why? Was anything changed that would 'turn off' the reviewers, such as removing an input IC? NO! Was it more ugly, or plain looking than the amps it replaced? No, not really. Did it do something wrong, like run out of current, have xover distortion, have low slew rate, or even really high odd order harmonic distortion? NO! So what made the HCA-3500 fail with my colleagues and the reviewers when they LISTENED to it?
OK, Jan, it's your turn. Tell me.
After all, it was MY design, at least it had my name and my circuit topology incorporated in it. It was based on my recommended changes to the HCA-2200 Mk2 to make a simpler and more powerful amplifier, and it measured well on my test bench and that of 'Stereophile', so it should be (if I am not a 'fool') perfect, at least in my opinion. Yet it failed with the reviewers and the marketplace. Why? Was anything changed that would 'turn off' the reviewers, such as removing an input IC? NO! Was it more ugly, or plain looking than the amps it replaced? No, not really. Did it do something wrong, like run out of current, have xover distortion, have low slew rate, or even really high odd order harmonic distortion? NO! So what made the HCA-3500 fail with my colleagues and the reviewers when they LISTENED to it?
OK, Jan, it's your turn. Tell me.
John, removing all the bypass caps and replacing ....all of them? or most of them? value changes?
what was the subjective and/or measured effect(s)?
what was the thinking behind going to a "minimal set", why that approach?
Thanks.
SY has warned of the kind of game and the capabilities of the performer a long time ago.
Nevertheless, a resistor placed in this position of that particular circuit should be made out of nothing else but "Zro" to it’s ninth stage.
Specifications of this material have not yet been released, but there is anecdotal evidence that it induces directionality effects.
Enjoy
George
AC!!!
NOW we see it all - in purest form! 😉
Did you know that A. Crowley was an international level Alpine climber before he ruined his mortal body?
NOW we see it all - in purest form! 😉
Did you know that A. Crowley was an international level Alpine climber before he ruined his mortal body?
People keep name-dropping Fourier, and hinting that the French gentleman cannot handle their music. Last time I looked, the requirement for a Fourier representation was that the waveform contains nothing worse than a finite number of finite discontinuities. Music is simple by comparison: no discontinuities at all, and band-limited.
Fourier theory seems to be a bit like schoolboy maths: those who are puzzled by it somehow think it is cool to advertise this fact in public.
Fourier theory seems to be a bit like schoolboy maths: those who are puzzled by it somehow think it is cool to advertise this fact in public.
It's an endpoint. It lumps together known and unclear stuff, icluding all the things you were speculating about. And it's the endpoint that matters, since that tells you what the real part will do in a circuit.
> So what made the HCA-3500 fail with my colleagues
> and the reviewers when they LISTENED to it?
Who LISTENED first ?
Your colleagues or the reviewers ?
> and the reviewers when they LISTENED to it?
Who LISTENED first ?
Your colleagues or the reviewers ?
- Status
- Not open for further replies.