Where did you find these high power resistors with -100dB 3rd harmonic distortion @1Khz and 50mV dissipation? Brand, series?
I'll take a couple, these may be valuable collector items.
Ed
If the feedback resistor was your worry, why you framed the chip amps?
It is an external component.
Don’t discrete amps have such a resistor too?
Wouldn’t it produce equal distortion there?
I’ll have a couple of drinks because in sober state I see these two posts not matching.
I wish to see them matching .
George
PS. Ironically, Scott was the first one here to target inappropriate low wattage feedback resistors
I hope so, fas42. Just i try to keep a scientific attitude. The first time , long time ago, i had heard of those rumors of the improvements given by some cones under power amps, i had done the following to figure out: Record the output of my amp, powered, both with no signal and with 1000Hz, and made big shocks on the amp. Of course, nothing ! I had not buy any of the golden plated cones. But i had take big cares to rigidify and dump my enclosure's walls the best i could and to isolate my CD transport from vibrations.
In the same spirit, i do not buy "Audiophile" cables. But i changed all the RCAs i could for mini XLRs, and added ferrite coil on all my cords.
Can-you justify by any measurement the advantage (on a sonic point of view) of your Blowtorch chassis, machined from plain unobatnium ?
Yes, you are coming from a similar angle: the fact you ditched the RCAs, added ferrites, tells me you can hear these distortion artifacts.
The trouble with the amplifier test you tried is that you expected to register an instantaneous difference. In general, it doesn't work this way, many of the issues take time to build up, the sound begins to degrade in a subtle manner over a period of time -- just like what happens if you rely on RCAs.
With the amplifier test, what I would have done is feed it an IM test signal, of high complexity if possible, and recorded the distortion spectrum over a period of time, every half hour or so. With, and without the footers. My assertion would be that you would be able to see changes in that spectrum occurring, if there was "something in it".
Frank
Chip amps have poor CMRR. The quality i.e. of the feedback resistor matters in discrete designs too, but less.
Do chip amps sound bad? no, some are quite good, mostly thanks to the very short signal, and to good behaviour at HF (unlike class d).
Or they have no idea how to measure. (Me neither but I'd like to learn)
Do chip amps sound bad? no, some are quite good, mostly thanks to the very short signal, and to good behaviour at HF (unlike class d).
Or they have no idea how to measure. (Me neither but I'd like to learn)
Yes, that's a way to look at it. The technique that I use that I've found very effective, is to consider every system I come across as being "perfect", but it doesn't sound that way because it's handicapped by weaknesses, often quite silly and simple in nature. So, the trick is to isolate, identify each and every weakness and rectify it, one by one by one. Doing that has always worked, got me sound far more satisfying than the vast majority out there deliver ...
Frank
Please actually produce this rather than make up these "adder" factors out of the air. Since there are plenty of PA's that measure in the -100's of dB's at full power under many different conditions, the burden of physical proof is on you, not some measurement of -120dB and then a shaggy dog story that makes it 60dB worse.
fas42,
I would sure like to know the mechanism at play that would give these slow changes over time that you are saying that you observe? So if you could actually measure this change in an amplifier caused by ultra-sonic or other vibrations over time, does that mean that this could be reversed by removing the vibration and the device would return to the previous state> Somehow I doubt very much that you have tested this hypothesis in any real manner, this sounds like a purely subjective result that you are stating as a fact. If what you are saying is true it would be both reversible and measurable over time. How about if you put the device suspended in a vacuum chamber and operated it that way, would the sound be different also over time from the normal state of having the device in the same room as the speakers? I truly believe that there is a difference in having spikes or feet under a speaker, the change in the surface area that the enclosure touches the floor and transfers energy into the structure. But I don't think that I have ever seen an actual measure of an amplifier changing output due to airborne vibrations affecting a component unless it was a component that could be vibrated and cause a measurable change, this would not take a long period of time to achieve the result. Let's hear your theory and proof behind these so called changes you have found over the years. RCA cables are notoriously poorly constructed and oxidation is the problem, but is there truly any other sonic difference if we are leaving out balanced signals between RCA and XLR terminated cables. If the impedance is not a problem what are you hearing, are you saying you hear the oxidation? Why wouldn't an XLR also be subject to oxidation over time?
I would sure like to know the mechanism at play that would give these slow changes over time that you are saying that you observe? So if you could actually measure this change in an amplifier caused by ultra-sonic or other vibrations over time, does that mean that this could be reversed by removing the vibration and the device would return to the previous state> Somehow I doubt very much that you have tested this hypothesis in any real manner, this sounds like a purely subjective result that you are stating as a fact. If what you are saying is true it would be both reversible and measurable over time. How about if you put the device suspended in a vacuum chamber and operated it that way, would the sound be different also over time from the normal state of having the device in the same room as the speakers? I truly believe that there is a difference in having spikes or feet under a speaker, the change in the surface area that the enclosure touches the floor and transfers energy into the structure. But I don't think that I have ever seen an actual measure of an amplifier changing output due to airborne vibrations affecting a component unless it was a component that could be vibrated and cause a measurable change, this would not take a long period of time to achieve the result. Let's hear your theory and proof behind these so called changes you have found over the years. RCA cables are notoriously poorly constructed and oxidation is the problem, but is there truly any other sonic difference if we are leaving out balanced signals between RCA and XLR terminated cables. If the impedance is not a problem what are you hearing, are you saying you hear the oxidation? Why wouldn't an XLR also be subject to oxidation over time?
Scott
There is a 3dB increase in thermal distortion for each halving of frequency so that is 9 db more for music than the test frequency. Then there is 6 db per doubling of power level that is another 12. Then there is the 6 db from the test method vs what you would get from using the same part in the high resistance side of a divider.
The sample shown was -100 when you add all the harmonics. So that would leave an Ohmite 1/4 watt resistor at -73. Now there are other makers who are worse and using a resistor 1/4 the value from the manufacturer's circuit you quoted would get you there.
Now the designer used the smallest wattage resistor in the most important place, so I suspect they were going for that. Of course if they just used one of the CC resistors marketed as "audio grade" that would get them there and then some.
If I find it I will post one of those plots. If you look at the folks who love CC resistors you will find some favorites have quite spectacular distortion plots. I am almost waiting for them to discover thermistors. Or has someone already used them in feedback loops?
But Scott you could have done the numbers yourself. It was never a secret that when CC resistors were dominant they couldn't be used in A/D converters much past 12 bits. N
http://www.tubecad.com/2007/12/23/The Sad Tale of a Half-Watt Resistor.pdf
Got anything new since 1957?
Got anything new since 1957?
Thanks for the thoughtful response, Kindhornman ...
Certainly no proof so far, I haven't the access to the right, sensitive enough gear to measure things, so my ears have to do the work in the meantime ...
In my experience the effect does not reverse if the mechanism causing the abberation is removed; perhaps if one waits long enough it will diminish in intensity, I haven't experimented sufficiently to say one way or the other. The simplest technique I've used to reset is to power down everything, wait and power up again.
Your experiment in the vaccuum chamber would work if the particular problem was due to external vibration; but if it was internally initiated, say from the power transformer vibrating at different levels, depending upon just about everything, then perhaps no.
My theory in this area, that has evolved over the years, is that seemingly subtle triboelectric mechanisms are behind a lot of these issues; it's not the answer to everything, but all experiments that precisely addressed this behaviour as a cause have given positive results.
Most certainly you can hear oxidation, this was one of my first "discoveries". Drove me crazy until I said, to hell with it, I'm going to hardwire everything(!) ... and that solved that one.
As you say, RCA connectors are terrible things; XLR are by design much better. The ideal is a gas tight connection, and the XLR has a much, much better chance of achieving that, if well constructed, the metal on metal pressure would typically be much greater.
Frank
Last edited:
I might just mention that my rough and ready system has the woofer/subwoofer literally 1 foot away, as the crow flies, from the CD player/system electronics, for reasons of getting reasonable integration with the stereo speakers, and laziness , and I run it at very high volumes. I wouldn't be game to do this on a normal setup, it would mangle the sound terribly, in the areas I worry about ...
Frank
, and I run it at very high volumes. I wouldn't be game to do this on a normal setup, it would mangle the sound terribly, in the areas I worry about ...Frank
- Status
- Not open for further replies.