> Let's stick with audio frequency effects please.
from guitar amp backplate :
- Reverb: Tube driven 4 counterwound springs long pan
- Vibrato: True pitch shifting varistor vibrato - single stage switchable to conventional tremolo
- Speaker(s): Celestion Gold
- Cabinet: plywood construction
How can you ' shift pitch ' with a varistor ?
from guitar amp backplate :
- Reverb: Tube driven 4 counterwound springs long pan
- Vibrato: True pitch shifting varistor vibrato - single stage switchable to conventional tremolo
- Speaker(s): Celestion Gold
- Cabinet: plywood construction
How can you ' shift pitch ' with a varistor ?
Actually a reasonable question.
On my 5 foot length of abused wire I get distortion of -155 dB. Your metal runs are at least another 45 dB shorter. Even at -155 you would still have 24 bits and room to spare. Let's see if the problem pops up when you get to 36 bits!
It would seem not at first glance but the answer is you actually could! All it would take is to apply the music signal to a voltage divider with a varistor in one leg. At the same time you would also apply a sine wave to get both the sum and difference results.
Now if you shift everything by say 40000 hertz then bandpass filter the result and then shift it again by 39900 you would end up with the original shifted by 100 hertz.
I suspect they only do it once with an oscillator frequency around 10 hertz.
I would rather be interested in 36 dB resolution re 1Vrms = 0dB 
. This is nothing but phantasy, and audibility of copper cracks as well. I would like to see any measurement that would confirm that 1 - 2 m of copper based coax or twisted pair had any measurable nonlinear distortion. I am sure you will give us no evidence of such thing.
. This is nothing but phantasy, and audibility of copper cracks as well. I would like to see any measurement that would confirm that 1 - 2 m of copper based coax or twisted pair had any measurable nonlinear distortion. I am sure you will give us no evidence of such thing.I would rather be interested in 36 dB resolution re 1Vrms = 0dB
Not only is it measurable it is even calculable! Coppers coefficient of resistance for temperature changes is 0.004041/C. What you are used to is copper that is thick enough that it can dissipate the heat. Given a thin enough copper wire and enough signal it is actually not hard.
However I think your question really is "Does this matter for audio?"
As the highest distortion I have measured for cables comes from the actual connectors and not the conductor that alone would give a hint. If you read my article on how I measured those distortions you would have been amused to see the problem was getting the EMI out. The EMI was at least three orders of magnitude greater.
Now I think it is possible to build audio gear that may be sensitive to defects in interconnects and easy to avoid those issues, that is not what you asked.
Now if you wish to make a wager that I can show thermal distortion in two meters of copper twisted pair, I will start with some 36 gauge and make up ten 2 meter twisted pairs and use those as resistors in my 10 resistor bridge setup. I suspect given enough voltage into that, it should be easy to measure the effect. Now is that typical of use, of course not. But my OPINION is that it is measurable.
Not surprising, since these results involved looking through a sea of spurs from line and who knows what tones.
As Samuel Groner points out at this level the confounders can pop up from just about anywhere. I don't feel you have done due diligence to identify all the potential errors at this level. The slightest common mode error is what is assumed to be a purely differential signal can destroy the result. None of the instruments you use are specified to have errors anywhere near this level.
I wish you mentioned it then. I can lend you my version if you want to play. I have show others the setup and let them play. The results are consistent even over a few years.
The preamp you used seems like it should work.
Last edited:
And you think that even if I am high on this measurement, is this going to cause anyone problems?
My conclusions have been that different solders don't make a difference and while it may be possible to design a system that is influenced by interconnects it is easy to avoid the issues. Is there any disagreement on that?
Last edited:
No, just that anything reported at that level is essentially random chance and you can make up any story you want.
Just a filter. Can model approximately with inductors and resistors. Nothing nonlinear.jneutron said:
Maxwell's equations are linear. Linear media (such as metallic conductors) are linear. Solutions are therefore linear, and superposition always works. There is no mechanism to generate non-linearity - for that you need non-linear media. This is second-year undergraduate physics (perhaps third-year EE - at least in the UK). I cannot comment on how this fits with "pay grades" in other countries.
nonlinear products of EM in linear media would be such a startling/interesting result that you would think grad level EM textbooks would use it as an advanced problem
or it would be subject of papers detailing it
again the recognized Ordinary Magnetoresistance Effect looks like it could have a very small effect in some conductor geometries
or it would be subject of papers detailing it
again the recognized Ordinary Magnetoresistance Effect looks like it could have a very small effect in some conductor geometries
Last edited:
So I can measure resistors with nice clean results but not a piece if wire.
You don't back up your measurements with a commesurate rigor that meets contemporary standards.
At least people are edging a bit closer to the elephant in the room - all these 'perfect' conductors, etc, end up in a connector, which is vaguely linked to another connector, in an extremely crude and imperfect interface - one can repeat the mantra that a chain is only as strong as the weakest link many times, but if there is no interest on focusing on that ...
- Status
- Not open for further replies.