Think of it within the context of figure 3. The loop area modulation and therefore the resistance, is a function of the absolute value of the current slew.
Proximity effect modulation of a cylindrical resistor's resistance is a function of the absolute value of the current derivative.
jn
Excuse me, these numbers are at G = <PLUS>1, the distortion making resistors must be imaginary. There is no current flowing in Rf in figure 4 save the (101x the error signal)/Rf. Come on guys, pay attention. Ed you mis-understood my joke, you said the thermal distortion was third order for a DRY resistor.
C'mon Scott, you know the drill. They are looking at the difference across a 10 ohm resistor with a gain of 100. The noise from just the resistor should be (back of envelope calc's) 100pV/Hz at the input. Add this to the 2.7 nV/Hz and you get 2.702 nV/Hz. (For those who have trouble with this addition please do a bit of homework before commenting.) With a gain of 100 that is 270.2 nV/Hz of noise. Now .00003% of 5 volts is 1.5 uV. So (1.5 uV/270.2 nV)^2 is a measurement bandwidth of 31 Hz.
Now they are measuring using an AP System 2, something I am familiar with. If I just go to the FFT analysis without stepping up the default resolution this level wouldn't even show up. Now at maximum rate the AP does go to -150ish re 1 V rms. But that takes 45 minutes or so. Without details on the settings of the AP one does not know if they were looking at noise or signal.
When they first did their setup they used higher value resistors. (So please don't give me they know what they are doing line... everyone makes missteaks.) I did tease BP about it and never got a reply, but the next time I looked they were using much lower value resistors.
ES
In Ed's test setup? No. I've delved in components and test setups using 400 V/nS with 8 amps per nanosecond, where the effects were right there in your face/slap you upside the head. But I've no access to an AP to try this at the low levels necessary in audio.
Let Ed do the test (if he wishes). It would be nice to confirm or refute a proximity based resistance modulation as applied to audio level signals and components.
jn
What math? This circuit works as shown just fine a 65k FFT has about 1Hz noise BW and the distortion components are easily resolved. Ed the 1.5uV is also X101 to the output not just the noise. 1.5uV of thirds is 150uV at the output (that's the point of the circuit) the noise in a 1Hz BW is ~.27uV. I think Bob Pease included the pictures when he did it.
Last edited:
Think one frequency in the Fourier domain the derivative has both signs, but it makes no physical sense to subtract AND add resistance by constriction of path. So I see seconds, sort of a rectifying mechanism.
Examine figure 3 again. Pretend the conductor they show is actually cylindrical in shape. Pretend it is skinning as they depict, such that current is flowing on the surface. Then pretend that there is a return current right next to it such that the current on the cylinder is closest to the return conductor. In other words, substitute a cylinder for their rectangular section.
Now, is the picture of positive going current or negative going current?
The answer is, it is a depiction of current crowding INDEPENDENT of the direction in which the current is changing.
That is the "clause" I refer to.
The resistance of the cylindrical resistor can be modulated higher than the base resistance by the proximity effect. It CANNOT be modulated to be lower than the base resistance.
The resistance will modulate in only one direction. Given sine wave excitation and cosine slew, the resistance will modulate based on the absolute value of the cosine slew.
jn
Bingo. That is why I mentioned ideal rectifiers needed in the simulation.
Scott...you have access to an AP?
jn
I try to understand this slang expression. Does-it has a connection with some sexual relationship with flies ?
Last edited:
Scott,
The AP System 2 as mentioned in the documentation doesn't do 65k FFTs!
The signal output voltage is 5 volts! That is where the data sheet shows the lowest distortion. As I suspect you know the falling distortion curve is really just showing the drop in noise level. (S+N)/N Above 5 ish volts the distortion begins to rise.
The 1.50 uV is the .00003% of the 5 volts out or what is suposed to be the measured distortion. If you want try the calc with the .000009% distortion number that is .45 uV! The claimed distortion is, if not at the noise level, very very close.
Now from the data presented in the manufacturer's data sheet and simple calcs, my suspicion is that we really are seeing noise distorted data.
My point is that modern opamps have surpassed many passive components!
Scott just join George and I in a bottle of Ouzo, you'll feel better.
The crowding depends only on frequency, not magnitude. Skin effect, and hence proximity effect, gives a geometric current pattern governed by frequency alone. Look at the maths in any decent EM textbook. That is why, as jcx(?) said, it can be modelled by adding some inductances to parallel paths.jneutron said:
Looking in the time domain, as you seem to be doing, will create complications when the phenomenon is best considered in the frequency domain. Of course the two must give the same results when done properly, but one will often be easier to deal with.
YAFD is an FLA - four letter acronym - which I just invented. It means Yet Another Fourier Debate. A debate which I don't intend to repeat, as my experience is that Fourier deniers cannot be convinced of their error.Esperado said:
Last edited:
Thanks for that comprehensive rundown on the comparative qualities of pots, John, that's the sort of info, gained from real experience, that's invaluable ...
Personally, I could not tolerate a system using a conventional pot for volume, except for a quick evaluation of capabilities ...
Frank
Thanks for the explanation, DF96. I hated acronyms, but just invented acronyms, you can't imagine
This said, you where not the first: Urban Dictionary: YAFD
IBIF.
Really ? Did-you buy your records on Mars ?
Sorry Ed you just obviously don't understand that circuit, you compute the resistors noise contribution (nil) and then say the resistors should be smaller.
Stop and think about it, the noise AT THE INPUT of the op-amp is 2.7nV/rt-Hz the distorton REFERED TO THE INPUT is 1.5uV they are BOTH amplified by 101, the distortion and noise of the souece see a gain of one. Trust me Ed you ARE wrong and I don't particularly like Ouzo or any pastis from any other culture.
Yes, I know the "dilemma" - every recording has already gone through a series of pots back in the studio; therefore why should one extra on the preamp make such a difference?
Part of the answer is that the pot's in the studio generally have to be reasonably decent, robust units; they wouldn't last long in the industry if they were too obviously flaky; another is that these attenuators are constantly being adjusted during the recording session, the contact points are refreshed frequently during this time; also, the sound of the distortion artifacts are now part of the recording's tone, ambience, the feel of the event, subjectively they belong to what happened in the studio, not what is going on in your room; and, a major chunk seems to be that these sorts of problems are worse when reproducing via speakers rather than recording, the fact that much larger pulses of current flow with PAs in the picture exacerbates things.
That said, many recent classical recordings are very dodgy in SQ, track to track the variation in level of artifacts that I usually associate with volume controls is quite striking -- perhaps because the engineers are getting too sloppy ...
The biggest reason, though, is that I know that my volume control makes a difference, I can hear it -- I can't do anything about what's on the recording, but, I can do something about the gear in my room. Therefore, I do what's necessary to eliminate that "annoyance" ...
Frank
Last edited:
- Status
- Not open for further replies.