No. The coriolis effect has nothing to do with some "memory" water should have
(which it does not) nor with other forces that move water in some direction.
Last edited:
Max, i admire your efforts to find a new evil and feed your anxieties. Just consider the amplifier with low IM that all self friendly music lover is supposed to use. With two signals of high frequencies and high levels, you read some 0.002% ?
What do you expect the IM level with 1/f noise ? One thing is for sure, there is a chance this noise you do not hear will be a little distorted by the signal.
Do some maths. Let us pick some figures out of the air: a decent amplifier might have 0.1% second-order distortion at full output (0.01% for SS, 1% for valves, so take the geometric mean?). Noise might be -80dB below full output. (If you don't like my figures, substitute your own). Second-order distortion scales with signal level. As the noise is 80dB below the signal, then the noise modulation will be 160dB below the signal. You would need very golden ears to hear that?Max Headroom said:
OK, not a fair comparison as we don't run at full output all the time. Drop the volume. Noise may stay roughly the same or fall a bit (depending on where in the signal chain it arises). Intermodulation will fall with signal level, so the noise IM will still be 160dB below signal.
Looking at the problem from the other end: what do you need to do to maximise noise IM from 1/f? First you need DC coupling to allow the noise to propagate through the circuit. Then you need active devices whose gain depends critically on their bias, such as BJTs (to them 1/f noise looks like a changing bias). If you can arrange it so their internal capacitance changes with bias too then even better: you can get phase IM too. Have I just demonstrated why valve audio sounds so good?
It's All Relative....
Dan.
Not all amplifiers are this 'blameless', especially when driving a reactive (loudspeaker) load.
Dan.
Very true. Look at the Wavac (at over $300k) for a perfect example. More conventional amps, however, do fall in the "blameless" category more often than not.
Deary, deary, me ... I do have to go through this painful exercise, do I?
,Okay. I used the term "memory" as a "fancy" way to indicate that the mass of waters ends up having a tendency to revolve in one direction or the other. From inertia mainly, thermal effects, even someone vibrating the floor under the wash basin at one point, you name it. It "remembers" everything that's happened to it, like a brick wall "remembers" how much sun has struck it by the temperature gradient through it, well beyond the time the sun has gone down ...
Getting closer ...?
Frank
Sampling Quality - Developer Zone - National Instruments
Pretty much any other 1/f noise white paper states the same.
Dan.
I wonder why audiophile amplifiers are obliged to measure so bad in order to sound sooo good, indeed.
depends I guess, certainly regulators, clocks and amplifiers generally have crappier PSRR down lower towards DC and below, but extreme HF/RF will get difficult too because stuff starts spontaneously leaving the PCB. but if we are talking audio frequencies, at least in my reading and application of Dacs, opamps, power regulators, the lower frequencies are the more difficult to keep noise and distortion low vs PS ripple.
by that I mean as a function of frequency. its not often linear though and youre right a lot of circuits and parts will have higher PSRR in lower, but not the lowest frequencies. I should stay away from general comments =) clocks and PSRR related jitter is always worse as the frequency goes down and that is more linear in that it gets steadily worse.
by that I mean as a function of frequency. its not often linear though and youre right a lot of circuits and parts will have higher PSRR in lower, but not the lowest frequencies. I should stay away from general comments =) clocks and PSRR related jitter is always worse as the frequency goes down and that is more linear in that it gets steadily worse.
Last edited:
Does the audio world not for a large part thrive on science-orientated, with some lack of creativity, and the non-factual driven, with way too much ?
Maybe audiophile amps do thing that audiophiles like very well . However they reproduce parts of the waveform that the audiophiles brain ignores badly . Thus "good sound" bad measurement. Take for example a signal that the brain process the first third and disregards the last 2/3 and and amp that does that 1/3 well then falls down on the 2/3 the brain does not process well . Resulting in a sounds great measures bad audiophile amp. There may also be an addon bit in that 1/3 that the audiophile finds pleasing which measure even worse . The use of physics rules to meta-physics does not work as well. So as we go down the road of the perfect apple tree all the audiophile wants is some of the fruit but not all.
Funny how a silly stupid statement by me can cause such a fuss. I only brought up the Coriolis effect to make a point. People seem to make statements about how in some instance in some place something is different that changes how they see something or interpret a result. As if physics has a boundary by your physical location on earth. I always told a joke years ago while working with Bayer chemicals about a development project they told me not to take because they could not do it successfully in their own laboratory. Well I took on that challenge. I thought otherwise. Well we were successful in doing what they said was not possible, we made the rocket part with material that they could not seem to get successful results with. When they asked me how I did it I told them very matter of factually that the gravity in California was different than it was it Pittsburgh and that is how me made it happen! That shut them right up, I never told them how I did it, it was our little secret. Just assign some result to a false premise and the uninformed have no response. Believe me the chemist in Pittsburgh knew I was pulling their leg, but they did not get their answer from me. The answer was mono filament fishing line hidden in the part. Nobody ever saw that and it remains a secret to them until today!
Mutual inductance. Sharing a common wire. Large loops. I'll leave out the secondary characteristics, you mentioned them.
Wrap or braid the three wires together from the xfmr to the bridge. Keep the center tap wire continuous past the bridge, then wrap or braid the rail wired with the center tap to the caps.
I did this kind of thing a decade or so ago but I used 5 mil copper foils, kapton sheet, scotch VHB transfer tape, and a 25 amp bridge with wire leads out of the epoxy instead of 1/4 inch air terminals.
The wiring characteristic impedance was about 2 to 3 ohms given the copper widths and dielectric thickness. The center conductor in the bridge was the center foil, the opposite conductors in the circuit were on either side of the center foil, a five layer construct.
jn
ps..course, my solution is more a high slew rate thingy.
Last edited:
does this happen with duel secondarys ?
Question does this also happen with multi secondary windings ? Will a duel secondary with a derived center point have the same effect and how much . Yes it not a direct answer to the question but is an option if the transformer is not already in hand.
- Status
- Not open for further replies.