Yes, in his latest article in Vol 9 he only tested non-polar, non-electrolytic caps. Distortion varied with bias voltage.
Another thought: conduction of AC trough a capacitor is only possible by moving charge from one plate to the other and vice versa. So in that sense, it's dielectric absorption that makes a working capacitor possible in the first place, no?
Jan
Something I might try. Two polyester caps in series, with DC bias through high value resistor to the mid point. Does it sound good?
Trailing bursts of noise falsely fill/intrude the silences.
In experimentation I have been able to profoundly subjectively change pink noise reproduction.....and no FR filtering involved.
BQP's do this also, but also add a 'signature'.
Some types of film caps and electrolytics add distortions....spectrally shaped noises according to the dielectric type may be part of the equation.
Dan.
Well, in that case, are you this Derek? The initials line up nicely. Google Afbeeldingen resultaat voor http://a.espncdn.com/combiner/i?img=/i/headshots/nba/players/full/246.png&w=350&h=254
From the example of the polar Elco, it seems to follow that indeed asymmetric DA can cause distortion. It makes it likely that when this asymmetry comes into play, there is distortion. As mentioned before, from the bare fact that there is relatively high DA, you cannot conclude that there will be distortion, see COG/NPO. So this is the evidence I see and the interpretation I give to it.
Would make a nice thesis to look at the absortion/release behaviour of other dielectrics and correlate it with cap distortion.
It is not so much a frequency-related nonlinearity, but rather a field strength related nonlinearity I am suggesting.
Are you still in joking mode? If there were a charge moving from one plate to the other, the cap would leaking and kaput. Except for electrolytics, where some leakage always occurs.
Air has no dielectric absorption, but thankfully, I had radio in my youth.
But I thought it is said to be a good linear model? Presumably it not only accords with measurements (I mean measurements, not speculation) but also physical models of dielectric polarisation relaxation - which I would expect to show exponential decay and hence can be modelled by essentially a low pass filter.Esperado said:
DA is modelled with a CR ladder because it has been found that this accurately models DA.jan.didden said:
More likely that DA provides a context for even-order (i.e. asymmetric) nonlinearity to be seen, by providing a temporary dielectric bias and hence creating asymmetry. If adding a bias voltage to a cap causes 2nd to appear where previously there was only 3rd does this means that the bias caused the nonlinearity or merely that it moved the working point to a different position on the dielectric nonlinearity curve?
Is this just another example of people looking at a time domain waveform after a signal has been passed through a filter, seeing a change and then frightening themselves and shouting "Distortion!"?scott wurcer said:
If a cap has a nonlinear dielectric then in most cases the response will be symmetric about zero electric field. Hence a symmetric cap with a symmetric dielectric can only produce symmetric distortion i.e. odd order. To get even order distortion you need to add some asymmetry e.g. add a bias voltage. If this is higher in value than typical AC signals (e.g. a coupling cap) then a little 3rd becomes rather more 2nd.SY said:
If the cap construction is already asymmetric (e.g. polar electrolytic) then a bias voltage could make things better or worse.
You have symmetry, so only odd order distortion can occur.PMA said:
And boy was it a visit!
An externally hosted image should be here but it was not working when we last tested it.
se
I've seen no reports or observations of this. If you saw my quadrature bridge experiment such an effect would have jumped out just like excess noise of resistors (which BTW vanishes the instant the stimulus is removed).
Not for the total behavior of all dielectrics, because it does not show distortions.
What's so special about air? Just one of an infinite series of possible dielectrics. And it's breathable, I give you that. ;-)
Jan
I think it has been established that this is not the case. DA and distortion are poorly correlated.Esperado said:
It only claims to model DA, so the fact that it does not model distortion is unsurprising. If DA had been found to have a nonlinear element in it then the CR ladder model could be extended to cover this too, but there is no evidence that this is needed.
Some dielectrics have DA. Some dielectrics are nonlinear. It is not surprising that some dielectrics have both features. What is surprising is that some people still seem to blame DA for nonlinearity when there is no evidence for this.
- Status
- Not open for further replies.