?? the term distortion has no implied mechanism. Its like db, without the modifier of linear or non-linear, or .... its a value-less word.
From the sidelines it seems to me to be you who is insisting that distortion means non-linear distortion, while RNMarsh has made no assignment like that.
Alan
back to lurk mode
For seconds by chance you can simply do the math, out = in - .01*in*in and then feed this into the same out2 = out - .01*out*out and you get 6dB more seconds and much less 3rds and 4ths. Simple geometric progression for the first step it looks additive.
I know it isnt additive as in doubling and I know it 3-6dB..... I cant believe you guys. Are you really so dense? My context is the original context of the published DA test using pulse. I have not changed that.... it was about the many coupling caps (aluminum electros ) on the input and output of every stage and the gain/fb cap. In a chain of equipment and stages within each there WERE a lot of them in series to the signal.
Any other distortions of the intent and meaning is just that. A distortion.
THx-RNMarsh
Where did I "insist" that? I was explaining Dick's use of the word, since it confused Waly. Dick termed my explanation "absolutely correct."
Not too dense to know that the most pathological case is the only one of any interest as you are now seeming to agree to.
who uses electrolytic coupling caps with audible fc
put the corner down below 2 Hz already
then try to identify even polarized electrolytics by ear
again though who after Bateman wouldn't use at least 35 V rated bipolar Al electros with full thickness oxide on both foils for 2 Vrms consumer line level
put the corner down below 2 Hz already
then try to identify even polarized electrolytics by ear
again though who after Bateman wouldn't use at least 35 V rated bipolar Al electros with full thickness oxide on both foils for 2 Vrms consumer line level
Now? My DA article tested both polar and film which emphasized the poorest was the polar behavior. .
We know how to eliminate or reduce it.... W.Jung told us that a long time ago and that THD shows the affect of larger values (the voltage drop across the cap thing).
There is nothing new to learn as far as I am concerned ... oh maybe the film science behind any particular characteristic or where non-linearity comes from.... Thats cool.
But who really cares? I mean really cares. We all know what to do now and for a very long time. Use a film if you need a cap and use a servo when you dont.
BTW - there are plenty of people you can find who dont hear any improvement. So, it isnt like it is universally accepted by Joe Public.
Try reading a little of this: Capacitor Study For Audio Signal Path
It all balances out in the end. Now you know my adopted family is in Nepal and I am following that and trying to get in touch with family who will, like the others, be running out of food and clean water soon. So, I am putting this subject behind me and moving on to much more important things in this world than anyone's unresolved 30 year old issues.
THx-RNMarsh
We know how to eliminate or reduce it.... W.Jung told us that a long time ago and that THD shows the affect of larger values (the voltage drop across the cap thing).
There is nothing new to learn as far as I am concerned ... oh maybe the film science behind any particular characteristic or where non-linearity comes from.... Thats cool.
But who really cares? I mean really cares. We all know what to do now and for a very long time. Use a film if you need a cap and use a servo when you dont.
BTW - there are plenty of people you can find who dont hear any improvement. So, it isnt like it is universally accepted by Joe Public.
Try reading a little of this: Capacitor Study For Audio Signal Path
It all balances out in the end. Now you know my adopted family is in Nepal and I am following that and trying to get in touch with family who will, like the others, be running out of food and clean water soon. So, I am putting this subject behind me and moving on to much more important things in this world than anyone's unresolved 30 year old issues.
THx-RNMarsh
Last edited:
It's not really agony. I experienced it working on an analyzer.
That said, we should just write a song.
White noise...
...in a gilded cage.
Then we can all pose in front of a house with columns.
It's a beautiful day, today and most all days.
Cheers,
It was my understanding that a square wave acts as a sum of:
A. frequencies
B. harmonics
C. Both
D. None
Then what about a triangle wave?
What about using a square wave with and offset?
Where do we want to learn about measurement processes and their
affect on revealing distortion?
A. from Ancient Aliens.
B. from playoff level hockey.
C. from Mayweather v. Pacquiao.
D. from N. Korea.
E. from Baltimore.
F. from Last Man Standing.
If I left out an answergroup please feel free to add one.
Yes, it can compound much worse than just additive. Each time the signal is processed, more distortion is created,
which is further distorted by the next stage. This is why identical stages in series (even with gain reduction between
the stages to keep the signal levels constant) can sound much worse than a single stage, not just "worse proportional
to the number of stages".
Last edited:
The only approach that makes sense to me is that of subtraction. Subtract a source, a cause of audible distortion, determine the next one and do so again, and keep doing this until the sound is good enough to always please one. Remarkably 
, this is extremely effective - especially on low cost equipment ...
, this is extremely effective - especially on low cost equipment ... TAA 4/1980, Second paragraph.... "Dielectric absorption was investigated as its nature of behavior appeared to resemble the subjective descriptions being given to capacitors, especially electrolytic types. "
For the material science major I footnoted this gem: IEEE Transactions on Parts, Hybrids and Packaging. Vol. PHP-13, No.4, 12/1977 "Spurious Signal Generation in Plastic film Capacitors' by J.Borough, J.Burnham, W.Simmons, and S.Webster.
And, this footnote: TRW Inc. June 1980 App note; Circuit Design Considerations for Low Dielectric Absorption Applications' by W.Canning.
THx-RNMarsh
Last edited:
And distortion in cascaded stages can also cancel. This is simple to demonstrate for 2nd. For third it is a bit squirrelly.
I'm trying to recall if Borough et al. was the paper about which Samuel swore me to secrecy.
For some reason I'm reminded of this ... https://www.youtube.com/watch?v=TbKCISKooBc ...
Series transfer curves with (even small) higher order curvature are much less likely to cancel the distortion,
but rather will tend to make a more dense, complex array of distortion products. A couple of FETS or triodes can
indeed cancel distortion pretty well because the curvature is of low order.
Last edited:
Blanket statements about nonlinear systems are likely to be fuzzy at best.
What we need is a comprehensive theory linking nonlinear systems and distortion. Single frequency excitations are usually fairly simple and fall out of nonlinear operator theory. Multiple frequency excitations are more complicated. It is not trivial stuff.
Recently, the major horsepower devoted to the analysis of distortion via electronic circuit analysis has focused on high frequencies and communications. There may be authoritative materials on audio, but I haven't seen them. The two areas ought to map reasonably well.
Brad
Definitely fuzzy, in more than one sense. Yes, I'd like to see that for audio, but at this point in time it's unlikely.
Of course, audio is always multiple frequencies, or more properly, not even frequencies since they are transitory and not of long duration.
Much more straightforward to work on such things experimentally. And more fun.
Are you heckling me? Or applauding?
Either way, be careful, attention may go to my head.
- Status
- Not open for further replies.