But Bear, do you suggest that damping factor is somehow determined by the type of output device rather than the total amp design?
I can design high DF with ANY device, including LATFETs. Or design with DF = 1, with any device.
You can have an amp with weak bass / low DF and find out that it uses LATFETs. But concluding that LATFETs therefor cause low DF is very, very sloppy thinking. Actually total absense of thinking.
Jan
Due to lack of creativity I don´t make these things up myself
https://books.google.nl/books?id=YT...=noise pushing signal above threshold&f=false
Well if you put the ouput devices outside the NFB loop that is your lookout. But Lateral amps with a few milliohm output impedance are 10 a penny.
Any thinking that I have is clearly muddled.
I rely upon those amongst this well read and traveled group to guide!
And, I drew no conclusions of my own.
That would be imprudent on my part.
I merely took a note...
...now torn up and crossed off!
I'm just trying to get at where these "turning points" and "thresholds" lie.
I've just been told that amps with <0.01%THD are both indistinguishable from each other AND that their distortion is swamped by the distortion produced by speakers. I think this is what has been said??
Or was that 0.1%.
Hard to recall now.
Yet at present we are discussing the audibility of a sine wave buried in ~-120dB noise. This is certainly of interest to me, but I'm trying to think about how to assemble my next system, and what the criteria need to be.
So many seemingly contradictory ideas and thoughts...
Actually, yes.
Yes for HF, as in >1500Hz to above 20kHz.
More to come, once someone steps up with some roughly calculated numbers.
First linear, then a rough approximation of the drop in THD as power is reduced... which is in and of itself an interesting issue. Where does the lower limit of THD for a given driver lie, does the "graph" start to look like a remote cutoff pentode?
_-_-bear
Yes, Bruno´s latest speaker is fully active and quite good actually; I heard them fill a pretty large room with solid bass and uncolored sound. How well they measure? According to a German magazine spectacularly well. More to come.
No research?
No understanding?
or no interest in sharing your knowledge?
So why post unless you want to add in a positive way?
Maybe as a moderator you can tell how many members are reading along
and not posting??
Well, regardless, MOST of them have FAR LESS expertise than you do, and also far less than many of the main contributors here. So, to do justice to the forum and its aim of DIY audio, providing information and explanation is appropriate. As is relating detailed higher-level physics and engineering back to something relating to the typical reader.
This isn't just a private conversation/club for credentialed experts, in which they can be pedantic, and be condescending to "the masses". Don't you agree?
Looks like an interesting dissertation.
Yes, a thorough analysis of stochastic resonance is not a simple hand-waving exercise. The optimal amount of noise and its spectrum are crucial, and you can easily go over the top and degrade things. What is also to be noted is that the nonlinearity in the system does not need to be a step function for the approach to work.
Levinson (the company not the man) was getting interested in SR for a little while but it was mostly conjecture. This was around the time, circa 1995, that I was working on the switcher for Harman R&D, and they were vetting the design.
Bear, it is an interesting question, which can only be resolved by measurements in an anechoic room (because of the low levels you want to measure in order to get as much data points as far apart on the curve as possible). At the end of next month I might have the anechoic room at the Technical University of Delft for a day together with Jan D., and this is a fairly quick measurement to do; ARTA has a subroutine for it. I got interested. All I can tell you now is that all distortion from all drivers I ever measured went down with level. I just don't know what the correlation is, nor if it is the same for all drivers. What I did notice is that many drivers exhibit something of a discontinuity, where a small addition in output starts to generate disproportionately more distortion. This point can be easily found by measurements in relatively noisy circumstances.
bcarso,
Even the stiction would seem to be a very nonlinear function as once you get a mechanical system moving even a small amount the stiction would change functionally. In a speaker it would be the mostly the rigidity of the base resin in the spider I would imagine over the surround, but to put a percentage to that I wouldn't dare say. Add to that the non-linearity of the typical overhung voicecoil and flux density distribution and you have a combination of nonlinearities to contend with. I can compound on those with other nonlinearities but I know you do understand all the interactions that are taking place at both the lowest levels and again at the limits of output of any mechanical audio device. If we did indeed have a very linear speaker system so many of the interdependencies between the mechanical and electrical systems could actually be predicted that are now so impossible to predict by calculation alone.
Even the stiction would seem to be a very nonlinear function as once you get a mechanical system moving even a small amount the stiction would change functionally. In a speaker it would be the mostly the rigidity of the base resin in the spider I would imagine over the surround, but to put a percentage to that I wouldn't dare say. Add to that the non-linearity of the typical overhung voicecoil and flux density distribution and you have a combination of nonlinearities to contend with. I can compound on those with other nonlinearities but I know you do understand all the interactions that are taking place at both the lowest levels and again at the limits of output of any mechanical audio device. If we did indeed have a very linear speaker system so many of the interdependencies between the mechanical and electrical systems could actually be predicted that are now so impossible to predict by calculation alone.
How does that apply to speakers ?.
Sure, in mechanical contact systems expect degrees of stiction, but a speaker driver is a 'floating' system, ie force, mass and compliance but no mechanical contact in order to cause 'stiction' ?.
Dan.
- Status
- Not open for further replies.