Re: Towards -200dB distortion?
I am very aware of the fact that these distortion numbers are not achievable in the real world.
The reason why I do this, is because the distortions of real world amplifiers seem to be still noticeable.
If the classic classA output stage gives me -120 in sim and it is let's say only -60 in real,
then with this circuit I hope to get -80 in real, when I have -160 (K3+) in sim.
Bernhard
pjacobi said:
I am very aware of the fact that these distortion numbers are not achievable in the real world.
The reason why I do this, is because the distortions of real world amplifiers seem to be still noticeable.
If the classic classA output stage gives me -120 in sim and it is let's say only -60 in real,
then with this circuit I hope to get -80 in real, when I have -160 (K3+) in sim.
Bernhard
fscarpa58 said:
Federico,
yes, that was the first version when I was still happy the thing works...
It gave only current gain, so for 10V output some preamp 10V output would have been needed.
I have a new circuit that gives a voltage gain of 4.
I choosed that number arbitrary just to see if it would work with voltage gain also.
And it showed that there is no difference in performance.
With right values on left and right you can set the gain to any other number.
Please note that two resistors have been added on the left and four values changed on the right.
Some values on the right are very unrealistic
, this is only to get K2 below 200dB.13713 ohm would require a multiturn pot.
With normal values best result shall be -160 for K2+.
unitgain said:
Darn, there was in image in there! It showed up while the post was in moderation, but now I can't see it. I'll try to describe it in plain English
Imagine a bridged amp. The error of the first amp is added to the non-inverting input of the second amp, and the error of the second amp is added to the non-inverting input of the first amp. This way the two halves will tend to cancel each other's distortion.
The question is, if it's so imple, why no one thought of it before?
Again, I'm not a circuit designer
Parallel mode is possible too, but only for line level.
The main amp must have a 1k resistor on its output and after that resistor the error amp output is connected.
So the low imp. output of the error amp will be able to override the high imp. output of the main amp.
No way with power amp IMHO
Anyway even on line level I feel some problems which I can not specify, just a feeling...
But I just got a new idea
The main amp must have a 1k resistor on its output and after that resistor the error amp output is connected.
So the low imp. output of the error amp will be able to override the high imp. output of the main amp.
No way with power amp IMHO
Anyway even on line level I feel some problems which I can not specify, just a feeling...
But I just got a new idea
unitgain said:
Darn, there was in image in there! It showed up while the post was in moderation, but now I can't see it. I'll try to describe it in plain English
Imagine a bridged amp. The error of the first amp is added to the non-inverting input of the second amp, and the error of the second amp is added to the non-inverting input of the first amp. This way the two halves will tend to cancel each other's distortion.
The question is, if it's so imple, why no one thought of it before?
Again, I'm not a circuit designer
Functionally equal I did that in my extended version, somwhere posted in this thread.
Because I do not like adders I used subtractors, but in the end it's the same.
There are now two versions of my new idea and one is really elegant, it could beat the distortion killer circuit for some reason.
Even it can be added to the existing circuit.
If it will work, I already have a name for it - Nelson will be pleasured for sure
The Exorcist
Schematic will be posted when sim is ok.
Even it can be added to the existing circuit.
If it will work, I already have a name for it - Nelson will be pleasured for sure
The Exorcist
Schematic will be posted when sim is ok.
One possible version
For simplicity no bridged amp, and op amp only.
A differential amp subtracts main input from main output and provides with its output a floating ground for the feedback loop of the main amp.
In the sim this gives 12dB better performance. 128db >> 140dB.
No values fiddled yet...
Still I think the main amp feels the action of the error amp.
Another thing would be to connect the PSU of the main amp also to floating ground.
For simplicity no bridged amp, and op amp only.
A differential amp subtracts main input from main output and provides with its output a floating ground for the feedback loop of the main amp.
In the sim this gives 12dB better performance. 128db >> 140dB.
No values fiddled yet...
Still I think the main amp feels the action of the error amp.
Another thing would be to connect the PSU of the main amp also to floating ground.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.