heard some people don't use GNFB, particular guitar amps. I tried it on tube guitar amps, it sounded more "in your face" for the lack of words.
GNFB also flatten the frequency response ( flat gain) and extend the frequency response until the loop gain goes to zero.
To OP, what are you really try to ask? This is like asking "AWD" of cars!!! In what context?
GNFB also flatten the frequency response ( flat gain) and extend the frequency response until the loop gain goes to zero.
To OP, what are you really try to ask? This is like asking "AWD" of cars!!! In what context?
Last edited:
No. You need the right amount, which may be lots. It needs to be applied properly, not just slapped in (or removed from) an existing design.MiiB said:
'Imaging' probably relates to equality of frequency response between channels - feedback will help this. 'Engagement' may relate to favoured distortions - feedback may reduce this by reducing distortion.AKSA said:
If you can't measure it, then one possibility is that it doesn't exist. Or it may just be that you are not measuring it because you are measuring the wrong thing.
With gentile manner I understand thoughtfully applied. I cant help having the thought that the shorter the feedback path is, the less phase-shift due to device capacitance it has to work through.
Damir has made some good Non feedback voltage amplification circuits. They have low distortion, but need a light linear load to perform IRL.
I have made a gain of one OPS there I use a short cascode like feedback, The loop is really just around one/two devices and it holds excellent simulation results.
The one posted here does not really need the VBE and it can have multiple Fets in the output. It does PPM at 20KHz swinging close to the Rails. I have used jfet for the input as not to load the driving circuit.
Damir has made some good Non feedback voltage amplification circuits. They have low distortion, but need a light linear load to perform IRL.
I have made a gain of one OPS there I use a short cascode like feedback, The loop is really just around one/two devices and it holds excellent simulation results.
The one posted here does not really need the VBE and it can have multiple Fets in the output. It does PPM at 20KHz swinging close to the Rails. I have used jfet for the input as not to load the driving circuit.
Attachments
Last edited:
As a newbie I only try to understand electronics but why not give power to the user to control feed back; possibly from zero to whatever maximum is required depending on power required. Can a circuit be designed in such a way that one can have single ended low power to push pull to class AB ? Please be gentle electronics illiterate here. 
Regards.
Regards.
SY,
There is no such thing as "accurate amplification". Every sound that we hear as reproduced sound in our homes is just a mix of various distortions. Sometimes room acoustics distortions dominate and sometimes some other distortion mechanism is responsible for what we hear. Ultra low distortion amps does not help much if one considers the whole picture and sometimes such amps just reveal some other form of distortion making the sound even less pleasant. Therefore use of global negative feedback is not panacea. In many cases it just does not sound "right". The goal is not accurate amplification but listening pleasure and these two things are not in causal relation. The accurate amplification does not automatically mean listening pleasure. If accurate amplification is all that is needed I would be first to accept negative feedback as universal solution. But things are much more complicated than that.
Here's an article of AudioNote and time based distortions due to feedback: The Negative Effects of Feedback
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.