Error feedforward - need help please.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I've been thinking about adding error feedforward to my Class A amp. As per the simplified schematic, what I might do is monitor the difference (i.e. the error) between gate drive signal and source output signal by means of T1A and send it in anti-phase to the other end of the speaker connection via T1B and a high current buffer. If the mosfet source lags behind by say 1 volt (aka distortion), this 1 volt error gets transferred by the transformer to the buffer and gets added to the total voltage driving the speaker. Repeating myself here. :cannotbe:

What I need to know though, is what would be a good little low impedance, high current buffer that would be suitable? I would like something that would run off say 5 volts and swing
+ / - 2.5 volts, preferably with complementary low voltage low rdson fets like are used in synchronous buck converters for example, running in class A. Standing current of several amps would be OK hence the hope of running off a high current 5 volt SMPS. The output current would need to be about 4 amps peak.
 

Attachments

  • feedforward class a.png
    feedforward class a.png
    2.7 KB · Views: 443
This is a tricky thing to do (I am working on something similar). I have just created my own simple class-A buffer..... yes it has distortion products as well, but the overall results is less. Making it stable and effective is not easy. I am looking at some form of micro control in the background.
 
Keep us all posted as to how it progresses, alvaius. Maybe there will be some idea in it I can plagiarise for my own use. :devilr:

Actually I made a bit of a mistake; it's not feedforward so much as Error Takeoff. There was an article in Wirleless World in the mid 70's about this, Sandman's the man. Does anyone have this article?
 
AX tech editor
Joined 2002
Paid Member
Circlotron said:
Keep us all posted as to how it progresses, alvaius. Maybe there will be some idea in it I can plagiarise for my own use. :devilr:

Actually I made a bit of a mistake; it's not feedforward so much as Error Takeoff. There was an article in Wirleless World in the mid 70's about this, Sandman's the man. Does anyone have this article?

I probably have the Sandman article, need to check this evening, will let you know.

Jan Didden
 
I think, that the implementation of that distortion-less high current buffer will be a non-trivial task. You will loose the main advantage of your initial follower – simplicity and harmonic content. How will you compensate distortion generated by the buffer itself?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.