# Feedback question

Discussion in 'Electronic Design' started by Andy, Mar 29, 2009.

1. ### AndyGuest

Hello,

Hope this is an appropriate newsgroup. I have been wondering what need there
would be for feedback at each stage of a multiple stage amplifier? I've seen
it done. Why not just take feedback from the output of the final stage and
send it back to the input? A little mathematics with the feedback equation
reveals no difference between the two approaches.

The only ideas I have at the moment are that it is to (a) reduce the chance
of any particular stage bursting into local oscillation or (b) match the
impedance of each stage to the prior and following stages.

I'm not sure this is right though so was wondering if anyone out there knew?

Andy

2. ### AndyGuest

Thanks,

I know about pole-zero compensation, I've used simple compensators to stop
switched mode power circuits becoming unstable, by adding phase boost at the
right frequency and increase the phase margin.

I guess another poster's mention of phase shift, even multiples of 360
degrees is a big potential problem, I wouldn't like to sort that much phase
shift out in a feedback compensation circuit, so I can see it makes sense to
apply feedback every stage or every two stages.

Feedback of course works on phase as well as non-linearity so I can see
local feedback will keep the overall phase shift under control in the
forward path. I guess I wasn't thinking about HF effects enough.

The non-linearity will be larger in the forward path of a global feedback
amplifier too, and I suppose that means larger harmonics will be generated
spuriously in the forward path if feedback is applied only right at the end
and being a higher frequency than the signal they could be more difficult to
handle, they may in fact exceed the bandwidth of the amp.

I suppose one of the problems I have is knowing whether this is rule of
thumb stuff or done by analysis. Applying feedback locallysufficiently to
keep the phase shift under control I reckon I can figure out if I tried.
Knowing how much local feedback to apply to stop non-linearities causing
problems sounds horrific to analyse though :0)

As for pole-splitting, I've heard of it but never had to get involved with
its consequences, it's in Art of Electronics, I'll read up on it.

If anyone has a good reading recommendation i.e. amp design for dummies I'd
be interested.

Good response to my question btw,

thanks everyone,

Andy

3. ### AndyGuest

Yes, you're right about being too simplistic.

My back-of-envelope feedback equation calculation did not use the right
inputs.
I neglected what would happen if phase shift exceeded 180 degrees at gain
Of course cascading several stages with significant phase shifts at the
cut-off frequency will add the phase shifts at that frequency, as feedback
won't be able to function with gain =1. Depending on the amount of feedback
used that could lead to oscillation if global feedback is used.

So simplistically :0) could we say for an n stage amplifier the unity gain
( cutoff ) phase shift in degrees per stage must be kept below 180/n ( or
better for adequate phase margin, 120/n ?). Depending on amount of feedback
again of course, that's for 100% feedback which is the worst case.

Non-linearities are an additional problem on top of phase shift I suppose.

4. ### amarkGuest

Often enough, DC feedback helps prevent thermal runaway. Nothing
prevents the output/input feedback from being the dominant loop in the
amplifier.

5. ### Allan HerrimanGuest

Try to find a paper with the title "Nested Differentiating Feedback
Loops" by Cherry.
IIRC some audio power amps were made using these techniques that had (for
the time) very good distortion performance.

BTW, the "differentiating" part of the title meant that the localised
feedback was AC coupled; DC feedback was only applied at the global level.

Regards,
Allan

6. ### EeyoreGuest

Well, I've adopted this technique myself, working on the basis that I want each
stage to be as linear as possible in its own right. It also extends the
bandwidth and lowers the phase shift of each stage as you correctly deduce
making overall stability easier to control

Now ask about pole-zero compensation. No, it's not dome for impedance matching
where I've done it.

Graham

7. ### Guest

It is pretty hard to compensate more than two gain stages in a loop.
Not impossible, since there are op amps with 3 gain stages, but often
less is more, keep it simple stupid, or any number of rules of thumb
comes to mind. You might want to read up on pole splitting.

8. ### Guest

I'm not sure it is a slam-dunk that you get better linearity by
applying feedback per stage rather than one big loop. I'd have to see
the math. As you know, you take a gain stage that only has 2nd
harmonic distortion and apply feedback, you get an infinite series of
harmonics, though very small due to feedback. Then you are feeding
that signal full of harmonics into another gain stage.

I am sure compensation of 3 stages in one loop isn't going to be easy.