Maker Pro
Maker Pro

SMPS feedback compensation with error amp simpler than type 1?

eem2am

Aug 3, 2009
429
Joined
Aug 3, 2009
Messages
429
Hi,
You have type 1, 2 & 3 error amps for SMPS feedback compensation...
But what is the error amp called that is just a high gain inverting amp?...ie, not an integrating error amp.

Also, can you confirm, that using this (unknown name) error amplifer, there is no gaurantee in eg a Buck battery charger that you would reduce the chances of instability, by using it, instead of a type 1, 2 or 3 error amp?
 

cjdelphi

Oct 26, 2011
1,166
Joined
Oct 26, 2011
Messages
1,166
on a low voltage buck converter the output will have a resistor divider across it, the voltage might be so low it needs to be amplified via an opamp first, no clue why they call them error amplifiers though, and if a voltage threshold is high enough a signal is sent back to the regulator to hold off on the PWM signal until the comparator of the feedback voltage is low enough to begin switching again

and visa versa
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
no clue why they call them error amplifiers though,
It is called error amplifier because it amplifies the error, i.e. the deviation of the output voltage from the expected (set) output voltage.
the voltage might be so low it needs to be amplified via an opamp first,
Typically an error amplifier doesn't amplify the divided voltage (if that were the case, one could simply design the voltage divider differently). The error amplifier amplifies the difference between the divided output voltage and a reference voltage. This amplified signal is the used to control the regulator's output such that this difference, the error, becomes minimal.
As the error amplifier sits within a feedback loop, stability is critical. It needs to be fast for good regulation. It also requires a good phase margin to avoid oscillation.
and if a voltage threshold is high enough a signal is sent back to the regulator to hold off on the PWM signal until the comparator of the feedback voltage is low enough to begin switching again
Usually the regulation is done by modifying the duty cycle of the pwm signal, not by turning the pwm completely on or off. This can be done in case of no-load or very low load conditions, but increases output ipple.

@cjdelphi : You may need to brush up on your switch mode regulator fundamentals.
 

cjdelphi

Oct 26, 2011
1,166
Joined
Oct 26, 2011
Messages
1,166
It is called error amplifier because it amplifies the error, i.e. the deviation of the output voltage from the expected (set) output voltage.

Typically an error amplifier doesn't amplify the divided voltage (if that were the case, one could simply design the voltage divider differently). The error amplifier amplifies the difference between the divided output voltage and a reference voltage. This amplified signal is the used to control the regulator's output such that this difference, the error, becomes minimal.
As the error amplifier sits within a feedback loop, stability is critical. It needs to be fast for good regulation. It also requires a good phase margin to avoid oscillation.

Usually the regulation is done by modifying the duty cycle of the pwm signal, not by turning the pwm completely on or off. This can be done in case of no-load or very low load conditions, but increases output ipple.

@cjdelphi : You may need to brush up on your switch mode regulator fundamentals.

of course it has to be completely on or off in a duty cycle or it wouldn't be a duty cycle, I'm talking 70k times a second with a comparator, and again depends on what IC we are talking about, for example feeding 3v via the feed back pin of an lm2596, so while granted it's never truly off, we do have period of off and on time aka duty cycle

show me a duty cycle where it's never off.... that's called DC
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
of course it has to be completely on or off in a duty cycle or it wouldn't be a duty cycle
This is not what you stated in post #2:
a signal is sent back to the regulator to hold off on the PWM signal until the comparator of the feedback voltage is low enough to begin switching again
Which clearly, at least in my view, states that the pwm signal as such is turned on and off. This is usually (with exceptions) not the case. The duty cycle of the pwm signal is modulated to regulate the output voltage.
 
Top