1
\$\begingroup\$

In case of switch mode power supplies, we uselly use "Error Amplifiers" (Type 1,2 & 3), my question is: The output voltage of this amplifier (Verr), is it the DC error of the Vref and Vout (i.e: Verr=A(Vout-Vref), A is the DC gain of the error amplifier)? if is not, please explain to me what is it and what is the role of DC gain in that case??

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
3

1 Answer 1

3
\$\begingroup\$

(paraphrased) Is the output of the error amplifier a DC error (i.e., \$ V_{err} = A (V_{out} - V_{ref})\$?

No. Now would be a good time to find some reference material on op-amp circuit design and do some studying.

The "error amplifier" is a PI controller with some band-limiting on the proportional term. The band limiting is almost certainly there to reduce the amount of ripple in the output and avoid sub-harmonic oscillation.

Moreover, the error amplifier is inverting: the higher that \$V_{out} - V_{ref}\$ gets, the faster \$V_{err}\$ will trend downward.

You can tell this by looking at the feedback network: there is not a DC path from \$V_{err}\$ to the \$V_-\$ input of the op-amp: both paths have blocking capacitors. That means that the output always integrates.

It is likely that the R4*C2 time constant is chosen to be longer than the PWM frequency, but shorter than the desired settling time of the power supply. C3 and R4 are chosen to stabilize the supply while having it respond as rapidly as possible to variations in the load.

\$\endgroup\$
4
  • \$\begingroup\$ If you are right, how Verr contributes or compared to Vsawtooth to generate PWM signal?? \$\endgroup\$ Feb 1, 2019 at 17:47
  • \$\begingroup\$ The higher that \$V_{err}\$ goes, the higher the duty cycle. Do you understand what a PI controller is? \$\endgroup\$
    – TimWescott
    Feb 1, 2019 at 19:22
  • \$\begingroup\$ Yes, I do, and I know what the feedback theory does for system stability, but in here the output of the E/A is continuously compared to sawtooth with a fixed amplitude. So what I did not understand is : "what is the nature of the error that is compared to the sawtooth?? \$\endgroup\$ Feb 1, 2019 at 19:45
  • \$\begingroup\$ Calling it \$V_{err}\$ is a misnomer, if that helps. It should be called "control voltage". There is no explicit ground-referenced error voltage in that circuit -- the closest thing to that is the voltage difference between the reference voltage and the negative input terminal to the op-amp. \$\endgroup\$
    – TimWescott
    Feb 1, 2019 at 21:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.