1__TPS56121; compensation

Hi Udo,

First, the Compensation Calculator Tool (Rev. D) does use the K-factor method to calculate Type III compensation, not Type II.

I didn't see what values you input for the calculation, but from your description, I guess the reason you got the zeros above double pole frequency is likely because the cross-over frequency Fco and phase margin PM are way high with respect to the LC resonant frequency Fr. As we know the phase will drop by 180 degree due to the LC double-pole, depending on the Q-factor, the phase drop could be very fast for high Q system. So if cross-over frequency Fco is too high with respect to the LC resonant Fr  (e.g. Fco >10x Fr), in order to follow the K-factor way to get the phase boost, mathmatically the zeros will be placed after the Fr. However, it may lead to a drop in the loop phase to less than 0 degree below Fco while still giving positive phase margin at Fco required. Conditionally stable loop may cause oscillation and other issues.

So the issue is not the tool, but the user input for Fco and PM. Typically we suggest to set Fco = 2~5x Fr. You may check whether that gives the zeros before or close to Fr and the corresponding measurement data match the calculation.

Your way of using the tool by setting the zeros manually is also reasonable.

Regards,

Na

Hello Na,

many thanks for your answer.

Perhaps I used the tool in the wrong way and entered a f_crossover that is too high?!
Now I chose a much lower f_crossover, as you suggested. Then the tool places the zeros unter Fr; see the following table:

- at 18kHz there is a maximum for k_Venable
- above f_crossover=23kHz the zeros go above Fr

Regarding the value of k_Venable I still think that the calculation does a Type II amplifier compensation!
From the above calculations we see:
- zero_of_compensation * k_Venable = f_crossover
- f_crossover * k_Venable = double_pole_of_compensation

BUT:
# Linear Technology; Reference Reading #4; "THE K FACTOR: A NEW MATHEMATICAL TOOL FOR STABILITY ANALYSIS AND SYNTHESIS"; page 6
and
# Venable Industries; Venable Technical Paper #3; Optimum Feedback Amplifier Design For Control Systems; page 9
tell that the factor between zeros, f_crossover and double_poles_of_compensation should be sqrt(k) !
So, I can't agree with you that the tool calculates type-III-compensation.
Am I wrong?

best regards
Udo

Hi Udo,

I checked the calculation tool, the "kVenable" in the excel sheet is actually the square root of the K factor in the literature.

Regards,

Na

Hi Na,

thank you for your reply.
That correspons to my result and would show that the calculation does a Type II amplifier compensation.

I suppose this behaviour is due to the fact that the EXCEL-sheet is for various DC/DC-converter-ICs (with type II and type III amplifier compensation) and the formulas in the sheet can only be expressed for one type of compensation.

Now what shall we do?
Is there a plan to split this worksheet in two groups; one for the ICs that require type II and one for those ICs that require type III amplifier compensation in order to get the right k-value?

best regards
Udo

Hi Udo,

I am afraid your understanding is not correct. Since the "kVenable" in the excel sheet is actually the square root of the K factor in the literature, it confirms the spreadsheet is indeed Type III compensation. Actually the tool is for Type III only.

Regards,

Na

Hi Na,

ok, now I understand.
Thank you for the clarification.

best regards
Udo