Continue to Site

# How to adjust phase margin in ADS

#### mohamis288

##### Full Member level 3
Hello,
I want to measure the phase margin for a two-stage amplifier. I wish to adjust compensation capacitor using tuning tool in ADS and then measure phase margin from magnitude and phase graph. But, the problem I face is that every time I change compensation capacitor value, 0dB-crossing point of magnitude graph changes and I have to adjust marker position manually in order to cross 0 dB on magnitude graph and then read the relative phase to get phase margin. Isn't there a more straightforward method to perform this task?

Hello,
I want to measure the phase margin for a two-stage amplifier. I wish to adjust compensation capacitor using tuning tool in ADS and then measure phase margin from magnitude and phase graph. But, the problem I face is that every time I change compensation capacitor value, 0dB-crossing point of magnitude graph changes and I have to adjust marker position manually in order to cross 0 dB on magnitude graph and then read the relative phase to get phase margin. Isn't there a more straightforward method to perform this task?

I am not familiar with ADS - however, using a universal simulation program like PSpice it is possible to adjust a certain part within a circuit with feedback for a desired phase margin. This (new) method is described in the following document (chapt. 4).

The principle of this method is to insert a certain amount of phase shift into the closed loop and adjust a selected part of the circuit until the oscillation condition (zero phase margin) is fulfilled. Then, without this additional phase shift, the circuit has the required margin.

Here is an excerpt from the document:

4 Design Aid for a Specified Margin
Unlike the classical loop-gain-based procedures, the described phase-slope method
also offers the possibility to support the design of a system that requires a specified
phase margin. If both blocks HZ and GL contain a fixed phase shift that is identical to
the required phase margin (ϕZ = ϕPM ), any suitable component of the feedback loop
may be used for parameter stepping in place of ϕZ . However, it can be expected that
there will be a remarkable non-linear relationship between the stepped parameter and
the respective group delay magnitude maximum. Therefore, it is advisable to select
more than just two values for parameter stepping. Again, the desired component value
that ensures the required margin can be found using the goal function in (2).
For example, to provide the circuit in Fig. 1 with a phase margin ϕPM = ϕZ =
60◦ the load capacitor CL was stepped during the ac analysis over the range CL =
(0.1 . . . 1) nF in steps of 0.1 nF. Then, at the zero crossing of the goal function (2), a
value of CL = 0.285 nF can be identified to allow the required margin.

PS: If desired I could provide you (as the author) with a copy of the article (via email)

Last edited: