The phase for a pole or root at the origin is given as either + or - atan(w)

So why is this a constant 90 degrees? for w=0 this should be 0 degrees, for w -> inf this should go to 90 degrees, yet we plot it as a constant 90.

I am confused.

  • \$\begingroup\$ Active schemas can have constant 90 deg phase plot. About which schematic we're talking? \$\endgroup\$ – Looongcat Sep 8 '15 at 9:39
  • \$\begingroup\$ A pole at the origin is \$1/s\$. Taking this into the frequency domain gives \$1/j\omega = 0-j/\omega\$, which in turn gives \$\phi=arctan(-(1/\omega)/0)=-90^o\$ for all values of \$\omega\$ \$\endgroup\$ – Chu Sep 8 '15 at 15:31

Let us say transfer function is H(jw)=1/jw=-j/w. Its a single pole system with pole at origin.

The phase of any transfer function is given by atan(IMAG(H(jw)/REAL(H(jw))). So

Phase(H(jw))= atan(-w/0) =-90deg for all w.

Hence, phase starts with 90deg and remains constant thereafter as maximum phase lag a pole can give is 90deg.

The same may not be true for a pole at any other frequency. Lets say that pole is at w=a. so, transfer function is


Phase for this transfer function will be -atan(w/a)).

For w< Phase=0deg

For w>>a ==> Phase=-90deg

For w=a ==> Phase=-45deg.

Hence, in this case Phase is not constant, it starts with 0 and falls to -90deg.

Hope this helps..!!


Who told you that it would be "constant 90 deg"? It is not.

The open-loop gain of an opamp (phase-compensated) can be approximazed by a first-order lowpass (within the active region) with a rather low cut-off frequency (wc=10...100 Hz). Hence, the phase for w=0 is zero deg. The phase at w=wc is app. -45 deg (approximately, because there are other poles where the gain is <0 dB). Only for frequencies far above the cut-off wc the phase is app. -90deg.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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