How to analysis the stability of integrator/filter

[yyliang]

The stability of integrator/filter is decided by which factor? what is the infect of the feedback capacitor? Could we think that if the op amp is frequency stable and its GBW is large enough, then the integrator/filter is stable?

 

[willyboy19]

The stability of the integrator definitely comes from the opamp you are using. A stand alone integrator is almost unusable due to the lack of DC feedback path. You need to design the opamp carefully and make sure it is stable up to the targetted close loop bandwidth, not necessary the GBW -- sometimes you may find it is a waste of resource to make the opamp stable up to GBW when all you need is a fractional of that bandwidth.

 

[yyliang]

Thank you,willyboy19

What you mean is that when we design the op-amp used in the integrator,we need not to have a phasemargin larger than 45(or 60),why can we do it in this manner? would the op-amp oscillate in lower frequency than the close loop bandwidth? perhaps the feedback capacitor of the integrator(CF) can help to compensate the op-amp?

I have another question: the close loop gain of an inteegrator is Ci/Cf, so we can say the feedback factor of the negative feedback sysytem(the integrator) is Cf/Ci,which is always bigger than 1,but in a negative feedback sysytem,the feedback factor should be equal to orsmaller than 1,I am confused

Can somebody help me? Thank you in advance.

 

[nus_lin]

feedback factor is not equal to Cf/Ci. instead, it should be Ci/(Ci+Cf+Cg) where Cg is the input capacitance of opamp itself.