1
\$\begingroup\$

I do not understand "The circuit on the right is a feedback amplifier with loop gain T. Since all time constants in that loop are of the same order of magnitude, they create a system with several poles"

How do I derive the feedback loop gain T = (gm1)*(Rin) ?

Could anyone point me to some maths reference / equations ? Razavi book does not mention about the circuit on the right though.

enter image description here

Note: the current mirror on the right is a 4T Wilson current mirror https://en.wikipedia.org/wiki/Wilson_current_mirror#A_Four-Transistor_Improved_Mirror

Let me start with 3T Wilson mirror in BJT implementation.

enter image description here

A feedback path is thus formed that regulates IC3 so that it is nearly equal to the input current, reducing the systematic gain error caused by finite βF.

systematic gain error due to finite βF ?

\$\endgroup\$

1 Answer 1

Reset to default
1
\$\begingroup\$

Yeap----those topologies tend to oscillate. Various bipolar opamps in the semiconductor industry had those problems, from what I was taught in OpAmp design class.

\$\endgroup\$
3
  • \$\begingroup\$ From en.wikipedia.org/wiki/Barkhausen_stability_criterion, how do I prevent oscillation other than keeping the loop gain less than unity ? I understand that M1 provide 180 degree phase shift and M4 provides zero degree phase shift. \$\endgroup\$
    – kevin
    Commented Oct 13, 2017 at 6:53
  • 1
    \$\begingroup\$ You need a good high frequency model of the transistor; that includes the resistive paths from terminals (E B C) into the active region, thus the R+C phase shifts are (crudely) described; the real elements are usually "distributed" and messy. The specific silicon layout sets the parasitics; some bipolars have 2 or 4 or more base contacts, perhaps on both sides of the emitter. FETS with stripes can be better modeled, with that awareness. The substrate/well ties provide delays for charge collections and that modeling is your task. \$\endgroup\$
    – analogsystemsrf
    Commented Oct 13, 2017 at 16:29
  • \$\begingroup\$ I am reading "Ultra High-Compliance CMOS Current Mirrors for Low Voltage Charge Pumps and References" from Charlon. It is quite good. \$\endgroup\$
    – kevin
    Commented Oct 14, 2017 at 15:23

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.