Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

[SOLVED] [Moved]: series RC changes the dc gain

Status
Not open for further replies.
F

frankrose

Advanced Member level 3
Joined
Nov 27, 2013
Messages
883
Helped
237
Reputation
472
Reaction score
238
Trophy points
1,323
Activity points
7,820
Hi!

I want to use miller compensation in my OPAmp's 2nd stage. When I use miller capacitor only, the DC gain of the loop is 77dB, but when I use a series RC to shift the zero, the DC gain is decreasing to 68dB for different R values.
I don't understand why, a series RC shouldn't modify DC values. STB and AC analysis give the same results, the dominant pole is at 100Hz, I sample the loop gain at 0.1Hz. And I found that where the loop gain is smaller there the 1st stage has got a bit higher DC gain and the 2nd stage has got much lower DC gain. I attached a figure about the swept value of the R versus the DC gain. R is changing from 1Ohm to 25Ohm so basically it is just metal resistance.

rzero.png
2-Figure1-1.png
 

Is the resistor ideal or you have used a foundry based one ??
 

ideal from analogLib. capacitor also.

- - - Updated - - -

It is solved. Quite annoying. I had to connect together 2 ground symbol with a 0V DC source on the testbench. A gnd cell from analogLib and a vss_inherit cell from basic library... Now the DC gain is constant 77dB, independent from the resistor value. But if somebody could explain me how did it help would be awesome. It is not trivial. *Just Cadence things.*
 

One possibility is that the simulator does not always know the correct point to ground-reference a signal to.
Whether it should reference it to:

* a ground icon,
* or Vss,
* or Vdd, etc.

Or as a simpler possibility, some simulators require that your circuit have at least one ground icon, somewhere.
 

I think the simulator uses a different ground for the MSOFETs in the AC model, and can't connect it automatically to the analogLib's gnd. Maybe the substrate has one and the corresponding voltages are changed a gm, gmb or gds then. Still I don't know why I won't get always bad value if my or your statement is correct.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top