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.

Monte Carlo Mismatch Analysis Help

Status
Not open for further replies.

bunda_bindaas

Junior Member level 1
Joined
Mar 29, 2005
Messages
19
Helped
1
Reputation
2
Reaction score
0
Trophy points
1,281
Activity points
1,476
I am running the Monte Carlo mismatch analysis on a Miller OTA from -3 sigma to +3 sigma.

This kills my performance for the Miller OTA on some random patterns.

Basically the node labelled Y in the adjoining pic shows a lot of variation even when I increase the L of all devices. Even a 200 mV change on Y kills the mismatch analysis of the OTA.

Please suggest some scheme to get a better mismatch analysis result
 

You are you simulating this circuit in open-loop, aren't you ? you must do it in closed-loop (i.e. with the feeddback network).

Imagine that this circuit has a (open-loop) gain of 1000 and that the input referred offset voltage is 10 mV -> then the output voltage (with the inputs short-circuited) would have to be 10 V. As this is larger than VDD, the circuit saturates.

In practice any amplifier works with negative feedback, and what you want to measure is the offset voltage of the complete circuit (including feedback).

Hope this helps.
 

Thanks for the help but my OTA is not a standalone circuit. And as such the offset can not be compensated if it's used inside a system. Therefore if I have to run the mismatch analysis using Monte Carlo, without increasing the area, is there any way in which I can reduce the effects of mismatch?

I tried using a common mode feedback circuit to the output of the OTA. That helps in the mismatch analysis but it now becomes a closed loop system and its gain response shows peaking effects.

Can I use a CMFB circuit when the OTA is working as a comparator? What kind of analysis is required to characterize the OTA with the CMFB?

Any help will be appreciated.

P.S. I can print the results in case you need to see them
 

If you are using this circuit as an amplifier you must use feedback, and you don't have to compensate anything. When you apply the feedback network arround an amplifier its operating point is automatically adjusted (for an infinite gain, its input voltage becomes the VOS).

CMFB is only needed in the case of a differential output, which is NOT your case.

A high gain amplifier can be used as a continous-time comparator (again no CMFB is needed). However, if you have a clocked system it is usually more convenient to used a latched comparator, which uses positive feedback (faster&lower power).

Rgds
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top