[SOLVED] Problem designing Opamp

[Ravinder487]

I'm designing simple two stage miller Opamp with specifications, gain =4000V/V,CL=10pF and slew rate >18V/µs
I've designed two stages seperately.Gain of first stage is around 130V/V and of second stage is 56V/V.Now the problem is that when I cascade two stages I'm not getting product of two stages.
When I checked the ICMR of second stage it is from 330mV to 340mV and first stage OCMR is above 440mV.How can I make OCMR of first stage to be same as ICMR of second stage .To the best of my knowledge no book considers this issue .Can anybody help me in sorting out this issue.
Thanks&Regards
Ravinder Jakkidi

 

[calculus_cuthbert]

Is there a reason why you have designed the two stages separately?
What kind of second stage are you using? and is your first stage a diff pair with current mirror load?

 

[Ravinder487]

I have designed them together but when I probed at output of first stage I'm getting gain above 100V/V but my overall gain is around 200V/V.
The doubt that I've raised is used to be in my mind from long way back and I thought perhaps it would be the reason for my low gain so I separated out two stage and did the analysis.
Second stage is Common-source amplifier with pmos current mirror and my first stage is single ended differential amplifier (with active load)

 

[Ravinder487]

Dear Erikl, I'm really regretting for what I did ,but its not my intention.I've been waiting for single good suggestion since last 30Hrs and I couldn't get any .In disguise I've directly messaged you as your are only the person right from the beginning giving me good suggestions.Sorry for disturbing you and thanks for your good help ,I wish it will continue in future.

 

[erikl]

Ok Ravinder,

I can't tell you a lot without more info. May be you simulated the single stages without load. Now the second stage is an essential load for the first stage, which probably considerably decreases the gain. Just a guess!

 

[Ravinder487]

I don't think so.I have fixed my W/L ratios CS using gm/id methodology and I got gain of 200V/V,but when I parametrically swept my W/L ratios, at some point I got a gain of 13KV.Again I've separated out two stages and did the analysis, this time also the same problem.