Help me solve a problem with two stage cherryhooper amplifier

i used a two stage cherryhooper amplifer as a limiting amplifier. But it only works when i give small singal(<30mv) in tran simulation.
The total gain is 30db. i use jazz 0.35bicoms,frequency 10gb. i met convergence promblem if i give a large input.
i think it shoule achive limiting state with large input. Could someone help me?

even i also faced same problem in cadence with cherry hooper amplifier.

you used bicmos can you show schematic.....

i think i can help you.

thanks a lot raj, i will give you the schematic.
this is one stage of my amplifier

Added after 9 minutes:

sorry,maybe the pic is so small,i will give you a bigger one.
thanks again

it is the same architecture that i used.
and i also got problem in transient but i did not got any problem with ac simulations.

convergance is a kind of divide by zero error specially found in cadence software.

so i used ADS to simiulate it and got good results.
what specifications you are looking for?

i think i need bandwith>14Gb,gain>15db
i need a two stage >10G >30db amplifier

i think these spec can be achieved with given process.
just make sure that all mos transistors in sat region and i think gain would not be a problem but challenge is a b/w.
so far what specs you achieved?

actually i could get b/w and gain in one stage,also i could get a 30db 10gb with 2 stage,but i find it's hard to get a flat amplitude response. so the eyedigram is so bad.
have you met this question?

actually my requirements were less than what you require here so not a problem but because of BJTs used i think may be that's why it is source of noise.

or if you can send me the technology files then i can try woth this architecture and get back to you....

email: rajharvijay@yahoo.com
give me your email address also.

raj,i'm sorry i can't sent the architecture to you for some reason.
i'm glad to know you,my email is wang.yuanzhuo@gmail.com