Need information about designing CMOS OPAMP in 0,35 µm technology

[sykab]

Hi!

I’m trying to design, in Cadence, a single-ended CMOS AMPOP in technology 0,35 µm with a Bandwidth of 100MHZ. I'm thinking to design a simple OTA but I haven't decided yet.
I would like to ask if someone can give me or advice me some kind of information/book to help me:?:


Thanks

 

[steadymind]

Re: CMOS OPAMP 0,35 µm

go for telescopic cascode structure followed by a class AB buffer if you want to drive heavy cap loads and low resistive loads, and also swing.

Telescopic would be more efficient than a folded cascode.

all the best

 

[jecyhale]

Re: CMOS OPAMP 0,35 µm

It is up to the load and your application. Cascode OTA is a good choice, but OPAMP can also be designed. Just use two stage.

 

[sykab]

Re: CMOS OPAMP 0,35 µm

It is up to the load and your application. Cascode OTA is a good choice, but OPAMP can also be designed. Just use two stage.

I want to do a current source which specifications can be seen in the image.

Thanks for your reply.

 

[wael_wael]

Re: CMOS OPAMP 0,35 µm

i advice u first stage folded cascode (PMOS as input) and second stage is CS (sure NMOS the input), this circuit gives u low in settling time and high gain with phase margin
regards

 

[sykab]

Re: CMOS OPAMP 0,35 µm

i advice u first stage folded cascode (PMOS as input) and second stage is CS (sure NMOS the input), this circuit gives u low in settling time and high gain with phase margin
regards

What's the meaning of "CS".

Thanks for you reply

 

[gaom9]

CMOS OPAMP 0,35 µm

CS maybe is common source amp.
But the Bandwidth of 100MHZ is too large. Is it the unit gain bandwidth, but not the -3dB bandwidth.
I think it is hard to design an OTA with 100MHZ -3dB bandwidth. It will need a very large bias current to get a large gm.

Best regards!

 

[sykab]

Re: CMOS OPAMP 0,35 µm

CS maybe is common source amp.
But the Bandwidth of 100MHZ is too large. Is it the unit gain bandwidth, but not the -3dB bandwidth.
I think it is hard to design an OTA with 100MHZ -3dB bandwidth. It will need a very large bias current to get a large gm.

Best regards!

Yes, CS should be common source... it makes sense.

Thanks.