[SOLVED] Very Low Differential voltage amplification

[Fabien]

My amp is near 100dB with near 200MHz bandwidth for half a µA! But not stable at all!



I know it's challenging, but if it wasn't I would have been jobless!

 

[Ata_sa16]

Ofc it wont be stable compensation will decrease your bandwidth. If you design 2 stage without compensation your bandwidth becomes high but it is useless.

- - - Updated - - -

Wait i will find you a solution. I have to check my old designs of ultra low power opamp. I will come back.

- - - Updated - - -

Also you have to know what is your load capacitance. You have to include Cload in your design.

 

[Fabien]

Ofc it wont be stable compensation will decrease your bandwidth. If you design 2 stage without compensation your bandwidth becomes high but it is useless.

- - - Updated - - -

Wait i will find you a solution. I have to check my old designs of ultra low power opamp. I will come back.

- - - Updated - - -

Also you have to know what is your load capacitance. You have to include Cload in your design.

Yes, a load is included at the end of the design. But the amp is driving a comparator, so very small load, hopefully! That's the only way to reach such a high gain with low current!

Don't be in a hurry, I'm also working on it! Thank you anyway!

 

[Ata_sa16]

what is your CL value?

- - - Updated - - -

also cau you give me your tech hspice models ?

models of PMOS and NMOS

 

[Fabien]

what is your CL value?

- - - Updated - - -

also cau you give me your tech hspice models ?

models of PMOS and NMOS

The CL value is small, like input MOS transistor (Grid). 10fF is good enough.
I spent the day trying to design OpAmp with variable gain (resistor feedback), but seems to be so tricky! Not able to get InM = InP!
I'm sorry but I don't have access to the models of my transistors.

 

[Ata_sa16]

since you are using a comparator, i think 20dB gain is enough for you.
Try this you can reach 31dB min gain.

 

[Fabien]

Hi Ata_sa16, thank you for your help.
I try to design a telescopic opamp, just on diff pair. All transistors in sub th, I'm able to achieve 67dB from DC to 1MHz (-3dB) with a GBW around 1GHz. Even this, it's not enough, not sure a 30dB gain opamp will be suitable. **broken link removed**

Well, I think my comparator has to be improved. May be I'll try your design, the trouble with the telescopic cascode pair is that the inputs are very bias dependent. A small change of the Vdc move drastically the gain...

 

[deba_fire]

You could use a cross coupled pair for amplification like a strongARM latch.

 

[Fabien]

You could use a cross coupled pair for amplification like a strongARM latch.

Hi Deba, thank you for your reply! Never heard about StrongARM latch, but after a quick reading of papers it seems to be very interessting! I think it will be suitable for my need.
Thank you again!