Rail to Rail CMOS differential amplifier design

[darkseid]

Hi

I am trying to design a rail to rail input CMOS differential amplifier. I have attached the circuit file.
Bias current source: 50u A
All transistors have W/L = 32:1.

However the output gain is only 30 dB. Is there a way to boost the output gain. I tried using a cascode amplifier second stage but its not working. In fact is gain dropped drastically.

Kindly advice a solution

 

[bigstarstar]

I guess, for the single ended opamp, current should be mirrored from P10 to P8.

 

[darkseid]

Hi

P1, P5, P8 and P10 all have the same current of 50u A.

 

[analogLow]

Hello-

Interesting amplifier. What is the advantage in feeding the input voltage to two pairs of transistors? I think increasing the output stage resistance or transconductance should increase the gain.

Cheers,
M

 

[darkseid]

Hi

Feeding the input to two transistors, increases the input voltage range i.e. rail to rail.

I tried increasing the output impedance using a cascode stage but still its not giving the required gain.

 

[bigstarstar]

If you break connection from gate of P8 & P10 to P5 and connect gate of P10 to its drain, you may observe 6dB increase. Also, reduce the current of the current source (increase rout) and largely increase the aspect ratio of the N2, N3, P3, P4 (increase Gm) ~

 

[darkseid]

Hi
You mean to say: diode connect P8 and P10. But wont the diode load decrease the gain as the output resistance of diode configuration is very less i.e. 1/gm.

I will try modifying the current and aspect ratio and will get back.


Hi

I tried modifying the aspect ratio of the input nmos and pmos transistors from 32/1 to 256/1. Now I am getting a gain of 35dB (approx. 5 dB more). Decreasing the current is not making a big difference. If I further need to boost the gain to say, 50 dB or more, do I need another stage. if yes then what type of second stage should I use.

 

[bigstarstar]

You may choose an inverter amplifier as the second stage so that the output of your amplifier is also rail to rail. If you bias this stage properly you can get higher gain. I tried this structure and I got more than 65dB. However I strongly recommend you to choose folded cascode mode so that the higher gain will be achieved easily without extra stages and even the stability compensation.

 

[darkseid]

Hi

Can you suggest a circuit/tutorial for an inverter amplifier

 

[univer_solar]

Hi,
Could you post sche. with the value of W and L? Maybe it can be helpful for people to discuss and help you.

 

[darkseid]

Hi

I modified the circuit using high swing folded cascode amplifier. Now I am getting a gain of about 70 dB and a phase margin of 30 degrees. But the problem is that, My max DC input bias can go from 0.4V to 2.2V. The low side is ok but high side is too low. Is there any way to boost the high side voltage swing to at least 2.8V. I am operating from 0 to 3V..