Why the current is not right?

[wjxcom]

Hi all: I design a classAB OPA, this opa can be found in the paper"A High-Performance Micropower Switched-Capacitor Filter" writed by RINALDO CASTELLO and PAUL R. GRAY.

But when I design this opa, I found a problem:
In this opa (the schematic of this opa can be found at the attachment), for the bias circuit, for PMOS transistor, the current is 10u when the W/L of the PMOS is 12/2 and the m is 2, the current of NMOS is 10u when the W/L of the NMOS is 10/2 and the m is 1.

in the OPA, I think the current of M29 and M30 is 100u, so I make the w/l of M29 and M30 is 12/2 and the m is 20. but after the simulation, the current of M29 and M30 is only 60u!

I do not know why! help me please, thanx!!!

 

[Fom]

We don't know your operation voltage, transistor models.
But you have the simulation result. So check all currents and find where it is different from expected value.

 

[wjxcom]

Hi Fom: I think the transistor models can be found at the attachment, named csmc.lib.

I have simulate the circuit, and I do not know why the current of M29 and M30 is only 60u!!

regard!

 

[ieropsaltic]

Check the modes of operation of the transistors where the current is 60u. They may be out of saturation and into triode .If they're still in saturation but still have error, try increasing L of these transistors while keeping W/L constant .This would cause higher rds thus less current error (but I doubt that this is your case as 40% error is too large ,probably your transistors are out of saturation ) .

 

[hr_rezaee]

Hi
as I see in the schematic
M8 and M6 generate current of M30 nor vice versa.
in quaesent,
VGS2+VGS4=VGS8+VGS6
but size of M6 and M8 are 10 times of M2 and M4.
so, Vt2 is not equal to Vt6 and Vt4 is not equal to Vt8
and from schematic, Vds2 is not equal to Vds6 and Vds4 is not equal to Vds8
therefor current of M30 is not 10 times of M2.
if you want to have exactly 10 times you must set them manually with respect to simulation results.
(fine tunning)
regards

 

[wjxcom]

Hi ieropsaltic: from the simulation results, I found that these transistor DO in the saturation region! may be I should try increasing L of these transistors while keeping W/L constant as you said!

Hi hr_rezaee: thanx you to help me again. from the result, Vds2 is not equal to Vds6 and Vds4 is not equal to Vds8. you are my savior!

At the same, I have another question: transistor M1, M2, M5 and M6 is PMOS transistor, if the BULK of these transistor is connect with the AVDD(i.e. power supply) and not connect with the source of these transistor, then the current of M1 and M2 is 7u, rather than 10u! It is so strange! may be Vds of these transistor is making trouble!!!

 

[hr_rezaee]

At the same, I have another question: transistor M1, M2, M5 and M6 is PMOS transistor, if the BULK of these transistor is connect with the AVDD(i.e. power supply) and not connect with the source of these transistor, then the current of M1 and M2 is 7u, rather than 10u! It is so strange! may be Vds of these transistor is making trouble!!!

Yes, I think answer is Vds.
see simulation results.