Continue to Site

why the input mos of opamp is in cutoff when w/l is large

Status
Not open for further replies.
C

carlson

Member level 3
Member level 3
Joined
Mar 11, 2005
Messages
61
Helped
2
Reputation
4
Reaction score
1
Trophy points
1,288
Activity points
1,695
Hi,all!
I'm design the diff. op with resistive load. And I want to a big gm of input mos.
So I make the w/l very large, but the hspice show the mos into cutoff region.
I'm afraid it doesn't work.

Regards
 

When the w/l of the mosfets becomes very large while the drain current remains constant, the Vgs will be less than Vth. The mosfets work in the sub-threshold region (not cut-off region). Further enlarging the w/l has little effect to the gm if tail current remains constant.
 

    C

    carlson

    Points: 2
    Helpful Answer Positive Rating
Re: why the input mos of opamp is in cutoff when w/l is larg

Enlarge the tail current.
The differentail pair were in subthreshold
 

Re: why the input mos of opamp is in cutoff when w/l is larg

Thanks.
Is it still work normally?

Regards
 

Re: why the input mos of opamp is in cutoff when w/l is larg

if only the input pairs are into cut-off region, maybe there is no problem about the opamp. But if the loading transistors are into cut-off region, some important parameters, e.g. Gain, UGB, PM, would be changed.
Just my opinion
 

    C

    carlson

    Points: 2
    Helpful Answer Positive Rating
I notice the problem before. Acturally, input pair is in sub-threshold region but not cut-off when w/l is too large.
 

yes ,I agree flushrat point ,I have this experience!but the circuit work normal.
 

In theory, MOS differential pair is OK to work in sub-threshold region. However, the MOS pair can't be matched very well specially in sub-threshold
 

Re: why the input mos of opamp is in cutoff when w/l is larg

Hi, kwkam
why are the mos pairs matched very well in sub-threshold region?

Regards
 

Let's evaluate the current mismatch due to vt mismatch.
For saturation,
I=k(Vgs-Vt)^2
so
△I/I=△Vt/(Vgs-Vt) (1)
For sub-threshold region
I=Io*exp( (Vgs-Vt)/m(kt/q) )
△I/I==△Vt/m(kt/q) (2)
compare (1) and (2), Vgs-Vt >> m(kt/q)
so current mismach in subthreshold region is much larger. It is same to the gm of diff-pair in sub region.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top