Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

high swing MOS current mirror

Status
Not open for further replies.

bigsheng

Newbie level 2
Joined
Dec 18, 2007
Messages
2
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,299
I am designing a typical cascode high swing NMOS current mirror. Trying to determine on the transistor size here. In my opinion, the bottom transistor can be minimum length because its Vds doesn't change much due to the buffer of cascode transistor above, while the cascode transistor should use longer length because it can see large Vds change. But what I found out is many people are actually doing the opposite way: using a large L for the bottom NMOS and min L for cascode NMOS. Is there any reason for doing that? Thanks.
 

In my opinion, the output current is mainly determined by the bottom transistors, so the length of them should be large enough to avoid channal length modulation effect, while the cascode transistor's purpose is to shield the bottom transistor from output, it is alright to select a relative short length.
it's my opinion, pls figure it out if it was wrong.

hope it helps!
 

If possible, try to use long channel for both transistors because the total output impedance is ~ gm2*ro2*ro1 (gm2 being cascode device). However, you need to watch out headrooms constraint.
 

I agree the current is mainly determined by bottom transistor. However, since the drain of bottom transistor is shielded by the cascode transistor, actually there should only be very small Vds change of the bottom transistor. That's why I thought it is ok to use small length for bottom transistor.
 

An another point of view.
Small bottom MOST only is used in high frequency application.
If consider low offset, noise applications and bandwidth trade-offs mirroring MOSTs must operate at strong inversion level with small value of Gm. Usually this gives small W/L ratio (high ro1). Influence of Cascode MOSTs on offset and noise really negligible especially when ro1 high! So cascode MOSTs are short chanel devices with small area and large enough W/L ratio for gm2*ro2*ro1.

Good reference:

or
 

Well-said, DenisMark!
So if one uses small length for bottom transistor(low ro1), cascode MOST's offset and noise can not be negligible even at low frequency.
 

bigsheng said:
I am designing a typical cascode high swing NMOS current mirror. Trying to determine on the transistor size here. In my opinion, the bottom transistor can be minimum length because its Vds doesn't change much due to the buffer of cascode transistor above, while the cascode transistor should use longer length because it can see large Vds change. But what I found out is many people are actually doing the opposite way: using a large L for the bottom NMOS and min L for cascode NMOS. Is there any reason for doing that? Thanks.

Yes, it can see large Vds change is not due to the longer length, but due to the cascode structure provide very high impendence. so the length of the above is not improtant, but the bottom is important.
because the bottom will provide good match and current stable.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top