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.

current mirror lenght

Status
Not open for further replies.

student_of_analog

Junior Member level 2
Joined
Sep 11, 2011
Messages
20
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,424
Why should the lengths of transistors used in current mirrors be the same ?
Razavi on page number 139 says something about Ldrawn and Leff which I do not follow.
Anybody can help me understand this?
 

There is a difference between Ldrawn and Leff. If the length is not exact, it will leads to mismatch.
 

You could see mismatch in a MonteCarlo analysis, when you have different lengths of transistors on a current mirror.
 

transistor length should be the same to achieve proper matching.
 

Ldrawn - is the gate length you use in your schematic for design. It is "Ideal" length. It is defined as distance between Source and Drain diffusion edges.
Leff is effectivre gate length after the processing is done. It takes into account underdiffusion of LDD structure under the gate.

Leff will be changing device to device. The best you can model it with Monte Carlo which adds gaussian statistical distribution to this parameter.
If you look for some data for mismatch you can see that even two neigboring devices might have different Leff and so the matching will not be perfect.
 

The only really dependable matching is instance-ratio (N:M equal geometry,
and even equal-adjacent-geometry devices). Inconvenient for non-integer
ratios but that's not often a matching interest.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top