The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> About designing current mirror
https://designers-guide.org/forum/YaBB.pl?num=1283267001

Message started by 西楼 on Aug 31^st, 2010, 8:03am

Title: About designing current mirror
Post by 西楼 on Aug 31^st, 2010, 8:03am

Hello all,
When I designed a current mirror,my design is hard to match what I expect.
My design is expected to realize the function of a CCCS(current control current mirror).That is,when the input is I1,then two output would be 9I1,-9I1.
In general, my design is as follows:
I set mn=10, m31=90,m32=90
The problem of my design is dc current is very accurate,meanwhile ac response is very dissappointing.
How to improve or solve my problem?
Hope you can give your excellent suggestions,thanks a million. :)

Title: Re: About designing current mirror
Post by AnalogDE on Aug 31^st, 2010, 9:09am

Minimize L as much as possible (this reduces mirroring accuracy)
Minimize W as much as possible (Pay attention to headroom)

Avoid a 1:9 ratio and use multiple stages, i.e. 1:3:9

Title: Re: About designing current mirror
Post by loose-electron on Aug 31^st, 2010, 10:49am

Trying to get bandwidth out of a current source?

Use a fixed current source and put a differential pair (current steering) above the current source and get you bandwidth there.

Title: Re: About designing current mirror
Post by 西楼 on Aug 31^st, 2010, 8:02pm

Thank you for your kindness. You are right.As you said,W,L must be small as much as possible.When I change the W and L(randomly but of course be small),the ac response is much more better(simulation graph pasted below with R2/plus input signal;R0/plus,R1/plus output signals).It is 10:10 ratio.
However,the input current is also pull down.I have to tradeoff.Normally,input current can reach 20.8uA without any load.
The output signal,R1/plus,has a little ripple(the zoom in graph of R1/plus pasted below).
Besides,loose-electron,would you like to explain how to get the bandwidth of my design in detail? I can't quite follow you.Thank you again.

Title: Re: About designing current mirror
Post by 西楼 on Aug 31^st, 2010, 8:19pm

The zoom in graph of R1/plus is pasted below.

Title: Re: About designing current mirror
Post by loose-electron on Sep 2^nd, 2010, 8:06pm

You need to explain a lot more clearly the specifics of what you are trying to do.

THere are active amplifier methods of driving gate bias to create a wider BW structure, but then its not a "current source" it is a common gate amplifier.

A "current source" is generally a fixed bias structure.