The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> Characterizing a current mirror
https://designers-guide.org/forum/YaBB.pl?num=1140620340

Message started by Marc Murphy on Feb 22^nd, 2006, 6:59am

Title: Characterizing a current mirror
Post by Marc Murphy on Feb 22^nd, 2006, 6:59am

Hi all, what things are important when trying to characterize a current mirror? Here are the things that come to mind:

input resistance
output resistance
dc current transfer ratio
frequency response (applying AC current signal to input and observing current output)
output current vs. output voltage (dc sweep)

Are there any others that are important?

Thanks,
Marc

Title: Re: Characterizing a current mirror
Post by huber on Feb 22^nd, 2006, 7:39am

Mismatch?

Title: Re: Characterizing a current mirror
Post by Marc Murphy on Feb 22^nd, 2006, 7:54am

huber wrote on Feb 22^nd, 2006, 7:39am:

Mismatch?

Oh yeah! I think that is captured in the DC current transfer ratio (Iout/Iin). If you get a 1:1 transfer you have no mismatch.

Title: Re: Characterizing a current mirror
Post by Paul on Feb 22^nd, 2006, 12:25pm

Hi Marc,

I believe output current vs output voltage corresponds to the output conductance/resistance you also mention. I would add that noise may be an issue in some cases. I would also mention that Iout/Iin both includes static offset and mismatch.

Paul

Title: Re: Characterizing a current mirror
Post by Marc Murphy on Feb 22^nd, 2006, 1:47pm

Paul wrote on Feb 22^nd, 2006, 12:25pm:

Noise...of course...good catch! :)

Is a systematic offset a big deal? I mean, if you know what your current offset is, you should just be able to adjust Iin to get the Iout you want, right...?

Right, I would typically plot Iout vs. Vout and then take 1/derivative(Iout) to get Rout.

Thanks,
Marc

Title: Re: Characterizing a current mirror
Post by vivkr on Feb 23^rd, 2006, 1:39am

Hi Marc,

I would say that systematic mismatch is an issue if you want precise currents, because you may know the
systematic mismatch at one corner (process, voltage, temperature), but it is unlikely to be the same
across all (P,V,T) corners.

Whether the mismatch is critical or not is largely dependent on the application. For instance, it would
be fatal in a current-steering DAC, but may be a little less serious in an opamp.

However, a significant amount of systematic mismatch is a sign of a poorly designed mirror, and may
be an indirect sign that your output impedance etc. are poor. You should ask yourself why there is systematic
mismatch. You have the same VGS, same (W/L) or a fixed ratio. The only variation comes from the VDS
variation, and if this causes significant deviation in the mirror ratio, then you will run into trouble with
some other critical parameter such as gain or PSRR.

So pay attention to the mismatch is what I would say.

Regards
Vivek

Title: Re: Characterizing a current mirror
Post by sonny on Feb 25^th, 2006, 6:12am

Over drive voltage is important. It decide output inpedance and noise.
Longer L brings high over drive voltage and good current ratio matching.

Title: Re: Characterizing a current mirror
Post by jjsnail on Mar 9^th, 2006, 2:32am

hi vivkr
how to analyze the systematic mismatch of a current mirror?

can U give a example?