[sdiy] current mirrors

[Scott Bernardi]

Ooo! Ooo! I know!  (picture little kid jumping up and down with his hand 
in the air).
The Vbe vs. Ic curves should be matched.  I.e., for a given Vbe they 
should give the same collector current, and visa-versa. The transistors 
should be at the same temperature also since that greatly affects the 
Vbe vs. Ic relationship.
A two transistor current mirror will have an error due to the base 
currents. One way you can decrease that error (by a factor of Beta) is 
to add a third transistor with its B-E junction replacing the B-C short 
on the "input" side of the mirror.The collector can go to the supply 
rail (positive rail for npn mirror, negative rail for pnp) The 
transistor doesn't have to match.
Here's a link I found about current mirrors:
http://www.eas.asu.edu/~thornton/Teaching/ece334/lectures/lecture_16.pdf

Paul Schreiber wrote:

>Yes, that's the whole *point* of being "matched"  :)
>
>Now, exactly WHAT is "matched" in a 'matched transistor pair'?
>.....Anybody.....Buhler?........(yes, I know the answer, but I want you to find out yourself).
>
>Paul S.
>dang cold (cough cough)
>
>
>
>----- Original Message ----- 
>From: "Benjamin Riggs" <benj at iinet.net.au>
>To: "Synth-Diy" <synth-diy at dropmix.xs4all.nl>
>Sent: Wednesday, November 26, 2003 12:41 PM
>Subject: [sdiy] current mirrors
>
>
>  
>
>>i am hardly an expert when it comes to transistor theory, but i think i
>>understand the basic operation of current mirrors. what i would like to know
>>is if in the basic two transistor curent mirror, do the transistors need to
>>be matched?
>>
>>B
>>
>>
>>    
>>
>
>
>  
>

-- 
Scott Bernardi
sbernardi at comcast.net

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20031126/67678512/attachment.htm>