The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> biasing current
https://designers-guide.org/forum/YaBB.pl?num=1173691456

Message started by aaron_do on Mar 12^th, 2007, 2:24am

Title: biasing current
Post by aaron_do on Mar 12^th, 2007, 2:24am

Hi all,

I've found that the resistor variation from the model to the fabricated one can be enormous. For a biasing circuit like a bandgap reference, resistors are quite necessary (although I've seen some resistorless designs). This causes the results to be very different from expected. What kind of techniques do people use to get around that? Also, what is a good CMOS bandgap for 1.8 V supply?

thanks,
Aaron

Title: Re: biasing current
Post by monte78 on Mar 12^th, 2007, 2:35am

The point is that you cannot obtain an accurate value of resistors, but you can obtain an accurate matching of resistors. Usually you design a bandgape reference where your voltage reference depends on the ratio of resistors and on base-emitter voltages. You can search here for some ideas:

http://www.circuitsage.com/bandgap.html

aaron_do wrote on Mar 12^th, 2007, 2:24am:

Title: Re: biasing current
Post by Venkateshr on Mar 12^th, 2007, 5:47am

For the Voltage BG ref the question is over matching of resistors and not individual resistor tolerance.but for Iref it is Vbg/R like ckt which depends on R inversely and in first order itself.Here we should see to the variations of resistor with process (also with temp but this can be compensated) :'(

Title: Re: biasing current
Post by aaron_do on Mar 12^th, 2007, 8:53am

thanks for the replies.

As Venkateshr pointed out the current produced by the current mirrors in the BG circuit is sensitive to resistor variation. I did find a paper which seems to have a solution to this problem though...

"A new low voltage precision CMOS current reference with no external components"

it can be found from IEEE xplore

cheers,
Aaron