Why do you call the reference in Fig. 20.14 a Beta-Multiplier Reference (BMR)

instead of calling it, as you show on the top of page 625, a constant-gm reference?

 

The BMR is only a constant-gm reference when the MOSFETs follow the (long

channel) square-law equations. Since CMOS circuits in production for the

last 20+ years can't (accurately) be described using the square-law equations it’s

best to describe the reference by how the MOSFETs are related, one’s Beta is

multiplied up from the others.

 

Also note that the name BMR is similar to the name BGR (bandgap reference).

 

Return