|
发表于 2010-12-30 07:29:46
|
显示全部楼层
本帖最后由 fhchen2002 于 2010-12-30 09:34 编辑
No, the current through M19 and M20 are not necessarily equal.
But, you have the freedom to make the current roughly the same in your design.
These two devices are used to bias the output M25 and M26 devices.
VG of M25 is set at VDDA - VGS_M21 - VGS_M22 + VGS_M20. (approximately 1 "diode" drop from power.)
VG of M26 is set at 0 V + VGS_M23 + VGS_M24 - VGS_M19. (approxaimately 1 "diode" drop above ground.)
DC analysis
Assuming balanced inputs,
The current through M16, I(M16) = Ib4 - 0.5 * Ib1
I(M19) + I(M20) = I(M16)
I(M19) is mirrored from Ib8. I(M19) = (W_M19 / W_M23) * Ib8 (assuming same lengths for M23 and M19)
I(M20) is mirrored from Ib5. I(M20) = (W_M20 / W_M22) * Ib5 (assumping same lengths for M20 and M22)
Additional notes
Usually M25 and M26 output directly drives off-chip (i.e., they are large devices) and are subject to ESD design rule.
So are M19, M20, M22 & M23 subject to ESD rule, since we want good VTH tracking. (They are small devices.)
Yes, this OpAmp is able to drive 1 KOhm resistive load. |
|