Simple vs Current Mirror Opamps

[celebrevida]

In the project I am working on there are various 1 or 2 stage opamps.

(a) Simple Opamp (b) Current Mirror Opamp

I am seeing the following two types of first stage topologies.

But it isn't clear why (b) was chosen instead of (a) in certain sub-cells. I literally swapped out (b) with (a) and it seems to work just fine.

If we assume that there is a B:1 ratio in output current vs Itail in (b), then my understanding is that the current mirror opamp has a factor of "B" increase in Gm, GBW and SR but the same Av.

I guess it isn't clear why the previous designer chose (b) over (a) in those cases. If (a) is "good enough", why consume more area and power? Why am I not seeing?

Thanks in advance for any insight.

 

[dick_freebird]

(a) has a limited output swing relative to (b) which can
to rail-to-rail. If whatever's "catching" what (a) is "pitching"
can be happy with Vo range (which will be affected by both
Vi+ and Vi- pinning the tail to max(Vi+, Vi-)-VTN ), then
fine.

(b) will also deliver a lot more gain (at least near CM center)
as Rout can be higher (Best Vio will be had for (a) when Vdd-Vo
matches the Vgs of left hand load PMOS, meaning that the
Vo load will be not be in saturation and stage Rout, low.

 

[celebrevida]

Thanks for your reply and good point about (b) having larger output swing. But if the output swing with (a) is sufficient, then it seems there is no reason to use (b), right?

As for supporting larger Rout load, you could just as easily use (a) but with larger sized FETs, instead of using (b), correct?

I guess it sounds like (b) is really only needed if you need higher Vo swing and so just want to confirm this.