need help to evaluate two circuits

[kickbeer]

helo,

in the attachment below, there are two differential amplifier. One is differential amplifier with resistive load and the other one with active load. the differential amplifier with resistive load is the original one which i took from ieee article. the second one(with active load) is the modification i made. i did this because i want to increase the gain. My question is, did i modify the differential amplifier (with resistive load) correctly? i doubt this since i connect the drain of both M18 and M19 directly to the gate of M5 and M6.

 

[diemilio]

You're right... it's wrong. You need to somehow sink the current from M18 and M19.

diemilio

 

[kickbeer]

You're right... it's wrong. You need to somehow sink the current from M18 and M19.

diemilio

how about the other one in attachment?

 

[diemilio]

That one is right because now you have an AC path for the currents of M3 and M6. The thing is that there's no advantage between this topology and the one with the resistors (from the gain point of view) because the impedance at those two nodes (gates of M4 and M5) will be dominated by the impedance of the diode connected MOS, so it will be close to 1/gm which is usually quite small.

Why are you trying to use this topology? Is there any reason in particular? Is there any advantage of having this configuration instead of a diff amp followed by two inverting stages or a folded cascode amp for your application??

Regards,

diemilio

 

[kickbeer]

i'm trying to increase the bandwidth but somewhat in the simulation was similar to that one with resistive load. What is your opinion? You can see the original article from which i took that circuit in the attachment