Is this a reasonable way to make a voltage buffer? Behaves well in simulation. (Voltage ranges and o...
l
Is this a reasonable way to make a voltage buffer? Behaves well in simulation. (Voltage ranges and output draw are as depicted in image.)
t
Well, what you have in principle is a single-transistor source follower, where the source voltage follows the gate voltage by keeping the drain current constant. If you then consider that you want to get rid of the voltage drop, you can do the opposite thing at the output, and regenerate the original voltage across the same type of transistor with the same current through it. If you draw that up, you'll find that what you have is a simple/standard 5-transistor follower amplifier. But, you are choking off all the current in your circuit by having a tail current going through an nFET with 0V Vgs. It is not driven by the current going in at the top. It is running in deep subthreshold which is why you get a nearly zero drop across the gate-to-source voltage of your source follower nFET. It probably works but also is going to be very noise-sensitive.
l
So what's the less silly thing to do? This kind of thing? It is indeed quite sensitive to temperature.
Ah apologies you said already "If you draw that up, you'll find that what you have is a simple/standard 5-transistor follower amplifier." Thank you!
t
This thing:
l
Yep yep 👍