current-mirror
In designing a Common-Source amplifier, why is the inclusion of a current source considered most popular method?
Consider a single stage CS amplifier where I bias the gate with an appropriate DC voltage and derive the Id through Vdd. Is it considered a good design? In doing so, I am replacing my current source circuit with a Vdd from which i can derive the drain current.
Here is a unconventional explanation (you can find many other conventional explanations on the web)...
The basic idea is to oppose two current sources (transistor collectors); so we have to connect them in the same loop and to vary their currents in opposite directions:
As a result the voltage of the common point between the sources will "move" extremely vigorously (a fully geometrical phenomenon):
A good example of this configuration is the classic CMOS stage where the two transistors are connected with their sources to Vdd and Vss (ground) and their drains are joined:
But in the emitter-coupled (long-tailed) pair, the collectors of both transistors (T1 and T2 below) are turned up. So, we should turn the one collector (the T1's in this example) back down to stand against the other (T2's) collector. For this purpose, we have connected the simple current mirror (T3-T4) in the picture below:
See also this paper and its more sophisticated version.
The connection is basic to how a current mirror works. The two gates are connected together so that both FETs are driven to the same level. This assumes well-matched FETs, as would be the case with them being next to each other on the same IC.
The way the current mirror works is that the desired current is forced D-S thru M1. With its gate connected to its drain, the gate will automatically be driven to whatever it takes to support the desired current. This same voltage is then applied to the gate of M2, which in theory then allows the same current.
The purpose of M1 is to provide the right gate voltage for the desired current. The D-G connection is the feedback that allows that, in fact requires that, to happen.
Best Answer
If you want to blow up that transistor in a hurry, yes that's a great design. Vdd at the drain of a CS amp will allow huge amounts of current to go straight into your CS amp. If this is an NMOS, once you turn on the transistor by raising the voltage to a given level, the resistance of the source/drain path is virtually 0. Let's say it's actually 0.1 ohms and your Vdd=1V. This would allow 10amps of current through your transistor. You'll let the magic smoke out.
The other issue with putting Vdd at the drain is that you now have no method to use the output of the amplifier. All of the current comes from Vdd and drops straight into ground. There's no longer a place to tap in and measure voltage or current.
The reason why a current source is used is to prevent exactly what I describe above. A current source limits the amount of current that can be sent into the CS amp. An ideal current source also has an output resistance of infinite so you'll maximize the gain of a CS amp. Having this other component as part of the CS amplifier allows the voltage to swing freely between Vdd and Gnd without being pinned to either.