1
\$\begingroup\$

I am referring to Prof. Behzad Razavi Lecture https://www.youtube.com/watch?v=GAvZLyii3aw&list=PLiDoPUX9nLkIw9EnIv_3K19wlcyJ6msYd&index=2 (Minutes 19-26)

Prof. says to use nMOS when we draw a current from a node and supply to ground and to use pMOS when we draw current from Vdd(supply voltage) and supply to another node. But why is it done like this? Why can't we use nMOS as current source to draw current from Vdd to another node?

\$\endgroup\$
  • \$\begingroup\$ A simple and partial answer is that the load connects to the drain in either case. The drain is dynamic in terms of voltage or current supplied and can approach the source voltage if saturated. \$\endgroup\$ – Sparky256 Jul 27 at 5:46
0
\$\begingroup\$

For a current source the small signal impedance provided by it is ideally infinite. In case of transistor implementations, it is crucial which terminal we are "looking into" when we are defining the small signal impedance of the current source. If we use an nMOS between Vdd, and a node to which another MOS perhaps is connected, we would be measuring the small signal impedance "looking into" the source, which gives an impedance of 1/gm, quite low, thus deviating from the actual nature of a current source. Whereas if we use a pMOS to implement it, we would effectively be "looking into" the drain, giving an impedance of ro (due to the channel length modulation, or infinite if it is excluded) which is quite a high value. You can use similar observations to ascertain the behaviour between a node and ground and which MOS to use. Hope it helped.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.