141N. 1st Order high-frequency TTC analysis of common source and common drain
HTML-код
- Опубликовано: 8 сен 2024
- Analog Circuit Design (New 2019)
Professor Ali Hajimiri
California Institute of Technology (Caltech)
chic.caltech.ed...
© Copyright, Ali Hajimiri
I've been seeing the series from the beginning and understand well all of them until this video. The calculation of R_pi, Ru, R_theta, Gm, etc. is difficult to understand.
@ 30:26 what about the current path through R1, as we only null the independent source?
@23:00 while calculating the resistance R_theta there seems to be a mistake. As per my analysis the resistance seen by C_theta should be "(R2/(1+gm*R2))+R3" instead of "(R2+R3)/(1+gm*R2)", in other words only R2 would be scaled downby (1+gm*R2) and added to R3. Am I correct.
@19:30 you draw the model for the transistor with the capacitor and resistors attached so that determining the time constants is not time consuming the next time around, what I am curious about is that how do you take that C_theta into account, because most of the resources I have seen include a drain-bulk or source-bulk capaciatance instead of the C_theta capacitance.
C_theta basically is the series combination of drain_bulk and source_bulk capacitors.
Why do we need to consider CPI and CU separately?
Teacher, can you turn on the CC subtitles?
Does anyone have the proofs for all the equivalent resistances?
@23:21 I was not able to find this formula for Gm exactly. Is this exact? I'm getting a factor (beta+1)/beta on R2
Hello, Do you mean the Gm for the BJT counterpart?
@@jjjjjjjjjjj12345 from wat I remember, it was about a generic formula for both BJT & MOSFET. The formula should work for mosfer setting alpha -> 1 (beta -> infinity), but my nitpick was only significat whith BJT.
That's because you're right... :) There is in fact that term before R_2. My bigger concern here is the derivation of these formulas (for R_mu specifically). I mean how can we derive them fast and effectively? I used the Blackman impedance formula but what I obtained had a quite different - a bit more difficult - form. I also applied Middlebrook's extra-element theorem. The result was equally different and equally difficult. All 3 of them (the Blackman, the Middlebrook and the Hajimiri) gave me the same result though numerically. By the way, I only know that you're right about that (beta+1)/beta factor because back then when I first read Mr. Hajimiri's IEEE article about this I left some hand-written side note there exactly about this difference... I must have used nodal analysis tool then.
This is just the emitter generation gain, but without the collector resistance. Prof. Hajimiri talks extensively about this in the BJT amplifier videos early in the series. The gain with emitter degeneration is basically -(total resistance on C side)/(total resistance on E side), which without the resistance on C side is 1/(rm + RE). And, as always, it's just an approximation assuming ro is infinite. When ro is infinite, then in fact what happens is that the gain is actually Gb*vb - Ge*ve, where Gb and Ge are complicated expressions, and depends on ro