Ok this is bothering me, I was asked to test for a 4 resistor PMOS FET with and without the feedback resistor (Rs), and explained for the gain differences.

as you can see the gain when without the RS resistor is smaller.

My explaination: with RS: increase in Vin will increase VGS + the voltage drop across RS(which depend on ID); increase in VGS->increase in ID which in turn creates a negative feedback towards VGS; VGS end up increasing with a smaller amount; ID increase by smaller amount; lower voltage drop on RD, more voltage output

Without RS: increase in Vin will increase VGS; ID will then increase and more voltage drop at RD; Vout would be smaller.


So with the existence of the RS resistor providing negative feedback, the gain is greater??? (anything wrong with my analysis?)

 

Were you instructed to remove Rs or say , just bypass it completely or partially with a capacitor (say 10uF)?

Negative feedback is normally used to reduce the gain to a lower value whilst gaining some benefits such as improved bandwidth, input impedance etc.

With Rs = Rd and no bypass, I'd expect a mid-frequency gain of around 1. If Rs comprised say 11K bypassed plus a series 1K un-bypassed the mid-frequency gain would be ~ 12.

Just making Rs a "short" has some undesirable effects.

 

Sorry - I ignored the loading effects of R6 [100k] and the series input R5 [50k]. The gains would be slightly lower than my estimates. If the intention is to find out the effect of feedback, then the trend of decreasing gain with increased feedback would be the same.

Hope this helps

 

It might be worth checking the amplifier gain over a range of frequencies (as well as 10kHz) in your simulation.

Is the 10kHz signal you have a realistic value for the design? Perhaps try 100Hz to 10 kHz to see what happens.

I don't have your simulation software to check this.

 

Were you instructed to remove Rs or say , just bypass it completely or partially with a capacitor (say 10uF)?

Negative feedback is normally used to reduce the gain to a lower value whilst gaining some benefits such as improved bandwidth, input impedance etc.

With Rs = Rd and no bypass, I'd expect a mid-frequency gain of around 1. If Rs comprised say 11K bypassed plus a series 1K un-bypassed the mid-frequency gain would be ~ 12.

Just making Rs a "short" has some undesirable effects.

Got it dude, thanks~