Very large gm (transconductance) values

[george2525]

Hello.

Ok so im looking at some feedback circuits and saw this one in a book. the derivation for B is made assuming that gm is enormous.

(see below)

How can gm be so large? ive never heard of this before! the circuit does not seem weird to me and ive never calculated a gm greater than order of millisiemens before

I have no idea. Tried google. not mentioned in any article I can find

Any ideas?

 

[george2525]

ps I would have said B = r just by looking at it but the logic here realy bothers me

 

[(*steve*)]

I think that the assumption that gm is large means that Vgs remains constant. I believe this is assumed to make things cancel out nicely.

 

[(*steve*)]

Hmmm. That ignores the constant current sink which acts as an infinite value resistor connected to an infinite -ve voltage. (with a certain definite ratio between those infinities)

edit: Ignore this, I must still be asleep. It doesn't change anything I said above.

 

[george2525]

Hi thanks for the reply. why would constant Vgs result in large gm?

also I would be so grateful of you can explain the notation in the schematic which I have never been sure of

- do you know what the arrows at the top and bottom of the circuit mean? I have never seen this explained in a book. I have been assuming they are going to DC supplies and can then be grounded for small signal analysis. is that right?

- also, are the current sources here for the DC bias?

- does the infinity symbol on the capacitor mean infinite capacitance? This is used quite a lot in some books but ive yet to see it explained.

- is the voltage source for both AC and DC?

If you know these things it will help me a lot as unexplained notation causes me so much frustration

Thanks

 

[(*steve*)]

Hi thanks for the reply. why would constant Vgs result in large gm?

It's the other way around. The higher the gm, the smaller the change in Vgs required to achieve a given Id.

because the source load is a constant current, Vs assumes the appropriate value with respect to Vg so that Id is the value of the constant source load. If the Vg changes, Vs changes to maintain this.

The circuit you have also has a capacitor connected to the source. This will cause a different small signal behavior.

Assuming a high gm means you can assume Vgs remains constant even under this condition ( I think).

 

[LvW]

May I ask you: Which book?
To me, the figure together with the given "explanation" looks a bit weird.
Examples:
* gm is "very high" (what does this mean)?
* basic gain is "high" (??). That`s all? No gain expression?
* I suppose the resistance "r" in the circuit is identical to rs=1/gm ?
* The assumption that r=1/gm would create a seres-series feedback effect is simply wrong!
(I think, even the assumption that 1/gm can be modeled as a resistor in the emitter leg is not correct! Formally, 1/gm is a "transresistance" but not identical to a classical two pole resistor).

My recommendation: Never blindly trust any formulas in textbooks. Even printed papers can be wrong.

 

[george2525]

its from a lesser known sedra book that is supposed to clear up the lack of explanation in the main book. its called K.Cs something. sorry i dont have it to hand

and yes I am aware of how vague textbooks enjoy being