Please Let me know the difference of betaeff & gm in spe

[Y.C Park]

betaeff

HI ALL !!!

I'm wondering about the difference of effective beta and gm, You know that
gm=beta*Vp Vp=Vdsat
But I do not understand the result of Spectre operationg point.

Please Let me know the exact meaning of betaeff.
betaeff = µ×(εox/tox) Is it right ?

Thanks ahead.
Y.C Park

 

[chinito]

spectre betaeff

if beta you're refering to is mu*cox then you can do a dc sweep and plot gm as a function of dc voltage. since you already know gm relationship to beta, do the scaling to gm and that's your beta plot. I never used beta effective that spectre gives out. beta is voltage and frequency dependent, so you really have to simulate and plot beta and look at it at frequency/voltage of interest.

 

[Y.C Park]

betaeff

Thanks chinito

I have one more question about hand calculation versus simulation result.
You know that hand calculation is only level 2 equation. Is it right?

I do the hand calculation spice level2 but there is far from the simulation result.
What shall I do?

I think hand calculation is only give to intuition and which parameter be controlled.
Is it right?

Please give me information.

Thanks a lot.
Y.C

 

[ambreesh]

gm, gds, hspice

Dear Y C Park,
If you are off your results by 20%, that is good enough.
For example, yuou use level2 equations. My foundry doesnot supply us with λ, we use level1 equation for Id calculation. Usually the results are off by 20%.
But if you are reffereing to frequency responce, yes thing go get out of hand.

 

[Y.C Park]

gm calculation pspice

Thanks for replay !!!
It will be helpful to me.

Analog Circuit Design is very long long trip.

Thanks again.

 

[chinito]

betaeff meaning

Once again, if you're refering to discrepancies between small signal parameters (beta, gm, gds, ro etc.) and simulated results for those parameters then it is OK (and quite common) to see difference of about 10-20%. Once again, it's relative. These parameters don't mean much by themselves. We use them in other figures of merit and our design specs. These parameters help us figure out these figures of merits and specs; and, put us 10 - 20% within our target. Don't get hung up on these parameters and their accuracy w.r.t. what you see in simulation when you plot them. Just know tweaking what will do what i.e. linearity of mixer is a function of overdrive voltage, so you can increase current to increase overdrive. hence, you can improve linearity (IIP3, OIP3, 1dB compression point etc.)