NMOS Cascode Logic

[grigione]

Please can you help me for this homework?
Size the following circuit so that it achieves a 100 ps delay (50-50) using 0.25 μm devices, while driving a 100 fF load on both differential outputs. (VDD = 2.5V) Assume A, B and their complements are available as inputs. Other data available for NMOS are following Vt=0.43(V),Vdsat=0.63(V),k'=115x10^-6 (A/V^2),lambda=0.06(v^-1),technology 250nm


I computed current charge/discharge Ic=C*dv/dt with C=100fF and dv/dt=(vdd/2)/T con T=100ps and Vdd=2.5V ,because current In for two nmos depends on w/l then In=Ic but I don't know how to compute current for two NMOS series,can you help me?

 

[Harald Kapp]

I don't know how to compute current for two NMOS series,

The current through two (or more) components in series is always the same for each component.

 

[grigione]

I know that current through two (or more) components in series is always the same for each component,but what is the current expression ,in this case,for two nmos series with same W/L ?

 

[Harald Kapp]

Does this discussion help?

 

[grigione]

If Vds of NMOS (A) is Vdd/2 and Vds of NMOS(B) is Vdd/2 then Vds of equivalent NMOS with W/2L is it Vdd/2?the current of nmos series is
Idn=k'*(W/2L)*vdsat*((vgs-vt)-0.5*vdsat)*(1+lambda*(vdd/2))
is it correct?

 

[grigione]

mos a and mos b on saturation velocity zone and (1+lambda(vds)) for lenght modulation....but what is right value for vds?because we are in saturation zone then vds (equivalent mos) = Vdssat (a)+Vdsat(b)=0.63*2=1.26 ?please...any help??.

 

[Harald Kapp]

If Vds of NMOS (A) is Vdd/2 and Vds of NMOS(B) is Vdd/2 then Vds of equivalent NMOS with W/2L is it Vdd/

No, it would be Vdd as teh full voltage will drop across one single equivalent resistor.

but what is right value for vds?

When fully turned on, as is the case in logic circuits, Vds is near 0 V as only small leakage currents will flow (unless you have a load that draws noticeable current, of course).

 

[grigione]

Therefore in this case the factor for lenght channel modulation (1+lambda*vds) will be equal to (1+0.06*2*0.63) where lambda is 0.06 and vds=2*vdsat where vdsat=0.63..is it correct my way of thinking?

 

[grigione]

any help??

 

[Harald Kapp]

Sorry, it's been a long time I have dealt with such in-depth analyses.