FilipeTavora
Newbie
Hello,
It may be just a silly question, but once my initial results are a little bit strange I decided to ask here for help. I am designing a folded cascode opamp using gm/id methodology at a 0.35um process. First of all, I downloaded the spice model bsim3v3 version 3.2.4. level 53 from the foundry website, which is intended for Eldo (Mentor simulator). Since I don't have access to professional tools (Cadence, Mentor, Tanner...) at the moment, I am trying to get by with the old free SPICE software, in my case, NGSPICE.
I wrote a spice script to extract and plot gm/id(W/L) x Id/W/L, Va (Id/gds) x Id/W/L and Av (DC gain) x Id/W/L for nmos transistor, as can be seen on attached files nmva_gmoverid_extraction_code.txt, nmos_gmoverid.png, nmos_va.png and nmos_DC_Gain.png, respectively. The curves seams to be OK for this nmos. So, I tried to do the same thing for the pmos, but the curves seams a little bit odd to me (See attached files with "pmos" prefix). So, my questions are:
1) Is It normal to have gmoverid curves with the kind of shape depicted on pmos_gmoverid_extraction.png (I never saw this shape on literature)?
2) How do we set the testbench to extract PMOS gmoverid curves properly?
Thanks and regards,
Filipe
It may be just a silly question, but once my initial results are a little bit strange I decided to ask here for help. I am designing a folded cascode opamp using gm/id methodology at a 0.35um process. First of all, I downloaded the spice model bsim3v3 version 3.2.4. level 53 from the foundry website, which is intended for Eldo (Mentor simulator). Since I don't have access to professional tools (Cadence, Mentor, Tanner...) at the moment, I am trying to get by with the old free SPICE software, in my case, NGSPICE.
I wrote a spice script to extract and plot gm/id(W/L) x Id/W/L, Va (Id/gds) x Id/W/L and Av (DC gain) x Id/W/L for nmos transistor, as can be seen on attached files nmva_gmoverid_extraction_code.txt, nmos_gmoverid.png, nmos_va.png and nmos_DC_Gain.png, respectively. The curves seams to be OK for this nmos. So, I tried to do the same thing for the pmos, but the curves seams a little bit odd to me (See attached files with "pmos" prefix). So, my questions are:
1) Is It normal to have gmoverid curves with the kind of shape depicted on pmos_gmoverid_extraction.png (I never saw this shape on literature)?
2) How do we set the testbench to extract PMOS gmoverid curves properly?
Thanks and regards,
Filipe
