The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Modeling >> Semiconductor Devices >> PSP MOSFET
https://designers-guide.org/forum/YaBB.pl?num=1155918151

Message started by nikunj on Aug 18^th, 2006, 9:22am

Title: PSP MOSFET
Post by nikunj on Aug 18^th, 2006, 9:22am

I want to simulate different MOSFET charactresistics(Id-Vgs,C-V,etc) through cadence spectre. i m using PSP MOSFET Model. i have installed this model library dynamically but haveing some problems on programming part(like creating netlist for how to swap VGS & VSB ...)

Can anyone help me....

Title: Re: PSP MOSFET
Post by Geoffrey_Coram on Aug 21^st, 2006, 10:10am

Why do you want to swap Vgs and Vsb??

What programming language are you using? Are you writing a perl script to generate a spectre netlist for the simulation?

Title: Re: PSP MOSFET
Post by nikunj on Aug 22^nd, 2006, 10:32am

I am looking for I-V(D.C) characteristics of MOSFET model. Say example for transfer characteristic I am plotting Id ver Vgs (range is 0 to 1v) for different values of Vsb.

More problem I m suffering from is Dynamic loading of MOSFET. I am using cadence spectre professional version in which PSP MOS MODEL is not there inside the tool. So I have installed dynamically. Installation is perfect but when I was running netlist it gave error like “devise is not recognize”

Do I need to make interface which behaves like handshaking between my library to cadence?? Or I need to generate KEY or CRACK the devise model for cadence?? If I have to do this then question is what is meaning of dynamic loading??

If you could give me some pointers, it will really help

Title: Re: PSP MOSFET
Post by nikunj on Aug 22^nd, 2006, 10:33am

Title: Re: PSP MOSFET
Post by Geoffrey_Coram on Aug 22^nd, 2006, 1:05pm

What do you mean "installation is perfect"?

Do you have a .cmiconfig file that loads the appropriate libraries? What do you see when you type "spectre -help psp1000"? You should get something indicating what version of the SiMKit is being loaded, and then some documentation on the model.

Title: Re: PSP MOSFET
Post by nikunj on Aug 22^nd, 2006, 3:01pm

Correct....
.cmiconfig file is updated properly so when i will type this command spectre –h psp1010
I am getting msg

“This Device is supported within altergroups
This devise is dynamically loaded from the shared object
/home/gnikunj/SIMKIT_HOME/libsimkit_spectre_3.0.so”

Then synopsis and all documentation related to parameters will come.

Here I didn’t get any msg regarding to the version of simkit..
So what is problem? What u suggest…

Title: Re: PSP MOSFET
Post by nikunj on Aug 22^nd, 2006, 11:42pm

hey thanks Geoffrey_Coram,
the problem is solved.. i was not including model code in netlist thats why it gave error
say for mos1100
#ahdl_include "mos11010.va"

is it possible to include 'c' code instead of verilog code in spectre netlist?

Looking forward to hearing from you...

Title: Re: PSP MOSFET
Post by Geoffrey_Coram on Aug 28^th, 2006, 6:49am

Sorry for the delay in responding.

Please double-check your instance line and your model card. You shouldn't need to #ahdl_include the Verilog-A source code.

Note you need a model, you can't do this:

m1 (d g s b) psp1010 type=n ...

you have to do something like

m1 (d g s b) my_nmos l=1u w=1u
model my_nmos psp1010 type=n ...

Title: Re: PSP MOSFET
Post by nikunj on Aug 29^th, 2006, 11:19am

For Id-Vd char. Of PSP, I will get some typical results ….it looks like Id-Vg char nt Id-Vd..

Here I m considering Vgs=0.7v Vsb=0.0v & I m varying Vds from 0 to 1 v and corresponding measuring drain current
Netlist ://
simulator lang=spectre

VGs (g 0) vsource type=dc dc=0.7
VDs (d 0) vsource type=dc

M1 ( d g 0 0 ) npsp w=10e-6 l=10e-6
model npsp psp1011 type=+1

dc1 dc dev=VD start=0 stop=1.0 lin=20

save VD:currents

when I will type spectre –f nutascii <filename> I will generate .raw data file also,

Title: // Spectre transfer char. circuit for psp1010
Date: 4:44:06 PM, Tue Aug 29, 2006
Plotname: DC Analysis `dc1': VDs:dc = (0 V -> 1 V)
Flags: real
No. Variables: 3
No. Points: 21
Variables: 0 dc V
1 VDs:n A plot=0 grid=0
2 VDs:p A plot=0 grid=0
Values:
0 0 0 0
1 0.05 5.0000062717331e-14 -5.0000062717331e-14
2 0.1 1.0000049019495e-13 -1.0000049019495e-13
3 0.15 1.50003455393141e-13 -1.50003455393141e-13
4 0.2 2.00024022330751e-13 -2.00024022330751e-13
5 0.25 2.50166540052325e-13 -2.50166540052325e-13
6 0.3 3.01152020978476e-13 -3.01152020978476e-13
7 0.35 3.5796590224208e-13 -3.5796590224208e-13
8 0.4 4.55057471006125e-13 -4.55057471006125e-13
9 0.45 8.30365196150691e-13 -8.30365196150691e-13
10 0.5 3.12644373269282e-12 -3.12644373269282e-12
11 0.55 1.86746180781748e-11 -1.86746180781748e-11
12 0.6 1.25567730061252e-10 -1.25567730061252e-9
13 0.65 8.6099747370198e-10 -8.6099747370198e-10
14 0.7 5.89935901098427e-09 -5.89935901098427e-09
15 0.75 3.97065948524656e-08 -3.97065948524656e-08
16 0.8 2.44109423589462e-07 -2.44109423589462e-07
17 0.85 1.12858076768285e-06 -1.12858076768285e-06
18 0.9 3.4322256987129e-06 -3.4322256987129e-06
19 0.95 7.14847075844511e-06 -7.14847075844511e-06
20 1 1.17955670880355e-05 -1.17955670880355e-05

i think Datas are comming wrong..Can u suggest me where I m doing wrong???

Really Thanks a lot for your time and consideration.

Title: Re: PSP MOSFET
Post by Geoffrey_Coram on Aug 30^th, 2006, 6:56am

nikunj wrote on Aug 29^th, 2006, 11:19am:

model npsp psp1011 type=+1

Try

model npsp psp1011 type=n

Title: Re: PSP MOSFET
Post by nikunj on Aug 30^th, 2006, 8:50am

If I am not mistaken then

psp1010 is for global geometry scaling
psp1011 is for global binning scaling
psp1010e is for local scalling

As I am working with parameter extraction of PSP mos model using Genetic Algorithms and (Variations). For my first stage, I am considering only global geometrical model because I am interested in data coming from global geometrical parameters only.

Can u suggest me is there problem in model itself or somewhere I am doing wrong??

Title: Re: PSP MOSFET
Post by nikunj on Sep 1^st, 2006, 12:41pm

thanks lot for helping....problem is solved...i got cottect dc I-V curve
bye

Title: Re: PSP MOSFET
Post by Geoffrey_Coram on Sep 4^th, 2006, 8:55am

Geoffrey_Coram wrote on Aug 30^th, 2006, 6:56am:

nikunj wrote on Aug 29^th, 2006, 11:19am:

model npsp psp1011 type=+1

Try

model npsp psp1011 type=n

For some reason, "type=+1" selects a PMOS device, I believe. I think the help for the device says to use n or p, so +1 is not really supported. This is unrelated to your choice of scaling rules.