problem of simulation of VCO in SPICE

[cici]

Hi,

I am doing simulation in SPICE. It is a ring VCO. WHen I just did
..TRAN analysis, there is "parameter out of range or the wrong type" on
screen. What is the problem? I am using TSMC25 models. It works well
in my current mirror.

Thank you!

 

[Jim Thompson]

Hi,

I am doing simulation in SPICE. It is a ring VCO. WHen I just did
.TRAN analysis, there is "parameter out of range or the wrong type" on
screen. What is the problem? I am using TSMC25 models. It works well
in my current mirror.

Thank you!

"SPICE" what?

...Jim Thompson

 

[cici]

"SPICE" what?

...Jim Thompson

It is SPICE3. Thanks.

 

[Jim Thompson]

"SPICE" what?

...Jim Thompson

[snip]
It is SPICE3. Thanks.

I don't know Spice3, but I'd guess it can't handle higher level
models. Isn't TSMC_0p25 in at least BSIM3.3.2 format? My
five-year-old ones are and I think recent ones are in EKV format.

...Jim Thompson

 

[Mike Engelhardt]

cici, Jim:

I don't know Spice3, but I'd guess it can't handle
higher level models. Isn't TSMC_0p25 in at least
BSIM3.3.2 format?

TMSC models are hspice. hspice added a number of
parameters to the Berkeley BSIM device, mostly dealing
with how to compute dimensions. If you fully specify
all dimensions(with areas and perimeters) you can
probably ignore the warnings. If your spice3 treats
them as errors, just comment them out of the model.
You can verify with another simulator when you get
a chance.

My five-year-old ones are and I think recent
ones are in EKV format.

I've never seen EKV models out of TSMC.

--Mike

 

[Jim Thompson]

cici, Jim:


TMSC models are hspice. hspice added a number of
parameters to the Berkeley BSIM device, mostly dealing
with how to compute dimensions. If you fully specify
all dimensions(with areas and perimeters) you can
probably ignore the warnings. If your spice3 treats
them as errors, just comment them out of the model.
You can verify with another simulator when you get
a chance.


I've never seen EKV models out of TSMC.

--Mike

I have a TSMC EKV model set here dated 4/20/2004.

...Jim Thompson

 

[Mike Engelhardt]

Jim,

[...]
My five-year-old ones are and I think recent
ones are in EKV format.

I've never seen EKV models out of TSMC.

I have a TSMC EKV model set here dated 4/20/2004.

Wow, thanks, that's amazing. What level number do
they use? I had been getting the impression that EKV's
popularity was going south. I suspect that's not part of
their mainstream model release, but a special library
with limited distribution.

--Mike

 

[Jim Thompson]

Jim,

[...]
My five-year-old ones are and I think recent
ones are in EKV format.

I've never seen EKV models out of TSMC.

I have a TSMC EKV model set here dated 4/20/2004.

Wow, thanks, that's amazing. What level number do
they use? I had been getting the impression that EKV's
popularity was going south.

Level=55. PSpice threw a curve ball and designated it as Level=5,
which was an old model level designator that was previously abandoned.

I suspect that's not part of
their mainstream model release, but a special library
with limited distribution.

--Mike

I suspect you are right... it came from a client of mine with a
special relationship with TSMC.

...Jim Thompson

 

[cici]

Jim,

[...]
My five-year-old ones are and I think recent
ones are in EKV format.

I've never seen EKV models out of TSMC.

I have a TSMC EKV model set here dated 4/20/2004.

Wow, thanks, that's amazing. What level number do
they use? I had been getting the impression that EKV's
popularity was going south.

Level=55. PSpice threw a curve ball and designated it as Level=5,
which was an old model level designator that was previously abandoned.

I suspect that's not part of
their mainstream model release, but a special library
with limited distribution.

--Mike

I suspect you are right... it came from a client of mine with a
special relationship with TSMC.

...Jim Thompson

Thanks for your information.

My tsmc25 models is level 11. version3.1.
When I use this model to simulate a simple current mirror in PSICE3,
it doesn't have such
problem. Only when I simulate VCO,this problem appears. I checked the
topology, it is correct.