Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

hspice model problem

Status
Not open for further replies.

cyrax747

Full Member level 3
Full Member level 3
Joined
Nov 8, 2012
Messages
167
Helped
13
Reputation
26
Reaction score
11
Trophy points
1,298
Location
Bangalore
Visit site
Activity points
2,494
Hi Friends,

I am using hspice tool with bsimv3 level 49 parameters.my doubt is "Is there any restriction on specifying the channel length i meal l parameter because by using this model file i am unable to go less than 65nm.My job is getting aborted.Please clarify
 

There may be a restriction by the overlying model files from the fab.
 

Maybe its like this... if the SPICE model is of a 65 nm process, then it represents the device details of a MOS transistor which can be fabricated aka manufactured with minimum dimensions of not less than 65 nm. Hence maybe the job aborting issue. Other than that it will be better if you can share some details viz the technology node being used at your end and the version of SPICE engine (for eg. a old spice engine, viz. lets say, something made in before technology nodes below 65 nm came to existence may give error etc. )

Hobbyiclearner
 
I am using hspiceui for windows level 49 model file bsim3v3.technology node used in my spice file is 130n.
 

Hi,

Another doubt is at present for tech nodes < 130nm bsim4 is being used and also bsim6.How to get this model files?? In the website they have given c code !! I dont understand.
 

I am using hspiceui for windows level 49 model file bsim3v3.technology node used in my spice file is 130n.

Pls confirm the release dates of your SPICE engine and the technology node first...........

- - - updated - - -

Hi,

Another doubt is at present for tech nodes < 130nm bsim4 is being used and also bsim6.How to get this model files?? In the website they have given c code !! I dont understand.

Provide additional details of the site pls....Usually the files maybe propitiatory and require permissions for downloading.........
 
Last edited by a moderator:

HSPICE -- A-2008.03 32-BIT (Feb 26 2008) tool
July 29, 2005 Release of BSIM3v3.3.0 Model.

site used for downloading...h**p://www-device.eecs.berkeley.edu/bsim/?page=BSIM4
please provide how to get bsim4 model file??
 

HSPICE -- A-2008.03 32-BIT (Feb 26 2008) tool
July 29, 2005 Release of BSIM3v3.3.0 Model.

site used for downloading...h**p://www-device.eecs.berkeley.edu/bsim/?page=BSIM4
please provide how to get bsim4 model file??

OK now you have to check if HSPICE 2008.03 version actually supports dimensions less thn 65 nm (try contacting synopsys support...)


for BSIM4 lokk here...

https://www-device.eecs.berkeley.edu/bsim/?page=BSIM4_LR

hobbyiclearner
 

Dude

I have visited same website and downloaded the bsim4 file,but it contains everything except parameters.It contains c code manual i mean explanation of parameters etc.
 


From PTM, i am using already but they are predictive models,I think BSIM versions will be good.
 

so from ptm i am using level 54 model bsim4 version.but are they developed by berkeley guys?? they are preductive assumptions for future technology,i mean not like researched by berkeley guys and official ones.
 

Level ≥ 49 are BSIM Models, so why not use the PTM models?

Hi erikl,

Just wanted confirmation for my knowledge... are model files greter than level 49 PTM models or BSIM models.... your last pot just confused me.. pls give some details...

Thanks,
hobbyiclearner
 
Last edited:

so from ptm i am using level 54 model bsim4 version.but are they developed by berkeley guys?? they are preductive assumptions for future technology,i mean not like researched by berkeley guys and official ones.

cyrax, you have to differentiate between the BSIM models per se, which are developed by the Berkeley guys, and which describe the methods and equations how the analysis tools (simulators) have to calculate the device properties from the parameters which are given in the model files - and the BSIM model parameter files themselves, whose parameters could either be estimated - as in the PTM models - or be extracted from silicon (by statistical measurement methods) and calculated with the help of special extraction software and rather complex computation effort, by the chips' producing (silicon) fabs and foundries.

I.e. a BSIM model without parameters can't be helpful for device analysis, you need a BSIM parameter file for that. If the Berkeley guys offer such a BSIM parameter file, these parameters may - or may not - stem from true silicon extracted measurements - and you wouldn't know from which (company's) process; they ain't necessarily better than the estimated parameter values from (e.g.) PTM.

BSIM model parameter files from both Berkeley or PTM are good enough for academic/educational purpose usage; if you need to design for a special process from a real semiconductor fab/foundry, you will need their individual PDK and simulator model files, which contain the actual parameters of this very process, and you'll probably have to pay for that, and for sure have to sign an NDA.

- - - Updated - - -

... are model files greter than level 49 PTM models or BSIM models...

First see my answer to cyrax, pls., for differentiation between BSIM models and BSIM parameter files.

BSIM models are analysis models, e.g. called BSIM3, BSIM4, BSIM6, BSIMSOI.

BSIM (model) parameter files are those used for one (or more) of these a.m. BSIM models, with level=49 and above (e.g. 54 mainly for BSIM4).
 

OK... so if I have understood correctly... model files are a set of equations and parameter files are a list of constants etc. of these equations. For eg: in level 1
Id= un*Cox*W/2L(Vgs -Vt)*(Vgs -Vt) is an equation of the model file and constants like vt, un, Cox etc are available in parameter file.

Then there are various levels of analysis depending on complexity and thereby BSIM 1, 2, 3 etc., each having its own sets of model files and parameter files.

For basic simulation simply the parameter files of these BSIM model files etc are enough, whereas for making a real & complex IC one needs both the model files and the parameter files (included with the pdk).

Again, one more query pls. : if the model files are required ie. the equations are required for such cases, are they in putted into the spice simulator manually at the designers end or if simply one has to check if the simulator is compatible with that particular model file.

Thanks,
Hobbyiclearner
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top