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.

C or Matlab models for transistors

Status
Not open for further replies.

ben_IC

Newbie level 3
Newbie level 3
Joined
Feb 7, 2014
Messages
3
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Visit site
Activity points
33
Hello,

I have designed a circuit in Cadence Orcad. I need to have a model of this circuit in Matlab or C for further analysis and modeling of the entire system.

I have a good model for the entire system except two transistors that operate as a common source amplifier. There is a P FET that operates as a load, and an n FET that is driven by a signal. I have used MbreakN-X and MbreakP-X and right now I use a Level 3 spice model for them.

How can I get a very good model for this two transistors? What Ids/Vgs equations should I use? I know that basic Ids/Vgs equations in introductory VLSI books are not precise enough. Is there any C++ or Matlab m files for transistor modeling for this level?

I am very flexible on what spice model, which technology, or which level I use. The only important thing is that I need a spice model WITH readily available and easy to use equations for it so that I can simply put in C or Matlab and they give me similiar results.


Thanks in advance
B
 

Sir those devices are modeled best in BSIM4 or the other high level models. So instead of creating your own model for calculations, you can just gather the data from those models, and just use regression to connect the data points. This would be much more simpler than writing a model from scratch, because the models get very complicated, very easily.
 

I have designed a circuit in Cadence Orcad. I need to have a model of this circuit in Matlab or C for further analysis and modeling of the entire system.

Why not analyze the design in Orcad? After all that is one of its intend purposes.

BigDog
 

Sir those devices are modeled best in BSIM4 or the other high level models. So instead of creating your own model for calculations, you can just gather the data from those models, and just use regression to connect the data points. This would be much more simpler than writing a model from scratch, because the models get very complicated, very easily.

So what you say is that I use BSIM4 spice model in Orcad, then plot Ids vs Vgs for a transistor. And save data point and apply a curve-fitting/regression technique to obtain a model in Matlab or C++? Right?


Tesekkurler Arkadas!

P.S: I have been to Istanbul a few times and I LOVED it!

- - - Updated - - -

Why not analyze the design in Orcad? After all that is one of its intend purposes.

BigDog

My analysis includes Monte Carlo simulation, exhaustive search of system parameters, heuristic search of system parameters, etc. It is very very very complex and it is not something that I can easily implement in Orcad. After I get the optimal system parameters in C/Matlab simulations, then I will simply write them into the circuit that I have in Orcad.
 
Last edited:

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top