What means pin t of the Verilog-A code of BSIM-BULK model?

It's a thermal node (which is optional normally). By the way, you don't need to use the Verilog-A model if using spectre as bsimbulk is built in to the simulator (see "spectre -h bsimbulk"; also covered in the Spectre Circuit Simulator Components and Device Models Reference manual in the bsim6 chapter - bsim6 is an alias for bsimbulk).

Andrew

Thank you for your detailed reply.

By the way, I want to simulate my MOSFET and circuits using Virtuoso, not spectre.

How can I use bsimbulk model which is built in spectre at Virtuoso? (There is no bsimbulk in any library I have.)

Thank you

yysunj

yysunj said:

By the way, I want to simulate my MOSFET and circuits using Virtuoso, not spectre

Given that Virtuoso is not a simulator, this is going to be a bit of a stretch. I'm assuming you are just saying that you want to start the simulation from Virtuoso - the simulator itself is (probably) spectre, but maybe you're using something else?

Anyway, the first thing to determine is whether this is for a real technology or something made up just for experimentation purposes? If it's from a foundry, the chances are they have a PDK (Process Design Kit) which has a symbol library and associated device models (perhaps using bsimbulk) that will (or can) be referenced from the Analog Design Environment and then passed to the simulator (e.g. spectre) to tell it the relevant parameters. Just because the underlying model (bsimbulk) exists either natively in the simulator or as VerilogA doesn't mean that's sufficient - that model normally needs parameters to provide the process-specific coefficients to the device modelling equations used in the device model.

So, some questions:

1. Is this a real technology?
2. If it's real, can you ask the foundry for the PDK and models?
3. If it's not real, how are you going to come up with a suitable set of model parameters for bsimbulk? Maybe you have a representative model definition already?

If you just have a model file, you can use the process described in this other thread. Oh, hang on, that was an answer I gave to another question from you! Did you not read what I said there? This is the same scenario as for the TFT model - yes, it's a different device, but the principles as to how you'd reference the model are the same.

Andrew

1. This is not a real technology. It is made up just for experimentation purposes.

2. It is not a real.

3. Yes, I can get parameters from my Fab.

Thanks

yysunj

1. This is not a real technology. It is made up just for experimentation purposes.

2. It is not a real.

3. Yes, I can get parameters from my Fab.

Thanks

yysunj