gpdk045 drain-source breakdown voltage

Hi Joris,

Why don't you think the parameters are bvd/bvs? (bvd defaults to the value of bvs). There is no parameter called wbvds in bsim4 as far as I can see (I have seen some references in LTSpice to a MESFET model and some hints that Keysight might have a parameter called this in their implementation - but it appears to be a parameter that triggers a warning in those models).

Andrew.

Hello Mr. Beckett,

thanks for the quick reply. 

1. In the gpdk045.scs-file, the bvs-voltage is set to 10; regardless of the type of transistor; so I thought this could not be the Vds-breakdown voltage I am looking for since this voltage is typically defined as being 2 or max. three times the nominal supply voltage. (also depending on the current through the transistor; because of breakdown but also e.g. max. power dissipation in the mosfet).  In this case, even in the case of a 1.8 V nmos, bvs is more than 5 times the nominal supply voltage.

2. So in fact, I am wondering what does the "1.8 V" in a "1.8V nmos" (g45n2svt) mean then?

3. What are e.g. the max. Vgs and Vds allowed for a 1.8 V nmos in the 45 nm GPDK? 

4. What is the max. power that can be dissipated in this type of nmos (is this a function of W and L?) 
Will the simulator give warnings if these limits are exceeded?

5. I found the  "wbvds" parameter in the BSIM4-model card for in ADS. Apparently, Cadence specifies some parameters which are not defined in the ADS BSIM4-model card, and vice versa as well? I was surprised by this, since I assumed so far that the BSIM4-model was completely standardised?

Excuses for my many questions,

Thanks,

Greetings,

Joris

Joris,

Answering most of these questions is a bit difficult, given that gpdk045 is a completely fictional process. So the models are made up. I'm sure that little thought went into setting bvs - that is also the default value for bsim4 anyway.

The voltage type of a mosfet in the gpdk is typical for that size of geometry - usually the impants and oxide thicknesses may be different to allow them to better behave (or survive) particular voltage ranges. How well these have been characterized in gpdk045 is still up for question though, since there was obviously nothing to physically characterize!

The model should indeed be pretty standard. There are a core set of standard parameters, and a few have been added in our implementation either because of specific foundry requests, compatibility with other models, and so on. I can't comment on what Keysight do with ADS...

Kind Regards,

Andrew.

Hello Mr. Beckett,

thanks for the quick answer. I was suspecting that this was the origin of the problem; but thanks a lot for providing a definitive answer. 

I actually have another question concerning the high-frequency analog behavior of the mosfets

but since this is not related to the breakdown voltage I made another thread for it.

Kind regards,

Joris