The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Modeling >> Semiconductor Devices >> BSIM4 model Parameter llpe0
https://designers-guide.org/forum/YaBB.pl?num=1238930941

Message started by cmos.analogvala on Apr 5^th, 2009, 4:28am

Title: BSIM4 model Parameter llpe0
Post by cmos.analogvala on Apr 5^th, 2009, 4:28am

I have four corner moel file given by foundry for 90nm technology. This model file has BSIM4 models of MOSFETs. There is a parameter llpe0 in this model file. I want to know what is the physical significance of this parameter ? This parameter indeed affects dc response of a MOSFET. I could not find any parameter like llpe0 in BSIM4 manual .

Does anybody know what is physical significance of llpe0 ?

Title: Re: BSIM4 model Parameter llpe0
Post by Andrew Beckett on Apr 6^th, 2009, 2:10am

If using spectre, you can use "spectre -h bsim4" and then you'll see:

Quote:

536 llpe0=0.0 Length dependence of lpe0.

And also:

Quote:

73 lpe0=1.74e-7 m Lateral non-uniform doping at Vbs=0.

I think you should find lpe0 in the bsim4 documentation though?

Regards,

Andrew.

Title: Re: BSIM4 model Parameter llpe0
Post by cmos.analogvala on Apr 9^th, 2009, 2:50am

Yes lpe0 is there in BSIM4 model. If I want to perform montecarlo simulations and don't have foundary specific model file, should I vary llpe0 and lpe0 both ?

Title: Re: BSIM4 model Parameter llpe0
Post by Geoffrey_Coram on Apr 14^th, 2009, 4:56am

No, don't vary llpe0. llpe0 is a hack that, when used properly, can help connect lmin/lmax bins such that the final lpe0 value is continuous across bin boundaries.

However, I also don't think that most people vary lpe0, either -- only things like TOX, VFB, DL, DW.

Title: Re: BSIM4 model Parameter llpe0
Post by cmos.analogvala on Apr 23^rd, 2009, 8:06am

Thanks for the reply,
I have also seen people varying only tox, VFB, dL and dW only. However lpe0 being halo doping coefficient. Shouldn't we vary lpe0 for Montecarlo simulations ? Montecarlo simulation files supplied by foundary include variation in which all parameters ?

Thanks
CA

Title: Re: BSIM4 model Parameter llpe0
Post by Geoffrey_Coram on Apr 28^th, 2009, 4:58am

I think that VFB is an "average" flatband voltage, which includes the effect of pocket implant. I would expect lpe0 variation to be a second-order effect (how much does the channel doping differ from the pocket).

Probably most foundries use the basic four; I think I've heard of some that used VTH instead of VFB, but maybe there's a paper saying why this is wrong (correlation of TOX and VTH). Certainly, you could set up a MC sim varying each and every parameter in the BSIM4 model set, but you'd have to run lots more sims and accept the memory penalty (each device has a different value of *each* parameter).