[SI-LIST] Re: IBIS2XTK v. SPI2MOD

Hi Adam...

First of all, if an ibis model is available it's easier to generate xtk files 
from IBIS2XTK,
since you don't have to worry about producing spice output files to use with
SPI2MOD. 

As far as accuracy, from my experience I found them to be very close.
Both IBIS2XTK and SPI2MOD do not like non-monotonicity in the VT curves 
and will try to straighten this out. SPI2MOD doesn't like non-monotinicity in 
the VI curve, whereas IBIS2XTK doesn't like it in the following sums: 
HIGH+SHUNT and LOW+SHUNT. 

That's okay, because in one of my previous
simulations I compared the original spice simulation (which has non-
monotonicity) with xtk simulation (monotonicity is straightened) and 
the results were very close. That means in real life monotonicity doesn't 
affect
you very much because it will be reduced by introducing package capacitance
and all the inductances on the board.

Another thing that the XTK manual mentiones is that SPI2MOD is designed
to be used with enhanced models excluding the multi-VI devices. So if your
driver has two stages of pull-ups and pull-downs, or a more complex design 
then
SPI2MOD as it is doesn't produce accurate models. You have to run separate 
SPICE simulation for individual stages and becomes complicated, and I don't 
know
how to do that.

In short, if you have ibis models and you trust them, then use them.
Also, I don't think there is any special treatment for differential models.

If somone or someones disagree with me that's okay, I'm still learning and 
would appreciate any correction or addition to my assesment.

Thanks,.

Tariq Abou-Jeyab


------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List archives are viewable at:     http://www.freelists.org/archives/si-list 
Old list archives are viewable at: http://www.qsl.net/wb6tpu

Other related posts: