Minutes, IBIS Futures Group Meeting 10/25

Attendees
---------------
Bob Ross
Arpad Muranyi
John Angulo
Stephen Peters


Minutes and AR review       5min   Peters
 -- review of previous minutes

  none

 -- AR list
  * John, upload minutes to web page.
  -- last weeks minutes are not on web page, neither are the ones owes. ETA,
this
     afternoon
  * Stephen, PDF current spec and get uploaded to web page [not done]
Friday.
Bob noted that footers/headers need to be updated.

  * Lynne, publish update (ibis-ml_05) [Done]
 
Opens for new Issues         5min   All

  what's allowed in a control source 

Old Business                     20min  Peters
  -- update on 'power supply rail selection' method (Mail from Al).
According to Al, we need to use a passed in generic (model parameter) to do
selection. Agreed to do so.

  A/R Stephen to update the example in section 4

  -- close on allowing currents in element value expressions. 
Stephen brought this up in relationship to adding diodes.  Arpad pointed out
that controled/vcontrolled  can be used to do diodes.  Bob favors adding in
current control to controlled elements.  Discussion on adding new
elements...no, just expand control nodes for current elements.  Discussion
on  how to notate current control... single I(instance_name.pin) or force an
I(node node) notation. Former  would be used if no node list, latter would
involve netlist changes to circuit.  More discussion, agree
instance_name.pin notation is preferred. Finally agreed to get rid of
control node listing on controlled  elements... just go with equations.
They will be able to reference any node voltage or branch current  in the
model.  Also note that + current goes into the element.

Current allowed in equations for passive elements.  However, still limited
to it's own node voltages or  currents.

Resistors still only have "R". 



IBIS-ML Spec Review           90min  All
  -- review updates to section 5

update footers/headers
uppercase 'M' in model_type
chapter 2  CAE -> EDA
uppercase tline in example (what should we do in titles?)
added table or something about unit multipliers
fix example that uses 'kelvin' scale on temperature
examples that use Time*1k.. make example more realistic (multiply by e9) and
also define someplace the simulator state variables.
fix diode equation
summarizing statement on what kind of expressions are allowed for
each group (type) of elements. (table is a good idea).
Add 'G' admittance to examples.
define units for frequency -- which is in Hertz (may not be an appropriate
example)
Chapter 4, remove 'then' from if-then-else
upper case IV and VT, make them I-V and V-T
include example that shows a table lookup multiplied by a scalar
don't use the phrase "I-V curve", it's a table.

  -- tline 'ref1' nodal format compatibility with N-port matrix
Issue is what this ref node really means.  Example: if one wants to describe
a three xmission line coupled system over a 'perfect' gnd plane then is it a
3-row matrix, with the ref node in the element connected to ground, or is it
a 4-row matrix.  What if it's not a perfect ground plane...want to include
return path effects.  In this case ref1 /= ref2 and it seems to warrant a
four row matrix.  Bob and Arpad to study the HSPICE manual some more to
understand the correct way to do this.

  -- outline of chapter 6 contents

Next Meeting Plans

We will meet again on Thursday Nov 1.