RE: IBIS MINUTES 12/01/95

Thank you for your observation and question.

My concept of "unmated" is that the model should be self-sufficient (i.e.,
work properly without extra 'patches and band aids' and have connections at
one side only. So it would look like this ...

     Board1--> conn_model (unmated)

There would be no additional load required outside of the model.

For SPICE type simulators there may have to be some standard DC path, or
'assumed load' to enable an unmated connector model to function properly.
 That would not likely be required by most models, however, because there
would already be the necessary DC paths to satisfy SPICE within the model
itself or the bus. A standard load may be something that should be defined
for stand-off type stacking connectors that are long relative to the models'
minimum applicable edge rate.

At this point I have only proposed, for consideration, the concept of both
mated and unmated connector models as opposed to a mated model with an
interface connection such as is proposed in BIRD 31. Many details need to
be worked out until that is finally accomplished if the concept is
acceptable to everyone.

Hope this is helpful.

Regards,

 Hank
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 ----------
From: cpk
To: amp03847
Subject: RE: IBIS MINUTES 12/01/95
Date: Wednesday, December 06, 1995 9:51AM

Clarification:

Does unmated mean that you assume some kind of load at the other end

Board1--> conn_model --> assumed load (unmated)
Board1 -> conn_model->board2 - full model

Thanx
 - kumar

>
>
> Bob,
>
> First I want to congratulate you on spelling my name correctly. That
> almost never happens. Also, you have done an excellent job of capturing
all
> the discussions that took place in the last meeting.
>
> There are two points that I think were discussed that I did not see in the

> minutes. Perhaps they were too small to include but I would like to
clarify
> and confirm them.
>
> First, I thought someone mentioned the desire to discuss connector modes
at
> the Face-to-Face Meeting in January. I wonder if anyone has any specific
> points they would like to discuss. Since I am considering trying to
attend,
> this could significantly influence my decision. (Cross country trips for
a
> one day meeting are not popular with me or my management.)
>
> Second, when BIRD 31 was discussed, I recognized the need for many users
to
> have unmated models to simulate "not-fully-populated busses". They would
> not be used for such analyses as cables or I/O connectors when there is a
> complete interconnection. While we do not routinely supply unmated models

> now, there is no reason that they could not be provided. They would even
be
> simpler than "part of the mated model" which the BIRD suggests. What I
> recommend is that two types of connector models be defined. First, mated
> models for most interconnection applications. Second, unmated models that

> have no through connection for situations that demand simulations of
> "not-fully-populated busses".
>
> I hope these points are helpful to the forum. I enjoyed participating
with
> you all and look forward to future work together.
>
> Regards,
>
> Hank Herrmann
> Technical Staff Member
> AMP Incorporated, Electro-Magnetic Technology
> M.S. 106-14
> P.O. Box 3608
> Harrisburg, PA 17105-3608
>
> Phone: 717-986-5534
> FAX: 717-986-5643
> INTERNET: hank.herrmann@amp.com
>
Received on Wed Dec 6 07:54:04 1995