I also agree with Pat.
Your model may really be dependent on your design and simulation
methodology. This is why it's hard for socket vendors to give you the data
you want.
Things to consider are return path and even/odd propagation modes Self
inductance may be OK for low frequency but my experience suggests that you
can get a good deal of ground/power bounce and SSO pushout from a socket
that would not be modeled by simple self inductance. I also have found these
effects to vary greatly from socket to socket.
Richard Mellitz,
Intel
-----Original Message-----
From: Pat Zabinski [mailto:[email protected]]
Sent: Thursday, September 02, 1999 7:31 AM
To: [email protected]
Subject: Re: [SI-LIST] : BGA socket modeling
Matt,
I went through a similar process about six months ago, and
parasitics
varied greatly between the various vendors. For example, we
measured
one socket with self inductance of 1 nH and a different
socket at 7 nH.
Similar variations were found in C and R.
In addition, when we asked most (not all) vendors for
models, they
rattled off numbers from a generic sheet that covered all
sockets,
regardless of pin-pitch. I don't recall finding any socket
vendor
that had mutual parasitics either.
In other words, I agree with the others in trying to develop
your
own models. There are several methods; you'll have to
determine
the best approach based on your resources.
Pat
> Hi,
>
> anyone have any approximate figures for BGA test socket
LRC values
> or how to model it.
>
> I am trying to get info from the socket vendor, but all I
seem
> to be able to get is contact resistance and mechanical
info.
>
> matt
**** To unsubscribe from si-list: send e-mail to
[email protected]. In the BODY of message put: UNSUBSCRIBE
si-list, for more help, put HELP. si-list archives are accessible at
http://www.qsl.net/wb6tpu/si-list ****
**** To unsubscribe from si-list: send e-mail to [email protected]. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****