From: Eric Goodill (firstname.lastname@example.org)
Date: Wed Dec 15 1999 - 17:06:49 PST
"D. C. Sessions" wrote:
> Eric Goodill wrote:
> > I want to use the typical LVDS termination where you have two 50-ohm resistors in series placed between the two lines of the LVDS diff. pair and also have a cap at the junction of the resistors to ground. I have a jillion LVDS pairs to terminate at a high-pin-count ASIC, and I'm looking to squeeze more terminations per square inch. (There's no hope of putting the terminations inside the ASIC at this point.) I'm hoping some company makes a single three-pin part the has those two resistors and a cap even if not specifically for termination use.
> > How about any other ideas for even higher density?
> Well, a newline every 60-70 mcolumns would help :-)
Sorry, set wrap to 72 chars.
> Seriously, you should be able to skip the cap and just use a 100-ohm resistor.
> Especially if you're running both ends of the line on a common ground.
> Alternately, use a pullup pack (several 50-ohm resistors to a common pin)
> and one cap. The application you're describing sounds like all of the pairs
> come over a common path, so they should have pretty well-matched common-mode
Thanks for all replies. I just happened to pick D.C.'s message to
actually reply to.
I forgot to mention a few key pieces of info. Speed is about 200 MHz.
Running card-to-card via backplane. Each card powered by separate DC/DC
convertors. Point-to-point links.
My concern is that I'll have some common-mode (single-ended) noise at
the receiver. If there's no single-ended-termination path, then that
noise may cause some timing jitter in the receiver. I'm a little
worried about my timing margin right now, so I'd like to start with the
idea os terminating the single-ended mode as well as the differential
mode. I guess the main sources of single-ended noise I can think of are
driver asymmetry and crosstalk. I've heard of mode conversion, but I'm
not sure exactly how it would operate in my case.
-- Eric Goodill Siara Systems, email@example.com
**** To unsubscribe from si-list: send e-mail to firstname.lastname@example.org. 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 ****
This archive was generated by hypermail 2b29 : Tue Feb 29 2000 - 11:39:07 PST