Re: [SI-LIST] : Fast edge termination choice

About this list Date view Thread view Subject view Author view

From: D. C. Sessions (dc.sessions@vlsi.com)
Date: Fri Feb 25 2000 - 14:32:21 PST


Shayle Hirschman wrote:
>
> SI experts:
>
> I am interested in making a decision between series source termination and
> parallel termination (probably something like 100 ohms to vcc and 100 ohms
> to gnd) at the destination of a point to point route.
>
> I want to use the fast slew rate option of the FPGA as the clock is 150 MHz
> and there isn't much timing budget from chip to chip.
>
> Does series source termination defeat the purpose of selecting the fast
> slew rate option? Will it slow the edge transition time down, or will the
> signal still transition and travel as fast as if there was no termination,
> but simply be divided (thereby depending on the reflection at the
> destination)?
>
> I suspect the latter, except that I've heard people say that series source
> termination reduces EMI by slowing the edges down. Could it be that it
> reduces EMI, not by slowing the edges down, but rather just simply because
> it terminates? And that, therefore, as far as EMI goes, parallel
> termination would have a similar EMI result?
>
> If the edge rate does not diminish by using series source termination, then
> I would prefer that method since it uses only one resistor and has no DC
> bias current as does parallel termination.

Series termination reduces the magnitude of the *currents*
involved by nearly half, and reduces the time envelope in
the same way any effective termination does: by reducing
the Q of the network.

If you have short interconnects, you can actually get pretty
reasonable signal quality on a point-to-point net (you were
talking about a point-to-point interconnect, right? The
series term implies that) by putting one resistor in the
middle of the run. It sounds terrible, but manages to
work tolerably well by providing a loss mechanism and by
phase-splitting the reflections.

-- 
D. C. Sessions
dc.sessions@vlsi.com

**** To unsubscribe from si-list or si-list-digest: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu ****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Thu Apr 20 2000 - 11:35:11 PDT