RE: [SI-LIST] : Re: Controlling switching noise in full-swing CMOS buffers (was Despite output resistor to reduce overshoot and undershoot)

About this list Date view Thread view Subject view Author view

From: Zabinski, Patrick J. (zabinski.patrick@mayo.edu)
Date: Wed Dec 06 2000 - 10:29:40 PST


Linear output impedance does not improve noise (at that I am aware
of), but it does improve the shape of the waveform.

Most full-swing CMOS buffers rely on a source-series termination
scheme where the matching impedance is at the driver end and
the receiving end is high impedance. In such schemes, a half-amplitude
wave will propogate down the line, double in amplitude when it gets
to the high-impedance receiver, and reflect back to the driver.
The assumption is that the driver will then absorb the reflection,
because the driver's output impedance matches that of the line.

If your driver's impedance varies with amplitude (as most buffers
do), then some of the reflected signal will get reflected from
the driver, and bounce back to the receiver (again). Generally,
the re-reflected signal amplitude is low, so it's not much of
a problem. However, if your driver is far from a good match,
your driver is improperly designed, or your data rate is such
that the re-reflected waves start to constructively add upon
each other, life can get interesting.

Anyway, in the grand scheme of things, the other issues/tricks
I mentioned are likely to be of more significance than this issue.

Pat

> I do not quite understand why linear output impedance can reduce noise
> or improve
> signal integrity ? You do not need to share the circuits anyway :)
> Are u talking about the dynamic output control implemented by
> some large
> companies ?
> If so, actually I also thought of this but perhaps as u said, it is a
> kind of proprietary, so I have
> to figure out the other way round.

**** 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 : Tue May 08 2001 - 14:30:22 PDT