RE: [SI-LIST] : Combined single-ended/differential termination pa ckage?

About this list Date view Thread view Subject view Author view

From: S. Weir (weirsp@atdial.net)
Date: Sat Mar 03 2001 - 14:06:55 PST


Ron, Rich,

I think the risk here is DC imbalance building up too quickly. In order
for a very small cap to work, the code pattern has to be balanced well
above the pole. @ 50pF / 50 ohms this is only 2.5nS. Even @ OC-192 rates
this is not going to be enough. The imbalance will be the result of any
integrated asymmetry in the wave forms of the individual drivers. This
only nulls-out when we integrate across a period where the code sequence is
DC balanced. For most codes out there, a fairly low frequency pole is
needed to avoid jitter from the symbol pattern.

Regards,

Steve.
At 12:51 PM 3/3/01 -0800, you wrote:
>Hi Rich
>
>You make a good point. I like the idea of termination of only the high
>frequency
>signals without providing a path for the lower frequency garbage from inverter
> power supplys and such.
>
>Thanks for the suggestion.
>
>ron
>>-----Original Message-----
>>From: RMELLISON@aol.com [mailto:RMELLISON@aol.com]
>>Sent: Saturday, March 03, 2001 12:41 PM
>>To: rmiller@brocade.com; ldmiller@rhapsodynetworks.com;
>>ericg@redback.com; si-list@silab.eng.sun.com
>>Subject: Re: [SI-LIST] : Combined single-ended/differential termination
>>pa ckage?
>>
>>Ron, I believe I've read that common mode noise on the input of PECL devices
>>(and maybe some others) can produce jitter on the output. I've also noted
>>that when people call out a capacitor for providing a common mode
>>termination
>>(when the differential termination is two seriesed 50 ohm resistors,
>>capacitor to ground at the center tap) the value is usually .01ufd. It
>>seems
>>like the high frequency components are the primary offenders, due to reduced
>>common mode rejection of the input stage of the receiver, as well as
>>unbalances in the two sides of the driver. I believe it would be a
>>reasonable compromise to use a 30-70pf capacitor in order to provide common
>>mode termination just at the higher end of the spectrum. What do you think?
>>Rich Ellison

**** 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 Jun 21 2001 - 10:11:07 PDT