RE: [SI-LIST] : Far end crosstalk

About this list Date view Thread view Subject view Author view

From: Larry Miller ([email protected])
Date: Fri Oct 13 2000 - 13:58:57 PDT


There is another kind of far end crosstalk that is often more important.
IEEE 802 refers to it as ELFEXT (Equal level Far End Crosstalk).

Since 100 Mb/sec (100BASE-TX) and Gigabit Copper (1000BASE-T) systems use an
adaptive equalizer at the receiving end of a transmission link, what you are
interested in is how much crosstalk you get when taking into account the
frequency response (roff-off with increasing frequency) of the cable.

That is to say, you are interested in the ratio of coupled signal to desired
signal with the treble control on the equalizer turned up. Since the TIA and
IEC cable and cabling plant standards specify this link cable/plant response
it is possible to specify how much of this kind of cross-coupling you can
stand.

This is really much more indicative of what the receiver will see and
respond to.

Larry Miller

> -----Original Message-----
> From: Issa, Elie [SMTP:[email protected]]
> Sent: Friday, October 13, 2000 9:36 AM
> To: 'Peterson, James F (FL51)'; [email protected]
> Subject: RE: [SI-LIST] : Far end crosstalk
>
> Hello:
> <I guess my question would be, why does it have to be a ratio of
> anything?>
> Because CrossTalk is characterized as a Gain for the quite line. So it
> has
> to be a ratio.
> Since it is a voltage gain, it should be compared to the input voltage of
> the aggressor
> The voltage at the far end of the aggressor would include losses
> unrelated
> to cross talk which you are not interested in this discussion
> I believe the definition of the far end cross talk would be, FEXT =
> Voltage
> at the far end of the quite line / Vin of the aggressor line.
>
> You need to note that some time when you are trying access your device,
> you
> may need some fixturing. such fixturing would degrade & slow your rise
> time
>
> and attenuate the input voltage from the source.
>
> You need to characterize those two quantities when you are reporting your
> CrossTalk or shopping for a product with good crosstalk performance. The
> attenuated amplitude from the output of the fixture would need to be in
> calculation of Crosstalk as the input aggressor voltage and not the source
> voltage of the instrument . Also you would be interested to know the Rise
> Time(since cross talk is dependent on it). Here also should be the
> measured
> Rise Time from the output of the fixture and not the RiseTime of the
> instrument. Not to consider those factors would make CrossTalk (far end
> or
> Near end) look better.
>
> Regards
>
> Elie Michael Issa
> Signal Integrity Group
> Corporate Design Automation
> ISSD, Enterprise Computing Group
> COMPAQ Computer Corporation, Houston
> Phone: 281-518-5492 / Fax: 281-518-0652
> [email protected]
>
>
>
> -----Original Message-----
> From: Peterson, James F (FL51) [mailto:[email protected]]
> Sent: Friday, October 13, 2000 10:00 AM
> To: [email protected]
> Subject: RE: [SI-LIST] : Far end crosstalk
>
>
> Hello Ohio,
>
> I guess my question would be, why does it have to be a ratio of anything?
> Why can't it just be the amount of noise visible at the far end of the
> victim's transmission line?
>
> I guess if it is a rating of a t-line's capacity for crosstalk then you
> might go with a ratio (for x amount of voltage injected at the aggressor's
> t-line you will get x mV of coupled noise at the victim's far end). If
> this
> is the case I would think you would use the aggressor's initial signal
> level.
>
> Jim
>
> -----Original Message-----
> From: JOHN SAWDY [mailto:[email protected]]
> Sent: Wednesday, October 11, 2000 12:48 PM
> To: [email protected]
> Subject: [SI-LIST] : Far end crosstalk
>
>
>
> Greetings from Ohio.
> What is the correct definition for far end crosstalk for a transmission
> line
> (cable): is it the ratio of the voltage measured at the far end of a quiet
> line vs. the input voltage on the active line or vs. the output (far end)
> voltage on the active line, which is an attenuated version of the input
> voltage.
>
> FEXT = Vout quiet/Vin active or Vout quiet/Vout active?
>
> John F. Sawdy, Senior Applications Engineer
> Meritec, a division of Associated Enterprises
> voice mail: 440-354-3148 x 267
> fax: 440 -354-0687
> mailto:[email protected] <mailto:[email protected]>
>
>
> **** To unsubscribe from si-list or si-list-digest: send e-mail to
> [email protected]. 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
> ****
>

**** To unsubscribe from si-list or si-list-digest: send e-mail to
[email protected]. 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 - 14:29:46 PDT