Istvan,
- There are mating connector halves at both ends, and shall need to be
connected/unconnected multiple times.
- No vibrations, no moisture environment, no temperature extremes.
- One piece of connector [*].
- Return pins are considered, at least one for each diff. pair.
Thank you very much for your suggestions.
Best regards,
Sebastian.
[*] In the near future, this need for "high-speed on old/unspecified
connectors" could be extended to similar facilities with different
through-wall connectors (and maybe multiple connectors in a given
facility). For this, we could even afford some kind of dynamic channel
characterization by the system itself.
Istvan Novak wrote:
Sebastian,
Are there mating connector halves at one or both ends
that would need to be connected/unconnected multiple
times during the life of usage or can you assume it is
installed/connected once and stays that way? Do you
need to do this on one piece of connector or a number
of nominally same connectors? Is this a harsh environment
with a lot of moisture and/or potentially corrosive particles
in the air, do you need to consider vibrations in the
installation?
I would start designing a breakout test board with a decent pinout
with as many return pins (you need that even for differential
signaling in an old open-frame connector) assuming
optimistically a few GHz bandwidth, using a four port
VNA with up to maybe ten GHz bandwidth. You would
need to characterize losses, reflection and crosstalk for
the full scattering matrix. You may skip some matrix elements
where speculatively you can convince yourself that one
neighboring pair will behave very similarly. You would
need to check for repeatability and consistency over
connecting/disconnecting and unit-to-unit if those cases also apply.
Regards,
Istvan Novak
Oracle
On 4/27/2016 7:37 AM, Sebastian Garcia wrote:
Hi Istvan, thank you for the reply. BTW, thank you for your valuable
knowledge shared over many years on this channel.
OK, but what testing setup/procedures would you recommend?
Regarding crosstalk, it will be unavoidable, as will have ~20 pairs
communicating (~10 in each direction) through the connector interface.
The goal is to get the best of a given connector, having the freedom to
design all around: cabling, PCBs, interfacing standard, signaling, etc.
E.g., using a TDR setup the characteristic impedance of the connector
could be identified, then try to match the rest of the transmission line
to this impedance. That is, not doing a classic 50 Ohm (100 Ohm diff.)
layout/cabling.
Best regards,
Sebastian.
Istvan Novak wrote:
Sebastian,
If you can fully serialize the data so it is just one differential
signal, in that case loss
(and also reflection) might be enough. If you need to pass multiple
signals, crosstalk
would be equally important, though then likely it is very much
pin-assignment dependent,
meaning you can make it better (or worse) with how you assign the pins.
Regards,
Istvan Novak
Oracle
On 4/26/2016 7:37 PM, Sebastian Garcia wrote:
Hi,------------------------------------------------------------------
I need to pass some digital (std. binary) differential signals through
an
unknown connector installed in a wall dividing two rooms.
Old connector, no technical data available.
The communicating circuits' PCBs and cabling (between each PCB and the
connector) at both sides of the connector are to be designed.
The signals will need to be serialized, and would like to obtain an
estimate of what maximum speed can be robustly achieved with the given
connector.
For testing purposes, both sides of the connector can be accessed at
the
same time by the same instrument (with typical instrumentation probe
lengths).
So, I would like to hear advice from the list members with experience
in
connector testing, for characterizing the involved losses.
Best regards,
Sebastian.
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List forum is accessible at:
http://tech.groups.yahoo.com/group/si-list
List archives are viewable at:
//www.freelists.org/archives/si-list
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
