[SI-LIST] Loss requirements in Infiniband spec. (si-list Digest V2 #70)
- From: "Jay Diepenbrock" <jaydiep@xxxxxxxxxx>
- To: si-list@xxxxxxxxxxxxx
- Date: Mon, 11 Mar 2002 09:07:39 -0500
Matt Ruston said:
>>All:
>>Hi. I've been looking through the Infiniband 1.0a spec and need some
>>clarification.
I think it might have been appropriate to address your questions to the
InfiniBand Trade Association in general, or the spec. owners ("architects")
in particular. Perhaps you didn't know how to find them, but next time you
might try www.infinibandta.org.
In the meantime, perhaps I can help clear things up.
>>1. Chapter 6, Table 17 specifies a LOSS of 2.5 dB for an adapter board,
15
>>dB of total loss, and (Chapter 7) 10 dB cable loss. Adapter board loss is
>>later defined in section 6.6.2.2 as "the maximum loss from the IC package
to
>>the board connector is 2.5 dB at 1250 MHz, including the connector."
>>Footnote 2 in Table 17 and Chapter 7 (Table 21) talk about the cable loss
>>(Vcout) as 10 dB "measured at board connector pins with equipment and
>>fixture contribution de-embedded." Section 6.6.2.3 says "including the
>>connector loss".
>>What are they trying to say? The adapter board and PCB end of the
connector
>>get 2.5 dB and the cable and it's plug side connector get 10 dB (after
>>compensating for the fixture's PCB end of the connector)?
The adapter board is the system adapter board, with SERDES, etc., etc. The
"fixture" is the test card that one might use to test a cable by itself.
The 10 dB is for the cable and the associated board connetor only but not
the test card or system adapter board.
>>2. Loss is said to be measured at 1250 MHz, but edge rates can be down to
>>100 pS (implying higher frequency content). Do we just lump the effects
of
>>these higher frequency components into the deterministic jitter?
100 ps rise times are consisnent with a bit time (Unit Interval) of 400 ps.
>>3. I assume the loss is power referenced to a 100 ohm system.
Yes.
>>Thanks for any opinions.
>>Matt
Thanks.
Jay Diepenbrock
Senior Engineer
Interconnect Technology & Qualification
IBM Global Procurement, B8UA/061, RTP, NC
Phone: 919-543-8804 Fax: 919-543-3642
Email: jaydiep@xxxxxxxxxx
---------------------- Forwarded by Jay Diepenbrock/Raleigh/IBM on
03/11/2002 08:51 AM ---------------------------
FreeLists Mailing List Manager <ecartis@xxxxxxxxxxxxx> on 03/09/2002
02:00:36 AM
Please respond to si-list@xxxxxxxxxxxxx
To: si-list digest users <ecartis@xxxxxxxxxxxxx>
cc:
Subject: si-list Digest V2 #70
si-list Digest Fri, 08 Mar 2002 Volume: 02 Issue: 070
In This Issue:
[SI-LIST] Loss requirements in Infiniband spec...
[SI-LIST] capacitive coupling to cables
----------------------------------------------------------------------
From: "ruston, matt" <ruston_matt@xxxxxxx>
Subject: [SI-LIST] Loss requirements in Infiniband spec...
Date: Fri, 8 Mar 2002 11:26:40 -0500
All:
Hi. I've been looking through the Infiniband 1.0a spec and need some
clarification.
1. Chapter 6, Table 17 specifies a LOSS of 2.5 dB for an adapter board, 15
dB of total loss, and (Chapter 7) 10 dB cable loss. Adapter board loss is
later defined in section 6.6.2.2 as "the maximum loss from the IC package
to
the board connector is 2.5 dB at 1250 MHz, including the connector."
Footnote 2 in Table 17 and Chapter 7 (Table 21) talk about the cable loss
(Vcout) as 10 dB "measured at board connector pins with equipment and
fixture contribution de-embedded." Section 6.6.2.3 says "including the
connector loss".
What are they trying to say? The adapter board and PCB end of the connector
get 2.5 dB and the cable and it's plug side connector get 10 dB (after
compensating for the fixture's PCB end of the connector)?
2. Loss is said to be measured at 1250 MHz, but edge rates can be down to
100 pS (implying higher frequency content). Do we just lump the effects of
these higher frequency components into the deterministic jitter?
3. I assume the loss is power referenced to a 100 ohm system.
Thanks for any opinions.
Matt
------------------------------
Date: Fri, 08 Mar 2002 22:34:49 -0800
From: "Douglas C. Smith" <doug@xxxxxxxxxx>
Subject: [SI-LIST] capacitive coupling to cables
Hi All,
I have posted the latest article on my website at http://emcesd.com:
Cable Effects Pt 3: Capacitive Pickup by Cables in a System
Last month the Technical Tidbit on http://emcesd.com explored inductive
coupling to cables. The article for March discusses capacitive coupling
to cables in a system and how to distinguish it from inductive coupling
by measurement technique. Also shown is how little coupling capacitance
is needed to keep you from shipping a system.
Enjoy! (I enjoy writing these articles and hope most of you enjoy
reading them!)
Doug
--
-------------------------------------------------------
___ _ Doug Smith
\ / ) P.O. Box 1457
========= Los Gatos, CA 95031-1457
_ / \ / \ _ TEL/FAX: 408-356-4186/358-3799
/ /\ \ ] / /\ \ Mobile: 408-858-4528
| q-----( ) | o | Email: doug@xxxxxxxxxx
\ _ / ] \ _ / Website: http://www.dsmith.org
-------------------------------------------------------
------------------------------
End of si-list Digest V2 #70
****************************
------------------------------------------------------------------
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 archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
- » [SI-LIST] Loss requirements in Infiniband spec. (si-list Digest V2 #70)
