| All,
I think we should differentiate between MDI and connector. In my opinion, the MDI (Medium Dependent Interfaces) defines the interface between the PMD sublayer and the medium from the PHY perspective. A connector is an entity more complex that implements all the mechanical features needed to meet many other requirements that are not directly related to the PHY specification. For example, depending on how the connector housing is designed, the retention force, the locking force, the water proof, etc can be different. In general we can have several connector for a same MDI, as well as, different cable designs for the same optical fiber (e.g. same optical fiber can be assembled with different coating, jacket, strength elements, etc, which provides different characteristics against environmental, chemical and mechanical loads).
To give and example, in the case of 1000BASE-RHC, in the market there are several connectors, but all of them use the same MDI. Being more specific, the ISO 21111-4 specifies two different types of connectors for 1000BASE-RHC ports, but both of them use the same MDI, which consist always in SI-POF butt-coupling with no contact. Different types of connectors have different mechanical characteristics, but all of them same medium interface. Of course, both types of connectors are not mating compatible. In the same way, we can have different SI-POF cables that meet with the medium characteristics of 802.3 C/115, however some of them support up to 85ºC, others up 105ºC or more, and depending on the cable structural elements the mechanical loads that can support are different.
During the 802.3bv project we did not have the necessity of defining the MDI for RHC PMD, because all the different existing connectors are equivalent in this respect. More important, 802.3bv leverages the cable and connector components of MOST, already used in automotive.
I think 802.3cz is quite different. No components exist for automotive yet, and a big development of components exist for other markets, like data-centers. Automotive and data-centers are very different in requirements. . In my opinion, 802.3cz cannot leverage data-center components as is. Therefore we need contributions that provide arguments to support the use of PC and/or EBO interfaces in automotive applications. Depending on the MDI definition supported by the scientific data provided in such contributions we will be able to decide the IL, RL and any other related MDI parameter that, of course, affects the link budget. I really think that we need this information for a consistent specification of the PMD sublayer. I believe the following contribution summarizes very well the point:
Best regards,
Rubén Pérez-Aranda
CTO at KDPOF_____________________________________________________________
Knowledge Development for POF, S.L.A: Ronda de Poniente 14 2º CD, 28760, Tres Cantos (Madrid), Spain P: +34 91 804 33 87 Ext:110 M: +34 689 319 866
Hi Bob, Please tell us how to count straw poll today. Is it also roll call for about 80 attendees? I have another question on the purpose of straw poll. Is it just a straw poll? Or does it intend to write connector design in future 802.3cz standard? I think to select butt joint or EBO is too specific, and my understanding is it is out of scope of 802.3cz. We should specify minimum optical requirement for connector, not physical connector design. Best regards, Yuji Watanabe, AGC Colleagues, This is a reminder of our call on 18 November 2020. This will be the third P802.3cz Task Force teleconference meeting held in conjunction with the IEEE 802.3 Working Group plenary session. As a Task Force meeting (not an ad hoc), we will be able to consider technical motions. Our draft timeline puts us at the point where we should be adopting baseline presentations that we can use to put technical content into our initial draft. If you would like to take a straw poll or make a motion, please provide text to me or optionally, reply to the reflector with motion and any additional information you wish to provide other participants. Our TF Secretary has the actual straw poll / motion language already for the following (only summary is provided below): 3. Motion #2 — Begin TF review of draft when D1.0 becomes available. 4. Motion #3 — Adopt timeline. (The actual timeline proposed for adoption in the motion will be the result of discussion on 18 Nov teleconference. Because the meeting is being held as a teleconference, only 802.3 voters will be allowed to make and vote on motions. Anyone can respond to straw polls. When there is no objection, we will adopt motions by unanimous consent, one exception being adoption of any baselines where the motion will be by roll call. I look forward to joining with you on the call in ~21 hours. Chair, P802.3cz Multi-Gigabit Optical Automotive Ethernet
| IEEE 802.3 invites you to join this Webex meeting. |
Meeting number (access code): 173 094 1945 |
Meeting password: 4quHe6BXmP8 |
Wednesday, November 18, 2020 | 8:00 am | (UTC-04:00) Eastern Time (US & Canada) | 2 hrs |
Join by phone | +1-646-992-2010 United States Toll (New York City) | +1-213-306-3065 United States Toll (Los Angeles) | Global call-in numbers |
|
To unsubscribe from the STDS-802-3-OMEGA list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-OMEGA&A=1
To unsubscribe from the STDS-802-3-OMEGA list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-OMEGA&A=1
To unsubscribe from the STDS-802-3-OMEGA list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-OMEGA&A=1
|