[802.3_10SPE] MDI connector comment r02-14

Thread Links	Date Links
Thread Prev	Thread Next	Thread Index	Date Prev	Date Next	Date Index

Dear colleagues,

I would also like to share my opinion on this comment.

The first question is whether the definition of connectors is necessary for the standard to be technically complete. As authors before me have mentioned, IEEE 802.3 standards exist with and without definitions of connectors. The measurement definition of a reference plane can also be carried out without a specific connector. Therefore, there is obviously no technical necessity to define certain connectors.

We also have to look at the differences between IEEE 802.3cg and other standards. In the past, it was often a question of developing communication technology for generic cabling in office buildings. Here the definition of the RJ45 as a uniform MDI connector and as a connector in this cabling was certainly a milestone for IEEE, TIA and ISO/IEC.

Today, however, the situation with IEEE 802.3cg is different. The focus here is on connecting IoT devices. These usually have no connection with the generic cabling, but follow their own installation rules depending on the area of application and expectations of the user groups. Within these application areas, manufacturer-independent interoperability and thus the basis for market success is always given. Therefore, the example with the RJ45 is not suitable for advertising a uniform connector.

It must also be questioned to what extent the plugs in question can be regarded as uniform MDI plugs at all. The differentiation of connectors according to E-Class has no technical background but is based on an artificial argumentation. Each of these plugs is basically suitable to fulfil all E-Classes. Therefore, this division will not be relevant in practice. Furthermore, there is no interoperability between the connectors of standards 63171-1 and 63171-6. The goal of a uniform MDI connector is therefore missed anyway.

In addition, 63171-6 describes several connectors that are incompatible with each other. Here the goal of a uniform MDI connector is violated within just one standard.

The question also arises as to whether the connectors are technically suitable at all. Standards 63171-1 and 63171-6 only require a dielectric strength of 1000V, although IEEE 802.3 requires a dielectric strength of 2250V. Also, the requirements for signal integrity are not defined according to the basic standard 63171, but only similarly. These cover the spectrum up to 600MHz and are not matched to the spectrum of IEEE 802.3cg, which makes economic solutions more difficult.

Projects 63171-1 and 63171-6 are still at a stage where technical changes are possible. Referencing them in an immature state is therefore a risk. It may be that IEEE 802.3cg would end up referencing MDI connectors that do not meet the requirements.

The selection of these connectors by the cabling committees TIA and ISO/IEC took place at an early stage when these aspects were only insufficiently known. Therefore, their recommendations can no longer be used as an argument for selecting an MDI connector. Moreover, these committees represent only the world of generic cabling and not the applications in which IEEE 802.3cg is likely to become most popular.

In summary, this means that the referencing of certain connectors is unnecessary, the proposed connectors are currently technically unsuitable and do not achieve interoperability, and the market does not benefit from referencing these connectors.

Therefore, I am in favor of completing IEEE 802.3cg without referencing certain connectors.

Bernd