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

Re: [802.3_NGECDC] Single twisted pair, 1km CFI slide review



Title: Standard
Some comments that may help.

I will not be a Macau, but as I recall we said that the CFI would be in San Diego.  And it looks like I am getting funding to attend IEEE again.

On 03/01/2016 11:35 AM, McCarthy, Mick wrote:

Dear 802.3 colleagues,

 

Please see below for comments on the “Single Twisted Pair – 1000m” slide deck as posted in the NG-ECDC Ad Hoc Public Area:

http://www.ieee802.org/3/ad_hoc/ngrates/public/16_01/Single%201000m_CFI_rev5.pdf

 

Please accept my apologies in advance if any of my comments show an obvious lack of knowledge of the IEEE 802.3 process, this is my first engagement with a project at this early stage.

 

Slide 3

1.       Motivation: I believe it may need strengthening with some market drivers, perhaps including but not limited to:

o   Desire from industry to converge to Ethernet-based connectivity solutions

§  Easier to manage single network type from enterprise to sensor

§  Reduces need for gateway-type functions, translating from different network types

o   Need for Ethernet with greater distances to replace existing fieldbus implementations

§  Greater bandwidth needed than existing 31.25kbps solutions


I believe MS/TP can go to 100kbps, but I don't know if this goes out to its max reach of 1200m.

§  Existing deployments utilise 1200m cable length


We will have to look at the cable specs for other standards.  Hopefully we will add PoDL that they lack.

o   Single pair cabling to enable deployment on some existing infrastructure

§  Lower cost and faster uptake through reuse of installed cable base


If we cannot used installed cable base, then it will be a major barrier to deployment.  Thus working with installed cable is a must have.

§  Ease of use – simple to deploy with lower skilled labour

 

2.       CFI:

o   I find the current wording to be a little confusing. Is the intent to:

§  Expand on what is implemented in 100BASE-T1 and/or 1000BASE-T1 to serve a longer cable length or

§  Create a new implementation to address the 1000-1200m Industrial requirement or

§  Both?


I could see that if this technology has a lower cost of entry than 100BASE-T1, it will find its way into areas like CANBus-FD alternative.

o   Not sure that this is clear and would suggest narrowing the focus to deliver on what is reflected in the ‘Motivation’ and the title slide.

 

3.       Similar to #2 above, is the work on P802.3bp or 802.3bw pre-set as the appropriate starting point for creating a solution to address the motivating needs?

o   Where might one locate the information relating to ‘the expertise of 100BASE-T1 on study of very long reach.’? Is this information available on the 802.3bw public or private area for consumption? Again, could this be interpreted as pointing towards a proposed solution prior to a study group being formed?

 

4.       Based on how any final standard or implementation may look, some additions to PoDL may be required, a modified coupling network, different power classes, etc. This may be a redundant comment on my behalf as it is looking towards a solution – the intent to support PoDL in some form does not change.

o   The comment on ‘optional AN’ is confusing to me. Is the suggestion a single PHY to support multiple *BASE-T1 standards? Apologies for my lack of knowledge, but has this functionality been included in the 802.3bw and P802.3bp work?

 

Slide 4

·         Title: I would suggest ‘at least 1000m’ to be consistent with the opening slide.

·         At the risk of sounding like a broken record, are we expressing a desire to extend the reach of 100 and 1000BASE-T1 or meet the needs of a >=1000m cable length or both? These may be mutually exclusive aims and require a divergent focus. Perhaps this is an acceptable challenge for a Study Group to resolve?


I believe the intent is >=1000m (and 1200m to match MS/TP)  to do this may result in a PHY different from the other -T1 efforts.

 

Slide 9

I perceive a lack of consistency in my reading of this slide where it could be interpreted as suggesting extending existing 100 and 1000BASE-T1 with a target of 100m? This doesn’t appear to be aligned with the 1000m number that pertains throughout the earlier part of the document.

 

Slide 11

I would suggest that there is quite a lot of work to be done at a Study Group to validate or repudiate some of the points here.

·         Is the suggestion that Automotive EMC requirements are consistently 10dB more stringent than is required for all Industrial environments? I believe this would need careful study with very close and broadly representative industrial supplier engagement to ensure all potential target markets are adequately investigated.


I believe Ludwig has significant data on industrial environments.  Perhaps he can contribute real numbers.  Back when I did Automotive EMC (circa '85 at AMC), EMC in the engine area was REALLY BAD, requiring significant design considerations.  I doubt that today's worst cases are any better.

·         Are we running the risk of being viewed as proposing a solution within this slide?

 

 

Summary

While there are undoubtedly excellent aspects to this slide deck, I’m left with a feeling of mixed primary aims throughout – extending existing work or creating a suitable solution to address the motivating needs.

 

I would also suggest that aspects of low cost and very low power are going to be significant contributors to the success of this in the intended end markets, particularly if this is positioned as the 802.3 solution for a greater play in IoT (which I believe it has great potential to be). I also believe that support for Intrinsic Safety (IS) from the outset will dictate whether any eventual new standard would be considered acceptable for many in an Industrial environment. While IS functionality is not required in every application, absence of it may preclude any adoption of this potential future standard as it cannot be deployed consistently across the whole plant network structure.


I would also like to bring up that low power will also impact security and there may need to be similar work in 802.1.  At most development of other EAP methods (IETF) that match the ablities of devices that will be on these networks.  This of course is my area of direct interest.

 

Thanks you for taking the time to read my comments and I look forward to continuing the discussion on the reflector or in person in Macau.

 

Best regards,

Mick McCarthy


--
Robert Moskowitz
Owner
HTT Consulting
C:      248-219-2059
F:      248-968-2824
E:      rgm@xxxxxxxxxxxxxxxxxxxx

There's no limit to what can be accomplished if it doesn't matter who gets the credit