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

Re: [802.3_10SPE] Power Class Proposal



My focus is commercial buildings. The “Construction Specifications Institute” (web address is arcat) curates the specification template that is used for large construction project in the United States (and used as a model for smaller projects). Division 23 is dedicated to “Heating Ventilating and Air Conditioning”.  When public money is spent, the specification is often public information and sometime available online. Typically, one finds language such as “control wiring shall be furnished and installed in accordance with requirements specified in the National Electric Code and all applicable local codes.” Such language also matters for insurability. The combination of construction practices, insurance practices, and code makes for slow change; all of these parties need to be comfortable with a change before it happens. My experience is the codes, particularly the National Electric Code, are important.

 

-David

 

From: Bruce Nordman [mailto:bnordman@xxxxxxx]
Sent: Friday, May 17, 2019 1:26 PM
To: STDS-802-3-10SPE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_10SPE] Power Class Proposal

 

My focus is residential and commercial buildings - I have no experience with industrial (or elevators).

 

In 10 or 20 years, the vast majority of devices using SPE will be new ones, not the existing industrial or building automation ones.  I don't think we need to or should burden those new devices or the PSEs they connect to with a second voltage, so I concur with Lennart.

For these industrial/building-automation markets, they are only used by a limited number of people with a limited range of products - not the large numbers that use ordinary Ethernet today. 

I have two types of AC outlets in my house - 120V and 240V.  The 240V are for specialized devices in special locations.  I think this is analogous to the 48 vs 24V for DC.

I think a second physical layer of standard for power (communications could be the same) with a different connector would be appropriate for the 24V application.  The two standards could share a lot of content and in many cases have tables with a 48V column and a 24V column.  This should reduce costs and increase efficiency for everyone.  Why not do this?

 

My knowledge of electric codes/regulations is limited, but at my DOE lab, 50V is the safety threshold for AC but for DC it is 100V.  SPE switches could have a global setting to not go over 50V in cases where they have to, but have a simple switch to turn that limit off, when the limit is raised.

 

( I am not an EE, but is it worth considering +/-30 for the 24V sector so that one could provide 60V?  I wouldn't bring this into the ordinary buildings space of nominal 48V - I think that should be as simple as possible. )

 

There are companies selling 48V lighting today - in the last week I came across:    https://www.w.atx-led.com/

 

--Bruce

 

 

On Thu, May 16, 2019 at 3:13 PM George Zimmerman <george@xxxxxxxxxxxxxxxxxxxx> wrote:

Lennart –

Thanks for your thoughts on this, and I generally agree with your sentiment on interoperability.

However, while what you say is true about designing PDs, it ignores the impact of using a higher voltage on the installation issues.

 

I, too, have pause at keeping 3 voltages, but think that 2 can be managed, and that 2 voltages actually fits well with the market requirements.

I suspect that we cannot go with fewer than 2 voltages without limiting our market potential.  This is because single pair applications go beyond Information Technology environments and have to deal with regulatory issues in environments which are not as benign as enterprise LAN applications.

 

Generally, PoE was designed in a world where we considered SELV limits, similar to US NEC Class II inherently limited power sources.  Those set up a 100VA per circuit limit and 60 Volts as the threshold.

 

However, if you are in damp, dusty or other environments – and 10BASE-T1L is designed to go into these – you may have other limits.  These are best allowed by a 24 volt system.  There are also (as I know you are aware) a host of regulations which say “50 volts” (sometimes without either AC or DC) and a 50 volt system would avoid arguing about installations because of this limit.

 

A long time ago we asked the industrial crowd whether they could live with the SELV 50-60 V systems that normal PoE would give – the answer was that they generally worked with 24 volts.  The reason why was buried in various codes and regulations.

Looking through the NEC, below are a summary of a number of the limits:

 

30V limits:

For example, low voltage lighting where wet contact is likely to occur has a limit of 30 volts DC in the NEC (Article 411).  This includes landscape lighting in damp areas and lighting around pools (Table 300.5).

Class I power limited circuits are rated at 30 volts max (725.41A).

(Hazardous locations - fire or explosion hazards) called Class I, II or III locations (article 501) require covers for protection at greater than 30 volts, and in wet locations, greater than 15 volts.

Swimming pools require less than 30 volts for safe contact on continuous dc circuits (article 680)

 

42 volt limits:

Elevators allow flexible cords and other rules which make installation easier when voltages are less than 42 volts dc. (article 620)

 

50 volt limits:

Electric vehicle charging systems (article 625) have more stringent rules for connections supplying ventilation equipment when the voltage supplying the ventilation equipment is more than 50 volts dc.

Circuits under 50 volts get all sorts of special treatment like exceptions from requirements and special requirements making things easier: (articles: 200.7 (wiring), 250 (various grounding & marking), 409 (disconnects), 422 (listing of appliances), 430 (motors), 522 (control circuits in amusement environments), 625 (see above), 665 (heating equipment), 668 (battery power supplies), 708 (Wiring of HVAC, Fire Alarm, Security, Emergency Communications, and Signaling Systems.), and then there are articles 720 (circuits under 50 volts).

 

While in many cases, the NEC says a voltage limit (30 or 50 volts) based on AC, the lists above do not include those which say AC.  They are either for DC or they don’t specify AC or DC, so they apply to both.  While some of the 50 volts might be extended to 60 volts dc in the future, or better yet, class II, they are not currently.  SO, what this means is that permitting, licensing, labeling, listing, and various regulatory rules governing installations get harder at these voltage limits, making it harder to address the market.

 

In short, we need to be cognizant of more than just what it takes to build a PD.

 

The easiest path, I think, from a market acceptance standpoint is to specify 2 voltage limits.  The 24 volt limit seems an obvious choice.

The question in my mind is for the higher limit:

  • whether to go for a 50 volt limit and maximize ease-of-use, or go for a 60 volt limit and allow a little more power transfer. (what do we enable by that?)

 

-george

 

From: Lennart Yseboodt <00000b30a2081bcd-dmarc-request@xxxxxxxx>
Sent: Thursday, May 16, 2019 12:51 AM
To: STDS-802-3-10SPE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_10SPE] Power Class Proposal

 

Hi Heath, all,

 

The goal of 10BASE-T1L + PoDL should be to maximize interoperability while offering the performance needed for as many applications as possible.

We need to answer two questions to come to a conclusion on your proposal.

 

1. Do we need a multi-voltage power standard ?

2. What is/are the right voltage(s) ?

 

With regard to (1), we can learn from other standards that multi-voltage power schemes tend to be far more complicated and fail to achieve wide interoperability.

Take for example USB-PD. This standard supports 3 voltage levels. Despite being released already years ago, the market traction is low.

In fact, this USB standard is eclipsed by a proprietary fast charging standard (you can all guess which one), that offers a simpler mechanism to achieve high charging speeds.

 

For a PD design, supporting multiple voltage levels creates quite a bit of extra complication and testing. The PD can choose to only support one of the voltage, but this then goes at the cost

of not working with all PSEs.

Similarly on the PSE side, a multi-voltage PSE is much more complicated than a single-volt one. Also power budgeting becomes more complicated.

I believe that interoperability and product simplicity is best served by choosing a single voltage to be supported.

 

That takes us to question (2)... which one ?

 

Higher voltages are (far) more efficient at transferring power.

For example, we need to deliver 8W over 100m of single-pair cable with AWG24.

One supply is 24V, the other is 53V.

 

The 24V supply will source 13.1W, with 5W dissipated in the cable.

The 53V supply will source 8.4W with 0.4W dissipated in the cable.

Both deliver 8W to the respective PD.

 

Additionally, 8W is about the limit that a 24V supply can deliver in a stable manner (for this cable), the 53V supply can go above 40W.

 

Very few electronic designs operate from 50V directly. We know this problem from PoE.

But converting from 50-60V down to 24, 12, or 5V is not a complicated or costly issue, many efficient switch mode designs exist that can do this.

 

Conclusion:

- Operating from a single voltage will greatly simplify PD and PSE design and support interoperability

- Picking the highest voltage range allows us to deliver the most power and deliver power FAR more efficiently than a lower supply voltage can

- It is not complicated to build PDs that need 24V internally to support a PoDL voltage of 50+V

 

Kind regards,

 

Lennart

 

On Wed, 2019-05-15 at 23:45 +0000, Stewart, Heath wrote:

All,

 

I wanted to open a discussion on the power class proposal referenced in i-321.

 

http://www.ieee802.org/3/cg/comments/Comment_i-321_Stewart_3cg_clause_104_modifications_v1.pdf

 

The proposal assigns class power voltage, current and power values to make required corrections to cable resistance and to accommodate the following use cases:

 

  1. 24V nom
  2. 50V max
  3. SELV max

 

All other changes derive from the above introduced values. The proposal attempts to cover all suggestions offered during a large number of discussion during and between cg sessions in the Vancouver meeting.

 

In the time since the proposal was submitted a couple of weeks ago I have collected a number of additional inputs from the cg task force. My goal in sending this to the reflector is to kickstart public and private discussions with the goal of streamlining our efforts in cg next week.

 

It is worth pointing out the there are a number of possible end markets: eg building automation and industrial automation. We have gotten really good input from these end markets and strive to enable relevant cabling resistances while remaining aware of economic and technical feasibility.

 

While process automation is an exciting market as well, efforts to marry PoDL to process automation Intrinsic Safety (IS) requirements stalled early on. Therefore process automation requirements are not directly considered in this PoDL contribution.

 

The vast majority of the feedback from colleagues has regarded the 3 proposed voltage ranges. Mainly the comments address the appropriateness of supporting 1 vs 2 vs 3 voltage ranges.

 

Please feel free to reply on this thread or offline. I will attempt to come prepared to modify proposed text to reflect the will of the group.

 

Cheers,

 

Heath Stewart

 

To unsubscribe from the STDS-802-3-10SPE list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-10SPE&A=1


To unsubscribe from the STDS-802-3-10SPE list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-10SPE&A=1


To unsubscribe from the STDS-802-3-10SPE list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-10SPE&A=1



--

Bruce Nordman
Lawrence Berkeley National Laboratory
nordman.lbl.gov
BNordman@xxxxxxx
510-486-7089
m: 510-501-7943


To unsubscribe from the STDS-802-3-10SPE list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-10SPE&A=1


To unsubscribe from the STDS-802-3-10SPE list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-10SPE&A=1