Re: [802.3_4PPOE] fallout from the 20 ohm conversation

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

Hi Lennart,

Thanks for reviewing.

Please see my response to your comments below.

Thanks

Yair

From: Yseboodt, Lennart [mailto:lennart.yseboodt@xxxxxxxxxxx]
Sent: Friday, May 12, 2017 9:39 PM
To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_4PPOE] fallout from the 20 ohm conversation

EXTERNAL EMAIL

Hi Pete, Yair,

Updated comments. Yair, your minimum PD voltages are wrong.

Yair: I agree in principle. We need that Type 3 and 4 PDs class 1-3 will be supported by Type 1 and 2 PSEs with Vpse=44V for class 1-3 with 12.5 ohms. So these are the numbers I got.

Class 1: 42.88V è 42.8 (rounding down for worst case) instead of 42.9

Class 2: 42.07V è 42 (rounding down for worst case) instead of 42.1

Class 3: 39.94V è 39.9

See calculations attached.

Type	Class	Vpse	Rch	Pclass-PD	Vpd	Ipd	Pclass 2-pais	Pclass Spec	Diff
1	1	44	12.5	3.84	42.8806	0.0896	3.9402	4	0.0598
1	2	44	12.5	6.49	42.0717	0.1543	6.7875	7	0.2125
1	3	44	12.5	12.95	39.9478	0.3242	14.2636	15.4	1.1364
2	4	50	12.5	25.5	42.5000	0.6000	30.0000	30	0.0000

About rounding... convention has always been to have "nice" numbers on the PSE side and accept PClass_PD as calculated.

We have made a couple small roundings to PClass_PD (eg. Class 5 was rounded up to 40W).

Yair: Since this is a spec, and spec need to be accurate and sync, accurate and make sense with text and related Tables, we must make sure that the calculated values will not be higher than the fixed worst case numbers in Table 145-11.

So, to round up PClass is OK. Rounding done is not.

As a result, we can round the following up:

3.92W è 4W

6.72W è 6.8W

451.W stays 45.1W since the calculated value is higher than 45W. We can adjust PD number, I don’t like it, but we can do it. Are you preferring this option? In This case Ppd for Class 5 will be 39.94W instead of 40W.

See other solution to this issue in the document to add a text to clarify the source of potential differences between the calculated values and fixed values in Table 145-11.

Having easy to remember numbers is more important than being precise down to 10 mW.

Yair: It was never an objective to remember numbers….

As such I've rounded the numbers to make as much sense as possible (matches with Pete's PClass numbers).

Yair: Please see above for what is the definition in my opinion for “making sense” underlined above.

Yair: Regarding your comment in page 4: “Does not match D2.4 text, nor does it produce a valid sentence. Please check.”

The text without the changes is D2.4 text.

The modified proposed changes is to split the sentence to two parts.

The first part says the when you plug the Vpse_min and Rch you will get the over-margined values in Table 145-11.

The 2^nd part is explaining that the over-margined values where taken per the lowest PSE Type which is Type 3.

Kind regards,

Lennart

From: Peter Johnson <peter_johnson@xxxxxxxxx>
Sent: Friday, May 12, 2017 17:12
To: Yseboodt, Lennart; STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: RE: [802.3_4PPOE] fallout from the 20 ohm conversation

Yair, Lennart:

If rounding mathematically, I get same values as Yair except at Class 3, I get 13.9W.

If rounding UP to nearest 10th in order to be conservative at the PSE, it would be:

Class 1 4W

Class 2 6.8W

Class 3 14W

The calculation for Class 5 is 45.1W – rounding this down is of course, NOT conservative.

I don’t like the language “worst case PSE Type parameters”. Worst case is never very descriptive. Not sure new text is needed – it should be already covered in footnotes to the table I think.

Also, if we toss out Option B, then the footnotes for the table 145-11 will need be adjusted as they presently say “for minimum Vport_pse-2P” and these values are all for 50V Vport_pse-2P.

Regards,

Pete J

Sifos

From: Yseboodt, Lennart [mailto:lennart.yseboodt@xxxxxxxxxxx]
Sent: Friday, May 12, 2017 9:27 AM
To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_4PPOE] fallout from the 20 ohm conversation

Hi Yair,

Review & comments attached.

Looks OK but needs couple of fixes.

Lennart

From: Yair Darshan <YDarshan@xxxxxxxxxxxxx>
Sent: Friday, May 12, 2017 2:03
To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_4PPOE] fallout from the 20 ohm conversation

Hi all,

1. Based on the last conclusions, my proposal to exclude the 20 ohms’ cases form the calculations of Pclass fixed values in Table 145-11

as derived from the worst case of Type 3 and 4 PSEs connected to all PD types, stays as proposed. Agreed?

2. Please see attached Rev003 in which most of the reviewers choose Option A solution.

3. In addition, I add proposed changes for Vpd, as result of the consequences of (1) and (2).

I’ll appreciate your inputs.

Yair

From: Chad Jones (cmjones) [mailto:cmjones@xxxxxxxxx]
Sent: Friday, May 12, 2017 12:46 AM
To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_4PPOE] fallout from the 20 ohm conversation

EXTERNAL EMAIL

Excellent, that’s not what I got from your original message. But it’s been a long day. My reading comprehension might be lowered… ☺

Chad Jones

Tech Lead, Cisco Systems

Chair, IEEE P802.3bt 4PPoE Task Force

From: Geoff Thompson <thompson@xxxxxxxx>
Date: Thursday, May 11, 2017 at 5:18 PM
To: Chad Jones <cmjones@xxxxxxxxx>
Cc: Geoff Thompson <thompson@xxxxxxxx>, "Larsen, Wayne" <WLARSEN@xxxxxxxxxxxxx>, David Law <dlaw@xxxxxxx>, 4PPOE Reflector <STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx>
Subject: Re: [802.3_4PPOE] fallout from the 20 ohm conversation

Chad-

We are in violent agreement

Geoff

On May 11, 2017, at 2:10 PMPDT, Chad Jones (cmjones) <cmjones@xxxxxxxxx> wrote:

Geoff, we want to remove support for 20 ohms. I don’t want to include old revisions pick up old specs. Our objective clearly states that we are supporting only Class D from 118011 1995 and 2002, Cat5 and Cat5e, with the added restriction that loop resistance is 25 ohms. I just need to make the draft text match this objective.

We are not taking the most conservative approach. Our system requirements clearly state minimum cabling requirements. We are purposely excluding 40-ohm loop resistance cabling.

Chad Jones

Tech Lead, Cisco Systems

Chair, IEEE P802.3bt 4PPoE Task Force

From: Geoff Thompson <thompson@xxxxxxxx>
Date: Thursday, May 11, 2017 at 4:59 PM
To: "Larsen, Wayne" <WLARSEN@xxxxxxxxxxxxx>, Chad Jones <cmjones@xxxxxxxxx>, David Law <dlaw@xxxxxxx>
Cc: Geoff Thompson <thompson@xxxxxxxx>, 4PPOE Reflector <STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx>
Subject: Re: [802.3_4PPOE] fallout from the 20 ohm conversation

Wayne/Chad/David-

What Wayne says is true

and I believe a similar situation exists wrt the ISO/IEC standard.

The problem is that both of the cabling standards are, in effect, purchase specifications for new builds and installs.

That is somewhat different than the case with 802.3bt where (although we are specifying new attachment equipment) we are specifying operating that equipment on the installed base of cabling, much of which is not new. The most conservative approach would be to spec everything against the original smaller gauge of TIA Cat 3, DIW and old German quad wiring. I believe it is our collective opinion that (with exception to the note David mentions) we don't need to do that, rather we believe that the lowest appropriate common denominator at this point in time is legacy Cat 5. It is appropriate to reference old versions of the cabling standards where necessary to pick up those specs.

Geoff

On May 11, 2017, at 1:42 PMPDT, Larsen, Wayne <WLARSEN@xxxxxxxxxxxxx> wrote:

Hi,

Maybe you all know this, but ANSI/TIA-568-C.2 superseded ‘B.2, ‘B.2-1, and ‘B.2-10. ‘2-1 and ‘2-10 were addenda to ‘B.2.

Wayne

5/11/17

From: Chad Jones (cmjones) [mailto:cmjones@xxxxxxxxx]
Sent: Thursday, May 11, 2017 3:39 PM
To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: [802.3_4PPOE] fallout from the 20 ohm conversation

All, the 20 ohm loop resistance discussion trigger déjà vu for David Law, so he checked through his archives. Sure enough, we had this EXACT conversation during development of 802.3at. as a result of this conversation, we placed some specific text in AT to define the cable while restricting the loop resistance. This also lead to analysis of the objectives and review of D2.4 for compliance. It was found that we are not in compliance with the objectives.

I have put together a doc showing David’s email, the objective, the D2.4 text, the AT text and the late comment I will file to close this hole. The only question I have is for the cabling folks:

For Cat5e, we list 4 ANSI standards: ANSI/TIA-568-C.2, ANSI/TIA/EIA-568-B.2, ANSI/TIA/EIA-568-B.2-1, and ANSI/TIA/EIA-568-B.2-10. In D2.4, we only list ANSI/TIA-568-C.2. Is it sufficient to list this one standard or is there something added by including those other three?

Chad Jones

Tech Lead, Cisco Systems

Chair, IEEE P802.3bt 4PPoE Task Force