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

Re: [802.3_4PPOE] PD isolation requirements - Updated presentation for comments Comment r02-119, r02-70



Hi Andrea and Lennart,

It is not clear Lennart why you determine that to make it work, we need complete isolation on both positive and negative in your option 1? (with option 2, I am OK)

What is the technical basis of this determination (for option 1)?

The facts are:

  1. Dual signature PDs works fine with isolation only in the negative for the last 10 years.
  2. The PSE guarantee supporting PDs with isolation only in negative pairs. It doesnt guarantee supporting on the positive pairs!
  3. 99.999% of PSEs, shorts the positive pairs which effectively shorts the positive pairs in the PD so there is no point to require isolation at the PD on the positive pairs.
  4. PSE is required to have its switches on the negative pairs. As such, also the current measurements have to be in the negative pairs. Having switches in the negative pairs and measuring current on the positive pairs will not work due to the unbalance issue.

So asking PDs to have isolation on the positive should be optional since it is not required for interoperability and most important, there is no technical need for it (it will do nothing..).

 

It is no problem to require PSE to measure on the negative pairs per your 2nd option even if practically nothing will work if we measure on the positive pairs and switch on the negative pairs (it is the controllability criteria that will be violated).

 

I like Andreas text for its simplicity. Andreas text proposal (to remove the 10uA isolation requirement) might work but as Lennart says, the text need to cover the full voltage range not just Voff. I will work on it tomorrow as 2nd option to my proposal in darshan_02_0118.pdf.

 

Thoughts?

 

Regards

Yair

 

From: Lennart Yseboodt [mailto:lennart.yseboodt@xxxxxxxxxxx]
Sent: Tuesday, January 16, 2018 4:27 PM
To: STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_4PPOE] PD isolation requirements - Updated presentation for comments Comment r02-119, r02-70

 

EXTERNAL EMAIL

Hi Andrea,

 

One of the 4PID methods involves performing detection on an unpowered pairset when the other pairset is turned on.

 

In order to make it work we have two choices:

(1) Require complete isolation between the pairsets of dual-signature PDs

(2) Require that PSEs measure on the negative pairs and mandate isolation on the negative side of dual-signature PDs

 

Kind regards,

 

Lennart

 

 

On Tue, 2018-01-16 at 11:03 +0100, Andrea Agnes wrote:

Hi all,

 

sorry for previous mail, it was incomplete answer.

probably the way to reach a possible definition for dual signature PDs that can tie positives is not to add an isolation requirement because it is not an isolated solution.

Furthermore the definition of dual signature PDs include an isolation requirement, then I suggest 2 action in order to obtain the target:

   - remove the additional requirement: Dual-signature PDs shall have less than or equal to 10 μA of current between any one conductor of Mode A and any one conductor of Mode B when VPD, as defined in 145.1.3, of either Mode is less than VOff_PD min, as defined in Table 145–9. See Table 79–f.

   - modify the definition of dual signature PDs

    from: A dual-signature PD shall present a valid detection signature, as defined in Table 145–21, on a given Mode, regardless of any voltage between 0 V and 57 V applied to the other Mode. This requirement applies to both Mode A and Mode B.

    to: A dual-signature PD shall present a valid detection signature, as defined in Table 145–21, on a given Mode, regardless of any voltage between 0 V and VOff_PD applied to the other Mode. This requirement applies to both Mode A and Mode B.

 

Thank you for your attention

 

Regards

Andrea

 

2018-01-16 10:54 GMT+01:00 Andrea Agnes <andrea.agnes181@xxxxxxxxx>:

Hi all,

 

probably the way to reach a possible definition for dual signature PDs that can tie positives is not to add an isolation requirement because it is not an isolated solution.

Furthermore the definition of dual signature PDs include an isolation requirement, then I suggest 2 action in order to obtain the target:

   - remove the additional requirement: Dual-signature PDs shall have less than or equal to 10 μA of current between any one conductor of Mode A and any one conductor of Mode B when VPD, as defined in 145.1.3, of either Mode is less than VOff_PD min, as defined in Table 145–9. See Table 79–f

 

2018-01-11 9:44 GMT+01:00 Andrea Agnes <andrea.agnes181@xxxxxxxxx>:

Hi Yair,

 

In my opinion there are 2 critical points in your proposal solution:

 

a)The voltage VPD is measured between any positive conductor of a pairset and any negative
conductor of the corresponding pairset, then it is not possible to measure current between negative pairs using VPD definition.

 

b)If PD dual signature ties positives together (problem #2, point d) and VPD is applied only to pairset mode A
at the negative pair of mode B the current is very higher than 30uA (for example if VPD is included in classification range).
Therefore your solution doesn’t include the possibility to join positive together in the dual signature PD.

 

Andrea

 

Andrea AGNES | Tel: +39 0293517504

Analog & MEMS Group | Industrial & Power Conversion Division | IC Design Project Leader

 

STMicroelectronics srl

Via Tolomeo, 1 | 20010 Cornaredo (MI) | Italy

 

ST online: www.st.com | Follow us on twitter: @st_world

 

2018-01-10 21:06 GMT+01:00 Yair Darshan <YDarshan@xxxxxxxxxxxxx>:

Hi all,

Please find updated presentation for comment r02-119 and r02-70.

Ill appreciate your inputs.

Yair

 

Darshan Yair

Chief R&D Engineer

Analog Mixed Signal Group

Microsemi Corporation

 

1 Hanagar St., P.O. Box 7220
Neve Ne'eman Industrial Zone
Hod Hasharon 45421, Israel
Tel:  +972-9-775-5100, EXT 210.

Cell: +972-54-4893019
Fax: +972-9-775-5111

 

E-mail: <mailto:ydarshan@xxxxxxxxxxxxx>.