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

Re: [802.3_4PPOE] Your proposal for Single-signature definition in

Hi Lennart,
1. Please see in my mail below that I say the the current injection method will not work if we are looking for non 25K signature. I know that voffset will be change but the problem is that nobody checks for this or care about it especially when PSE is testing for valid signature i.e. 25K YES OR NO since we are doing differential test in which Vofsset is canceled.
2. As for using it (Voffset ) for the PD definition it may be possible if it was defined clearly that this is the intent and a clear pass/fail criteria was define which is not the case. That is why the proposed text for current will not work.
3. We still working on the text (me and Mike) to fix this.

-----Original Message-----
From: Lennart Yseboodt [mailto:lennartyseboodt@xxxxxxxxx] 
Sent: Monday, November 6, 2017 6:11 PM
To: Yair Darshan <YDarshan@xxxxxxxxxxxxx>; STDS-802-3-4PPOE@xxxxxxxxxxxxxxxxx
Subject: Re: [802.3_4PPOE] Your proposal for Single-signature definition in


Hi Yair,

But it does work...

If a voltage is applied to the corrupting pairset, a single-signature PD will "not show a valid detection signature" because Rdetect will be outside the limits of Table 145-21.

If a current is applied to the corrupting pairset, a single-signature PD will "not show a valid detection signature" because Voffset will be outside of the limits of Table 145-21.

Both corrupting methods are viable.

Kind regards,


On Mon, 2017-11-06 at 15:02 +0000, Yair Darshan wrote:
> Hi David,
> I have reviewed your proposal and I see that the current injection 
> method may not work if you are looking for invalid signature i.e.
> 25K.
> The reason is that when you inject two current levels in mode A (I1,
> I2 and e.g. I1>I2) and find valid signature (25K) on that mode while 
> the current on mode B is 0, and then you inject new current level I3 
> on mode B while you keep I2 in mode A, you will get valid slope of 25K 
> (I am ignoring the offset voltage for now) which means it shows a 
> valid signature (25k) while I supposed to get invalid signature (in 
> terms of resistance and no other parameters that defines invalid 
> signature).
> In general, any current you inject on mode B when mode A is conducting 
> will just generate offset voltage across Rsig but will not change the 
> slope of Rsig=25K. The reason for it is that detection is done always 
> in differential way (two values of current or voltage) which cancel 
> offsets.
> The best approach to define single-signature PD is to use the voltage 
> concept. It will always work. Please remember that the way we define 
> the PD is it single-signature or not should not imply and doesn’t not 
> imply how we implement connection check. Moreover, I am also 
> suggesting adding a text that say it explicitly.
> To summarize:
> I prefer the following definition:
> A single-signature PD shall present a valid detection signature, as 
> defined in Table 145-21, on a given Mode when no voltage or current is 
> applied on the other Mode, and shall present an invalid not present a 
> valid detection signature on that Mode when any voltage between 10.1V 
> and 57V at least one voltage between 3.7V and 57V is applied to the 
> other Mode or any current greater than 124μA is applied to the other 
> Mode. These requirements apply to both Mode A and Mode B.
> This definition doesn’t imply on the way how connection check is 
> implemented weather using voltage or injecting current.”
> NOTE—A valid detection signature meets every requirement in Table
> 145-21 across all specified conditions. A failure under any allowed 
> conditions of Table 145-21 is considered “not a valid signature while 
> “invalid signature” is when the requirements of Table 145-22 are met.
> 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>.