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

RE: [802.3af] Another Minor IEEE802.3af D4 issue - use of "PG"



Charles,
 
Definition of "ground" in the PD is basically implementation specific issue.
 
When the PD enclosure is a conductive material than the enclosure became "system ground" from the PD point of view and the RJ45 chassis lead should be connected to that enclosure.
The rest of the electrical circuit is DC isolated from the enclosure i.e. can be connected to the enclosure by HV caps.
 
When the PD enclosure is a plastic, you can assign to the RJ45 chassis lead the name "chassis ground" or "system ground" and connect it to the primary common or secondary common leads or both through HV cap.
Final architecture is depend on your final hardware and how to get min EMI etc.
 
In any case you can not connect the RJ45 chassis ground directly to the primary or secondary common.
 
Regarding the usage of different ground symbol, I agree that we should use the same symbols when it is relevant i.e. if we want to measure the signal with reference to chassis ground or the the hardware common point in the PSE or PD.
In practice, from high frequency ac voltage point of view, the chassis ground and the hardware common are connected together with high quality cap for emi control so actually, they are at about the same ac potential..
 
Yair.
 
 
-----Original Message-----
From: Charles Palmer [mailto:charles@xxxxxxxxxxxxx]
Sent: Thursday, April 24, 2003 6:46 PM
To: stds-802-3-pwrviamdi@xxxxxxxx
Subject: [802.3af] Another Minor IEEE802.3af D4 issue - use of "PG"

Dear team
 
Section 33.4.1 (Isolation) refers to "frame ground" and section 33.4.1.1.2 refers to "protective ground". Are these intended to be the same thing? If so would it be best to use a single term? (I hope this isn't opening a can of worms - the existing IEEE802.3 also uses both terms...).
 
In 33.4.2 (Fault Tolerance), figure 33-14 includes a node marked "PG" - presumably "protective ground". Is it appropriate to define PG within Clause 33, or to refer to it? I see that it is defined within Clause 14.3, but not apparently elsewhere within IEEE802.3. Where PG is used in the existing clause 14.3 it is first referred to in the text: "the ground for all common mode tests is circuit PG, the Protective Ground for the AUI".
 
Figures 33-15, 33-16, 33-17 and 33-18 show a ground or chassis symbol, but without the "PG" text, unlike Figure 33-14. Should the text "PG" be added? "PG" seems to be present in corresponding Clause 14 figures.
 
Figure 33-17 uses a different symbol for the ground or chassis node. Should this be the same symbol as for the other figures?
 
There are two instances of the ground symbols in both figures 33-17 and 33-18 - should any inference be taken about whether they are in fact connected together?
 
That's it - sorry to be a pedantic pain again.
 
 
 
Actually, the reason I've been looking at this is I'm trying to work out how to interconnect the RJ45 chassis ground, the "ground" on the primary side of the PD PSU and the ground on the secondary side of the PD PSU. It seems common practise to connect a high voltage cap from the transformer CT to the RJ45 chassis ground, but what happens when the PD is an a plastic box? Should the chassis ground connect by another capacitor to the secondary side GND, as I see in some PHY reference designs? Or by a low-resistance link? Or both?  If anyone who understands this can point me in the right direction I'd appreciate it (but not if you think I'm just looking for free consultancy!!! Actually, I'm happy to pay a bit to get this sorted...)
 
 
Regards -
Charles Palmer
 
Technical Director, DSP Design Ltd
email: charles@xxxxxxxxxxxxx
7 Tapton Park Innovation Centre, Brimington Rd, Chesterfield S41 0TZ, UK
ph: +44 (0) 1246 545 918