| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
Yair, The change you suggest goes beyond fixing the simple known problem, described, and attempts to fix something you think is a problem, but frankly, I disagree. The definition says nothing about the timing of power on. It says “accepts power
simultaneously” – if this was for the PSE and said “applies power simultaneously” you might have a point – but that is not what it says.
The existing wording is consistent with resolved comments. I searched all comments on D3.0 which used the word simultaneous. Comment i-258 uses similar language to distinguish clause 33 PDs: “PDs that simultaneously require power from both Mode A and Mode B are specifically not allowed by this clause.”
It is also consistent with comment i-202, , from Arkady, on a different part of the text, 145.3.2 (the comment says 145.2.5.7, which would be a PSE section, but the resolution says the section number is wrong). The comment objects to the text: "Single-signature PDs that request Class 4 or less shall be able to operate if power is applied to either PD Mode A, PD Mode B, or both Modes simultaneously. All other PDs may require being supplied over Mode A
and Mode B simultaneously to operate at their nominal power level." Note this says “power is applied simultaneously” – a different meaning than “accepts power simultaneously”, as described above. The resolution was: “Change text to:" Single-signature PDs that request Class 4 or less shall be able to operate if power is applied to either PD Mode A, PD Mode B, or both Modes. All other PDs may require being supplied over both Mode A and Mode B to operate
at their nominal power level."” Additionally, your solution definitely creates a new problem – deleting “simultaneously” would just say “accepts power on both Modes.” – that clearly can be met by accepting power on either mode, one at a time. All types of PD have been
required to do this since the beginning, and it is NOT what we mean for type 3 and type 4 PDs. This definition is already out of scope. You don’t have agreement that you’re fixing a problem, and you create a new one. We really shouldn’t touch text we don’t have to, and the “accepts power simultaneously” is NOT touched. -george From: Yair Darshan [mailto:YDarshan@xxxxxxxxxxxxx] Hi George, It will be problematic not to accept the proposed change. We had similar comment in the draft body last cycle if I remember correctly on the same issue that was accepted. It will not make sense to have a correct text and wrong definition. The current definition with simultaneously is good only for implementations that has no other choice but to power on simultaneously. In the real world, the majority of the applications will do a wide range of powering sequences from almost zero to almost any realistic numbers. That is why
“simultaneously” is limiting and more important, incorrect. Yair From: George Zimmerman [mailto:george@xxxxxxxxxxxxxxxxxxxx]
EXTERNAL EMAIL Yair - the text relating to simultaneously powering has been accepted for a while. I would not support changing it. George A. Zimmerman, Ph.D. CME Consulting, Inc. Experts in PHYsical Layer Communications 310-920-3860
Hi George, In your proposed text: “1.4.418aa Type 3 PD: A single-signature PD that requests Class 1 to Class 6, or a dual-signature PD that requests Class 1 to Class 4 on both Modes during Physical Layer classification. Additionally, the PD implements Multiple-Event classification,
and accepts power on both Modes simultaneously. (See IEEE 802.3, Clause 145). 1.4.418ac Type 4 PD: A single-signature PD that requests Class 7 or Class 8, or a dual-signature PD that request Class 5 on at least one Mode during Physical Layer classification. Additionally, the PD implements
Multiple-Event classification, is capable of Data Link Layer classification, and accepts power on both Modes simultaneously. (See IEEE 802.3, Clause 145).” The use of “simultaneously” is confusing. Some readers will interpret that you must turn both modes at exactly the same time which is definitely not the intent, especial for dual signature PDs. I suggest the following to remove the word “simultaneously”. We had the same issue in the draft (main body) in previous releases and we fix it in this way. Yair From: George Zimmerman [mailto:george@xxxxxxxxxxxxxxxxxxxx]
EXTERNAL EMAIL OK, so the path of various solutions has gotten convoluted, and Heath pointed out to me offline that we need to straighten out Lennart & Andrea’s different parts. This combination seems to work: 1.4.418aa Type 3 PD: A single-signature PD that requests Class 1 to Class 6, or a dual-signature PD that requests Class 1 to Class 4 on both Modes during Physical
Layer classification, implements Multiple-Event classification, and accepts power on both Modes simultaneously. (See IEEE 802.3, Clause 145). 1.4.418ac Type 4 PD: A single-signature PD that requests Class 7 or One minor thing is that we can expect a comment on both of these that it reads like any single-sig class 1 to 6 (type 3, or 8 for type 4) is a type 3,4 PD; while a dual-sig requires the additional features listed. The English doesn’t
parse uniquely like a logic equation. What I was trying to wrap my head around was whether there was a reordering or rephrasing that made it clear that the logic was: ( (SS = class 7 + class 8) + (DS = maxclass 5) ) * (features). Breaking these into two sentences might work: 1.4.418aa Type 3 PD: A single-signature PD that requests Class 1 to Class 6, or a dual-signature PD that requests Class 1 to Class 4 on both Modes during Physical Layer classification. Additionally, the PD implements Multiple-Event classification,
and accepts power on both Modes simultaneously. (See IEEE 802.3, Clause 145). 1.4.418ac Type 4 PD: A single-signature PD that requests Class 7 or Class 8, or a dual-signature PD that request Class 5 on at least one Mode during Physical Layer classification. Additionally, the PD implements
Multiple-Event classification, is capable of Data Link Layer classification, and accepts power on both Modes simultaneously. (See IEEE 802.3, Clause 145). From: George Zimmerman [mailto:george@xxxxxxxxxxxxxxxxxxxx]
This does not fix the problem, because both type 3 and type 4 still include single-sig class 6. Add to that, that this is trying to make the definition complex. I suggest we stick to fixing the type 4 definition, and your suggestion below (on type 3) may provide the fix, but it isn’t readily apparent. The problem is that dual-sig which classify as class 5 or 6 on a pairset are outside the type 4 definition. We should focus on including those into type 4. -george From: Andrea Agnes [mailto:andrea.agnes181@xxxxxxxxx]
ok I agree but dual signature PD Class 5 would became both Type 3 and Type 4. I suggest accordingly modification of Type 3 definition: 1.4.418aa Type 3 PD: A single-signature PD that requests Class 1 to Class 6, or a dual-signature PD that requests Class 1 to Class 4 on both Modes during Physical
Layer classification, implements Multiple-Event classification, and accepts power on both Modes simultaneously. (See IEEE 802.3, Clause 145). 2017-10-20 11:54 GMT+02:00 Lennart Yseboodt <lennartyseboodt@xxxxxxxxx>:
|