| Thread Links | Date Links | ||||
|---|---|---|---|---|---|
| Thread Prev | Thread Next | Thread Index | Date Prev | Date Next | Date Index |
|
All – I made a commitment at the January interim to provide an overview of MPoE (Clause 189 in 802.3da) which might be used if we wanted power discovery and classification for coax or even for balanced pair.
I was going to present it this Thursday at the ad hoc, but given time constraints, that there were no comments related to powering on d0b, and that this was the only presentation offered for the ad hoc, I offered to save everyone the hour on Thursday morning
and have the presentation posted. It is at
https://www.ieee802.org/3/dm/public/0326/zimmerman_3dm_190226.pdf The presentation gives an overview of how PoE (yes, 2-pair and 4-pair PoE) does discovery and classification, how Clause 104 (SPoE/PoDL) differs from PoE, and how MPoE works – similar to PoE. You can, of
course, get much more detail by reading IEEE Std 802.3-2022 clauses 33, 104, 145, and soon IEEE Std 802.3da-2026 clause 189.
Remember I said power clauses are complicated. However, most of the complication is on startup operation, detection, and classification. These are the features that take a while to spec and make power projects
hard. They provide several benefits, including: ‘cold power’ (no operational power until source and load identify they are compatible), voltage & power-draw indication and negotiation where applicable, and ‘policing’ (PSE monitors the power drawn by the PD
vs. what has been negotiated), and mechanisms to maintain or discontinue power when load needs vary. They also specify power supply attributes like short circuit currents, maximum and minimum voltages (at the PSE and PD), and loop resistances. Many applications
just apply power and don’t need these, particularly applications like automotive where an engineer configures the parameters of the power supply and the load during the design phase of the link. The features that are hard to specify are generally driven by
a ‘plug-and-play’ (field-installable and potentially user-installable) paradigm like we see with PoE. Many, if not most, automotive applications do not fit this paradigm – hence we likely don’t need the level of detail that goes into a power clause – either
for coax or to reuse clause 104. However, any application of power must deal with noise and the electrical effects on the transmit signal. Since power coupling is usually in the passive circuitry between the driver/receiver in the PMD electronics
and the MDI (as well as between the PSE/PD PMIC and the PI) where we specify the ports operation these aren’t at an externally observable point. The effects of the power coupling circuitry & added noise must be considered in specifying the PHY. The presentation
ends with some thoughts on what we need to specify in single pair powering, what challenges are faced, and what we need to do. I think the minimum is not that much – and is independent of the classification/detection used. We can probably get much of it by reference to the Type F noise and ripple specifications in clause 104 (for
both TDD & ACT high speed). It leaves a little work to do, but not too much, in my opinion. If you are interested, please review, and I will be happy to discuss, either directly or on the reflector. I will be in-person in Vancouver if you wish to discuss there.
This is relevant to both TDD and ACT. I don’t think we need to add much to the draft to enable PoC and SPoE/PoDL for both PHYs. Perhaps with some side discussion we can bring in focused, consensus comments
either in March or initial WG ballot. George Zimmerman, Ph.D. President & Principal CME Consulting, Inc. Experts in Advanced PHYsical Communications 310-920-3860 To unsubscribe from the STDS-802-3-ISAAC list, click the following link: https://listserv.ieee.org/cgi-bin/wa?SUBED1=STDS-802-3-ISAAC&A=1 |