Re: [802.3_4PPOE] Are diode bridges really needed (2).pptx
While some may say they are out to
eliminate the scourge of diodes, what is being suggested is eliminating polarity insensitivity and allowing back feed into the network. So in order to not pay for efficient circuits to accomplish these features, what is suggested is to move Ethernet from something that can be installed by anyone to a system that requires a trained technician because when you make things that will not always just work you are saying only those with proper training will succeed. What will happen to the PD when its input is powered with reverse polarity? My bet is either something really bad or these high current PDs will have to be some kind of polarity protection circuitry, so why eliminate the polarity insensitivity?
Nothing prohibits a individual or organization from creating a PD today that does not have polarity insensitivity, the only thing that is prohibited is calling such a device IEEE compliant. One fieldbus spec that shall remain nameless took 802.3at, dropped LLDP, and made that their power system and the world did not stop spinning. If there is an industry that want to take advantage of the exact same low cost / volume production electronics that is used for Ethernet that is great, write a building power standard that uses electricians and not office workers for installation. To chuck out the installed base for a new market is not progress. Let each industry decide whether it is cheaper to use diodes, steering FETs or not claim compliance to the IEEE standard.