Re: current limiting of 802.3 power
Let me be clear, I am not lobbying to not have the source limited, I just want
to limit it with practical technologies.
The 350ma limit is from the transformers, the limit that has been discussed for
patch panels has been 250 ma per conductor. The patch panel number has been
considerably anecdotal, and I don't imagine we can really expect solid
What I was trying to say with the last e-mail is that I believe it is
impractical to expect the PSE to cut off at exactly 350ma. The exact limit can
be more easily enforced by the PD engineers at design time, they must ensure
that it will not draw more than 350ma on a sustained basis. The PSE is
responsible for cutting power if the PD faults (or something else faults) and
draw more than the legislated limit, but there has to be some slop above 350ma
to make it practical for the PSE to do so.
I don't know that we need to say the over current requirements can not be met by
PTCs at this stage of the game, at least I can not say I know enough to dismiss
them. PTCs are cheap and effective in my experience, but this project is way
beyond my limited experience in the power arena.
Finally, it is not safe to assume that the device that determines the minimum
current is flowing can also measure the current. Nortel discussed in LaJolla
using transformer saturation characteristics to know that at least a small
amount of current was flowing, but the system they discussed, I do not believe,
was well suited to measuring the actual current draw. An exact cut off may not
come for free.