Dances with wolves I mean Pairs!
Hi Dan and Larry
I repeat that we agree with you- that using the "unused" pairs for mid span
insertion sounds more conservative and may be a better idea.
at the same time, I look at the signal pair for a new switch, and say.
there is a good chance that I will be able to place all the circuity inside
the phy, possibly come up with a way to use very little silicon to
do current management out of the center tap. and I am done.
what I mean is (what we called switches, and filter can go inside
the phy) now, the detection is done without any added parts on the
board, not only that, a bonus to this is 2000v isolation and better
transient performance, I think as good as it can be .
at this point, I don't think that I can have all the data I need to prove
it all, but we need to know more about your schemes so we can judge
how solid they are. You can Impress too!- you know.
for instance, I am curious to see how Dan's approach of using a diode
in the DTE will deal with transients, how low would the threshold of the
Transient detection device have to be to protect a diode. in there
where our approach is the classic magnetic differential signal kinda
of simple but will always get you there.