I need to get a signal strength measurement on an 850nm gigabit ethernet signal without interfering with the signal. I can get a splitter, but I'm not clear on what percentages are appropriate for the split. I need to split off enough signal to have a measurable reading on something that is probably a very low power signal to begin with. At the same time, I cannot steal so much of the signal that I affect the integrity of the signal.
Probably a 10% / 90% split is too conservative. Does anyone with experience using splitters in a gigabit ethernet environment have an opinion on a best split percentage to use?