Please read: "No, German Scientists Have Not Confirmed the “Impossible” EMDrive"
http://io9.com/no-german-scientists-have-not-confirmed-the-i... (HN:
https://news.ycombinator.com/item?id=9973528 (5 points, 5 hours ago, 1 comment))
My abstract: It's very strange that the thrust don't appear almost instantly when you turn on the device and disappear almost instantly when you turn off the device. It's Is suspiciously similar to what you would expect from a measurement error due to thermal effects.
The only comment in the HN submission is mine, and I'll copy it (with a few modifications) because it's (more) relevant here.
Some numbers to compare:
The German experiment got 20uN with a 700W magneton, that is 20uN/700W = 0.03 mN/KW. (Probably some energy is "lost" and the "real" efficiency is higher, but when you claim to break the physics law the measurements must be foolproof.)
The maximal theoretical output of a device that don't break the currently accepted physics laws is 1/c = 0.0033 mN/KW. So the German measurement is 10x bigger than the theoretical maximum.
The 18-month-to-Pluto and 4-hous-to-Moon calculation use 0.4 N/KW = 400 mN/KW, that is approximately one half of the maximum claim of any of this device family, but most experiments get much smaller results. The 0.4 N/KW is 130,000x bigger than the theoretical maximum and 13,000x bigger than what they got in the German experiment. So if a device with the same output of this experiment, the trip to Pluto or the Moon will be much longer.
(Disclaimer: Just to be clear, I think this is only an experimental error.)
