OPERA grabs spanner, fixes kit, and slows down neutrinos
Ye canna break the laws of physics, it seems
According to a breaking report from Science, last year’s famous faster-than-light neutrino finding has been attributed to a cable fault.
The report says insiders at the OPERA collaboration have found that “a bad connection between a GPS unit and a computer may be to blame”.
The experiment last year seemed to identify neutrinos completing their trip between CERN at Geneva and Gran Sasso 60 nanoseconds too quickly. This result, apparently at odds with Einstein’t oft-confirmed Special Theory of Relativity (which sets the speed of light in a vacuum as a universal maximum), sparked excitement and skepticism around the world in equal measure.
It also set the physics section of the Arxiv open-access scientific database alight, with dozens of papers re-analysing the results or, alternatively, seeking explanations.
Back in OPERA, it seems, scientists and techs were also paying attention to more mundane matters, including the equipment used to make the measurements.
Here, if today’s report is correct, we find the very mundane explanation for the measurements: the poorly-terminated cable between the computer and GPS receiver introduced a signal delay.
After tightening the connection (in other words, El Reg would speculate, eliminating a tiny gap between the end of the fibre and the sensor to which it terminated), OPERA researchers have found that the GPS data arrived 60 nanoseconds earlier than was measured during the original experiment.
"More data will be needed to confirm this hypothesis", says the report. ®
Not only are its neutrinos faster than everyone else's, but it was also the first large hadron collider to use tabs and restart exactly where it left off after a magnet quench.
There *is* something to see.
...and that's good science. As a couple others have pointed out, a strange result got investigated, checked, re-checked, and now apparently discarded as erroneous. This is science working as designed and scientists setting a good example.
That can only be a good thing given all the bad examples out there. There are plenty of fields where the politics trumps the science. AGW is the obvious target around here -- on *both* sides -- but reproductive health is one I care more about. Physics seems to have escaped the problem so far, possibly because the public doesn't understand it well enough for it to be fuel for demagoguery.
Damn, now I have this mental image of some weasel going on about how the Light Sympathizers' refusal to believe in FTL neutrinos is holding us all back from FTL travel...and another weasel claiming that the Neutrino Sympathizers just want to open a wormhole to the lair of Azathoth....
(wow, spellchecker claims I spelled Demagoguery right on the first try.)
Re: Tightening a connector?
The way things like this work is:
The GPS receiver outputs a pulse once a second (the 1PPS output).
You have a very stable oscillator at a high frequency, such as a cesium or rubidium oscillator. These types of devices can be "tweaked" a bit to run slow or fast, but once tweaked, will be very stable. That gives you a very fast clock (sub-nanosecond per click), that won't change in rate unless you tweak it.
Now, you need to tweak that clock to match the clock at the far end. You do that via GPS like this:
You make a 1PPS signal from your clock. You compare that with the GPS 1PPS signal - FOR A VERY LONG TIME (like, days), averaging the time difference between the pulses from your clock and from GPS. Any single pulse may show a lot of jitter (random noise-like errors), but if you average them over long enough you can remove the noise and get a good measurement of how they are drifting. You use that very averaged signal to make very small tweaks to your clock. Eventually, your 2 clocks are in sync. You can then use the very high frequency output of them to make very precise time measurements.
Now, if that 1PPS signal cable from the GPS is a bit loose, then the rise time of the pulse will be slower, and the pulse will act like it is delayed by the slower rise time. Thus, you will sync your clock with a built-in delay due to that slower rise time.
(If you want more info, look up NTP (network time protocol) and PTP (precise time protocol, a.k.a. IEEE-1588) and that will go into more detail about getting things synced up.)