60 ns delay in a faulty connection - how?

[Borek]

Yes, it is about OPERA experiment, but it has nothing to do with neutrinos.

Can someone try to explain how is it technically possible that faulty connection could be responsible for additional signal delay? I understand it can make the signal unreadable, but delayed by 60 ns?

Main reason I can't understand the situation is that even if the link is faulty, signal still propagates with the same speed - approximately 1 feet per 1 ns. 60 ns is equivalent 60 feet. Signal path didn't change that much. What is the source of the delay? What am I missing?

 

[yungman]

Is it digital signal? If so, it might be possible that the faulty connection like a high impedance connection cause the rise time to slow way down and take a long time for signal to past the threshold to trigger the next stage. You'll see it as delay.

 

[f95toli]

They are not actually sending a signal in the experiment. They "time stamp" the neutrinos at the source (i.e. they keep a record of when the pulse was sent), and time-stamp them again when they arrives at the detector.
Hence, the issue is not about delays; it is about whether or not they've managed to synchronize their clocks well enough, and if they understand all the delays involved in sending/detecting the neutrinos.

 

[maimonides]

FWIW: This comment on Cosmic Variance is the only hint at a reasonable explanation of the "faulty cable" I´ve seen so far.
http://blogs.discovermagazine.com/cosmicvariance/2012/02/22/neutrinos-and-cables/#comment-225749

 

[Borek]

Is it digital signal? If so, it might be possible that the faulty connection like a high impedance connection cause the rise time to slow way down and take a long time for signal to past the threshold to trigger the next stage. You'll see it as delay.

Thanks, now I see a possible physics behind, that's what I was missing.

Hence, the issue is not about delays; it is about whether or not they've managed to synchronize their clocks well enough, and if they understand all the delays involved in sending/detecting the neutrinos.

If I understand the situation correctly they used GPS to synchronize the clocks, so these lost 60 ns will mean clocks are not properly synchronized. I believe we say the same thing.

FWIW: This comment on Cosmic Variance is the only hint at a reasonable explanation of the "faulty cable" I´ve seen so far.

At least it means I am not the only person wondering about it.

 

[f95toli]

If I understand the situation correctly they used GPS to synchronize the clocks, so these lost 60 ns will mean clocks are not properly synchronized. I believe we say the same thing.

Yes, they use GPS. But the synchronization is not in "real time",which is what I thought you meant, there is no delay "compared to something" since they are not comparing two signals.
The issue is not so much the clocks themselves. it is that the total delay they have to account for in the timing is much longer than 60ns and if they've made a mistake when e.g. measuring the delay that comes from the signal traveling from the GPS antenna down into the tunnel this could -potentially- cause quite a large delay; although it is very unlikely that they would make such a huge mistake. Unless they've made a serious mistake somewhere their clocks should be synchronized to within 1 ns.

Another -in my view more likely -explanation is that there is an error in timing when it comes to generation/detection of the neutrinos; there is an awful lot of electronics between the detectors themselves and the clocks that time-stamp the data and that delay (which has to be added/subtracted from the time-stamps at each end) can be quite tricky to measure.