• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Merged Puzzling results from CERN

Dr.Sid said:
So the question now remains how to utilize poorly inserted connectors in interstellar travel.
Click to expand...

I'm waiting for the first science fiction spaceship drive based on the "theory of loose connections" first discovered back in the 21st century.

Come to think of it it's already been done in a way - brownie points for identifying who is working on the space drive and has just one more connection to make!
 

Attachments

  • connectiondrive.jpg
    connectiondrive.jpg
    23.2 KB · Views: 486
http://news.sciencemag.org/sciencein...es-faster.html
After tightening the connection and then measuring the time it takes data to travel the length of the fiber, researchers found that the data arrive 60 nanoseconds earlier than assumed.
Click to expand...

The first question I'm sure a lot of very bright researchers would ask would be, "I thought you said you checked all the connections many times, didn't you?"
I am not very familiar with fiber optic connectors. Does anyone have an explanation for how a loose connection can cause "data [to arrive] 60 nanoseconds earlier than assumed?"
 
Olowkow said:
http://news.sciencemag.org/sciencein...es-faster.html


The first question I'm sure a lot of very bright researchers would ask would be, "I thought you said you checked all the connections many times, didn't you?"
I am not very familiar with fiber optic connectors. Does anyone have an explanation for how a loose connection can cause "data [to arrive] 60 nanoseconds earlier than assumed?"
Click to expand...

The speed of light varies depending on the medium it travels through. A gap between the end of the fiber cable and the pick-up in the conenctor means it travels through a bit of air first.
 
Hellbound said:
The speed of light varies depending on the medium it travels through. A gap between the end of the fiber cable and the pick-up in the conenctor means it travels through a bit of air first.
Click to expand...

Well that could hardly account for 60ns. Light will travel roughly 60 feet in that time. Air slows down light only marginally.
Interesting theory someone mentioned on BadAstronomy was that the device sometimes failed to transfer the data (due to losses or reflection) .. and simple resent the data again. Such protocol can act on pretty low level, unknown to user, and without any effect if the connection is fine.
 
Hellbound said:
The speed of light varies depending on the medium it travels through. A gap between the end of the fiber cable and the pick-up in the conenctor means it travels through a bit of air first.
Click to expand...

60ns is a lot of air though (something around 60feet, as air's a pretty good vacuum from the POV of refractive index).
ETA:and air's a much better vacuum than glass in that respect!

Speculating wildly, a dodgy fibre optic connector could cause the signal shape to degrade such that the sensor triggers later. Rather than a nice clean square wave popping out, you now have a sine-wave with an effective phase-delay.
 
Last edited:
Dr.Sid said:
Well that could hardly account for 60ns. Light will travel roughly 60 feet in that time. Air slows down light only marginally.
Interesting theory someone mentioned on BadAstronomy was that the device sometimes failed to transfer the data (due to losses or reflection) .. and simple resent the data again. Such protocol can act on pretty low level, unknown to user, and without any effect if the connection is fine.
Click to expand...

Well, yeah, I didn't really think it through..but it all comes down to differing speeds anyway.

So I still claim to be technically right, even if I was functionally wrong :p
 
All we can do is guess, given that we don't know much about the equipment they are using. I am sure it is quite complicated. I tend to think that a loose connection would cause no signal, and perhaps this signal is crucial to the timing. Maybe some AND gate is totally missing its trigger, and assumes a HI, causing everything to be off by 60 ns.

I'd have loved to see Feynman sit down with the schematics of the experiment, take a sip of coffee, and say "Uhm, check connector on cable S3488/Q5677." Tech, checks it, and ..."OMG!, it's loose!":) 20 minutes later, up and running.
 
Olowkow said:
I'd have loved to see Feynman sit down with the schematics of the experiment, take a sip of coffee, and say "Uhm, check connector on cable S3488/Q5677." Tech, checks it, and ..."OMG!, it's loose!":) 20 minutes later, up and running.
Click to expand...

If I'm remembering 'surely you're joking, Mr Feynman' correctly, to paraphrase he'd not understand some symbol in the diagram, point at it and ask 'what about that?', expecting an answer of the form 'The Frobditz is needed because of Quuxon variation', but get 'ooo, we'd not noticed that before'.
 
nathan said:
If I'm remembering 'surely you're joking, Mr Feynman' correctly, to paraphrase he'd not understand some symbol in the diagram, point at it and ask 'what about that?', expecting an answer of the form 'The Frobditz is needed because of Quuxon variation', but get 'ooo, we'd not noticed that before'.
Click to expand...

What a guy.:)
 
nathan said:
If I'm remembering 'surely you're joking, Mr Feynman' correctly, to paraphrase he'd not understand some symbol in the diagram, point at it and ask 'what about that?', expecting an answer of the form 'The Frobditz is needed because of Quuxon variation', but get 'ooo, we'd not noticed that before'.
Click to expand...

This is from memory, so forgive me if some details are wrong ...

He was asked to review the design for the processing of uranium. A drawing was laid on the table in front of him. Not wanting to reveal to the others that he didn't know what the symbols on the drawing meant, he decided to get them to explain it to them. He pointed at something that he guessed was a valve and asked something like "what happens if this valve fails?"

That lead to a discussion of the effects if the valve failed and the belief by the others that Feynman had understood the process and recognized a problem. :-)

-- Roger
 
Last edited:
Darat said:
So all this could have been avoided by phoning any half decent IT support person, bloody hell the first thing I check if something doesn't work is "Is it plugged in?".
Click to expand...

I channeled "The IT Crowd" back in post 782 to envision how such a phone call would sound. :-)

-- Roger
 
nathan said:
Speculating wildly, a dodgy fibre optic connector could cause the signal shape to degrade such that the sensor triggers later. Rather than a nice clean square wave popping out, you now have a sine-wave with an effective phase-delay.
Click to expand...

That's been my only guess, but I'm skeptical that a fiber-optic signal from a timing unit would have such a sloooooooow rise time.
 
ben m said:
That's been my only guess, but I'm skeptical that a fiber-optic signal from a timing unit would have such a sloooooooow rise time.
Click to expand...

Well, not knowing the actual hardware it's rather hard to guess. Having worked with photodiodes myself, i know that they are not linear. Any drop in incoming intensity can (and probably will) have strange results.

Plus, the following signal conditioning circuitry is important as well. Being a few ns off at the physical receiver can cause additional delays in the conditioning circuitry, depending on design of the same.

Also keep in mind that a lot of things are designed with fixed expectations in mind when it comes to signal integrity and timing. If, for example, they use a PLL to recover the clocking, having a few bits constantly skewed can already cause some hickup in the PLL, depending on design.

60 ns is a lot or very little, depending on how you look at it. Rise and fall times at digital gates can vary wildly with temperature, signal level, tolerances, etc.

At least, from what experience i have personally with fiber-optical links, i find it plausible that such a link may introduce a 60ns error in timing.

Greetings,

Chris
 
Last edited:
jadebox said:
This is from memory, so forgive me if some details are wrong ...

He was asked to review the design for the processing of uranium. A drawing was laid on the table in front of him. Not wanting to reveal to the others that he didn't know what the symbols on the drawing meant, he decided to get them to explain it to them. He pointed at something that he guessed was a valve and asked something like "what happens if this valve fails?"

That lead to a discussion of the effects if the valve failed and the belief by the others that Feynman had understood the process and recognized a problem. :-)

-- Roger
Click to expand...

They were looking at the blueprints for the plant trying to find the source of a problem. In the plant, all valves had to be in pairs along sections of tubing. Feynman was looking at the blueprint and wondering if the symbol he was looking at was a window or a valve. That's how little he knew about how to read a blueprint. But upon realizing there were far too many of those symbols to be windows, he decided to confirm that it was in fact a valve by pointing at one at random and saying, "What happens if this valve fails?"

The other's looked at it and were amazed to find that this valve he pointed out did not have a matching partner. They thought he had solved the problem by merely glancing at the blueprint.
 
They thought he had solved the problem by merely glancing at the blueprint.
Click to expand...
I'm sure we have all been in this position...and we just quietly smile. Speculation of course, but this equipment probably has hundreds of fiber optic in/out cables and is significantly more complicated than merely a detector that can be delayed by a bad connection. I'm thinking one of hundreds of crucial signals was just missing, and consequently introduced the delay. The reason I suspect this is because one of the articles stated that they would have to confirm this as the solution by testing in several months. Why could they not just deduce that this was the problem? It must be very complicated from a troubleshooting point of view. Just my opinion.
 
You must log in or register to reply here.

Back
Top Bottom