Slashdot Mirror
CERN Engineers Have To Identify and Disconnect 9,000 Obsolete Cables (vice.com)
An anonymous reader writes: CERN, home to the Large Hadron Collider, has grand plans to update the world's largest particle accelerator complex in the next few years. But engineers have identified a barrier to the upgrade: there's no space for new cables in the injectors that accelerate particles before they enter the LHC. In the past, when parts of the accelerators have been upgraded or added to, engineers would often additionally replace the cables that connected them. In the process, they would leave in place the old cables that were no longer in use. Now, a heap of obsolete cables are blocking the way to install new ones needed for the accelerator’s next big upgrade. To make space, CERN engineers have set out to identify and remove the old, unused cables. All 9,000 of them.
Sadly, I was thinking the same thing... might be time to get started doing that before I also have 9000 cables to identify and remove... :P
9000 cable... and no labeling
Sadly this is common...
As a lab manager I had to institute a rule that ANY cable that didn't have a label was going to be removed when found with no warning. Any cable which was incorrectly labeled, was subject to be connected to what the label said, or if that wasn't possible, the label would be removed and then the cable was pulled for not being labeled. Label content was defined and all where trained on how to make proper labels, and retrained when they came to ask why their system suddenly stopped working.
Maintaining a lab is a daily discipline, like cleaning house. You have to pick up after yourself as you go along or at the end of the day the mess is huge. Hey, where you born in a barn? Your Mom doesn't work here, clean up your mess!
"File to fit, pound to insert, paint to match" - Aircraft Maintenance 101
One of the few silver linings of the Hurricane Sandy damage: they finally pulled tons of old copper out of the tunnels and cable-runs and replaced it with fiber, because there was finally no way to be sure which was obsolete and which was current-but-damaged.
Confirmed: http://moodybluesattitude.yuku...
I don't know for sure but I bet this was part of a penny pinching cost analysis up front.
I recall when moving to a new site setting aside some time/budget to ensure that every cable was labelled (so, for example, we could trace ethernet from port on switch to patch panel to underfloor cable to floor jack to desk cabling to desk port) and set up a simple database to keep the details.
Work was killed off by accountants as an expensive luxury, after all cables didn't move often did they?
Fast forward to a minor flood under the false floor taking out some (but not all) systems. Fortunately some of them were in the finance and commercial group.
Suddenly it was "why can't you reconnect me NOW??". Money was paid for an 'after the event' recording of wiring by external people (which cost about 5 times the 'saving' up front).
Still at least it was better than a LONG time ago [Vax and VT220 era] when I saw one person labelling connections by yanking out an RS232 cable from a patch panel, waiting for a call "My terminal's died", asking which room they were in and making up a label and then plugging it back with "I think that may fix it" and getting pathetically grateful responses in return.
I am reminded of the days of wire-wrap circuit boards. hunders of wires in a few colors at most forming a rats nest of interconnects on the back. All done by hand from post to post where you had to count pins by eye to find the right post each time. Chance of 100% correct wiring was geometrically vanishing.
The problem was not discovering the connections you had failed to make (which is easily done with a continuity tester) but finding the connections that were mistakenly wired the wrong pins.
So what you did was go find a filament transformer (these were high current low voltage transformers used to power the filaments in tubes). then you put one probe on one pin, and another probe on every other pin it was not supposed to be connected to. This is not as complex as it sounds since normally one pin is not connected to more than 3 or 5 other pins. So once you eliminate those, you can just slide the probe along the sides of all the other socket pins.
The current was so large that even a momentary connection would vaporize the wire if it was incorrectly wired. A continuity tester would not have worked well because the response time for the human to test all N^2 connections and look at the continuity tester was too long.
Some drink at the fountain of knowledge. Others just gargle.