Lighting for the CERN project

[treez]

https://en.wikipedia.org/wiki/CERN

...we all know that CERN is a huge underground radiation lab.
Radiation used at CERN will damage electronics.
Lighting is needed all throughout the radiation labs.
The lighting must not therefore have electronics in it....therefore, would you say it would be best to use a 50KHz mains wiring distribution system inside CERN, and then couple into it with ferrite couplers, which have no electronics in them....to the LED bulbs?..there would obviously be electronics insde the 50KHz current source gernerators, but you could simply put those in lead boxes.
Is this how lighting is done inside CERN?. Is this how it should be done?

 

[Dan Mills]

A simple florescent tube with a magnetic ballast and thermal switch starter is about as rad hard as anything gets (Same with magnetic ballast mercury arc or metal halide).....

I would be very surprised if small tunnel lighting was not florescent with cavern lighting for the detectors being metal halide or similar.

LED diode junctions are not normally considered to be radiation hardened.

You are overthinking the thing.

Regards, Dan.

 

[treez]

you mean led lights are less radiation proof than fluorescent?
Are power leds as sensitive to radiation damage as are eg opamps, microcontrollers etc?
LEDs are more efficient than magnetic ballast fluorescent...mag ballast fluorescent is not an option.

 

[Dan Mills]

Yep, semiconductor junctions are less radiation resistant then a simple glass tube with filaments and phosphor.

Power LEDs tend to be large geometry so are probably not massively rad sensitive, but they will still degrade, and that is before you consider the plastics used for encapsulation.

Why would magnetic ballast not be an option? It is simple, robust and has very well established reliability at this point.
The energy cost of lighting the underground facility while it is shut down for work is negligable compared to the energy cost of running the machine.

Now granted you might have to watch bringing a steel cored inductor too close to the bending magnets, but that is a managable issue.

LED is a useful tool in the lighting engineers kit, but it should not be the only one.

Further, remember that the LHC was fitted out ~10 years back and while Metal Halide, florescent and mercury discharge were established reliable technology then, Power LED lighting had a spotty track record at best (Mainly people failing to properly manage the thermal issues).

Even if you wanted to do LED in an envoronment like that (And it sems to me to be the wrong tool for the job) I would not be looking at 50KHz, far better to take a leaf out of airport runway lighting and do a constant current supply with series connected LEDs, put the supply gear on the surface and run a constant current loop down the tunnels in pyro or similar cable.

Edit: Just found some pictures of the beam line, and the tunnel is clearly lit with florecent luminares.
Looks like the CMS cavern is a mix of metal halide and high pressure sodium, about what you would expect in a large industrial space.

Regards, Dan.

 

[Dominik Przyborowski]

Inside the main tunnel a standard fluorescent lamps are used. In fact the radiation is high only inside experiments complex. The whole hadrons and EM particles died in calorimeters and only muons go outside. In addition the synchrotron radiation caused by 7 TeV protons circled in 27km pipe isn't so high.
Of course the dose inside a tunnel will killed each human being.

 

[treez]

thanks, though in other environments where lighting is required and radiation levels are hig, the system I speakof will surely be of use?....
such as isotera.com system

 

[Dan Mills]

Why, it sounds like a pain in the arse, last thing anybody wants around sensitive physics experiments is hundreds of watts of 50KHz (Plus I would suspect harmonics) power on an unshielded twisted pair, it will get into everything....
Further it is more electronics to go wrong, seriously KISS applies in spades in such facilities.

Oh and a simple magnetic ballast metal halide fixture has about the same order of efficiency as a white LED while being far easier to cool and available from any electrical wholesaler.

I was figuring the bulk of the radiation in the tunnels would be Xrays from the interaction of the H+ beam and the steering magnets (Possibly also the quadrupoles and beam cooling systems), plus a very little other stuff where a mean free path turns out to not quite be long enough, none of which seems likely to be a problem for simple lighting fixtures, the ash from the experiments themselves however I was unsure about, I could maybe see some gamma?

The place such a thing might be interesting, would be an Intrinsically Safe version for use in petrochemical and mining where having the ability to connect a hand lamp or such without any possibility of a spark has some virtue (The driver would need to be current limited however).

Regards, Dan.

 

[FvM]

The place such a thing might be interesting, would be an Intrinsically Safe version for use in petrochemical and mining where having the ability to connect a hand lamp or such without any possibility of a spark has some virtue (The driver would need to be current limited however).

Would be... if you manage to prove it's safety under arbitrary load conditions. This probably turns out more difficult than for a wired power supply with same power.

The intrinsic safety point has been already discussed in previous isotera "fan blog" thread at Edaboard.

 

[Dan Mills]

Granted, getting gear certified as IS is a well known nightmare.

Limit total power and design the couplers to have a sufficiently high leakage inductance such that they inherently limit the power that can be extracted from the field.
In fact if you used transformer coupling at the driver end as well, then that could use the same trick to limit the input power and a earth fault detector could trip things off if either leg got shortd to ground, not there yet by a long way, but the notion starts to have possibiilities.

I do not see the point in a situation where something like an intrinsic safety requirement is not forcing the issue however (But I am well known for a liking for things to be as simple as possible).

Regards, Dan.