Are all interfaces to the ring optically coupled?

The place to discuss the LHC. Commissioning, operation, issues, events ....
Post Reply
jmc2000
Posts: 160
Joined: Sun May 08, 2011 2:51 pm

Are all interfaces to the ring optically coupled?

Post by jmc2000 » Fri May 27, 2011 1:18 pm

Does anyone know if all the interfaces to the LHC ring are optically coupled?

Because of the enormous changing magnetic field, I'd like to know how they deal with induced pick up. It appears that one of the PLCs had to be replaced that controls the cryogenics and perhaps a voltage spike killed it. I'm also surprised that the PLCs are not doubled up to offer some sort of redundancy in the case of a failure, given the major delay they cause as a consequence and their relative low cost.

Kasuha
Posts: 570
Joined: Tue Dec 15, 2009 1:22 pm

Re: Are all interfaces to the ring optically coupled?

Post by Kasuha » Fri May 27, 2011 2:03 pm

Instead of asking so many questions about LHC design, why don't you read how the LHC is designed yourself?
LHC design report is a bit aged and many things have been improved since then but I guess it's a good place to start with.
If you feel it's not detailed enough, you can also look into baseline documentation.

jmc2000
Posts: 160
Joined: Sun May 08, 2011 2:51 pm

Re: Are all interfaces to the ring optically coupled?

Post by jmc2000 » Fri May 27, 2011 5:43 pm

Kasuha wrote:Instead of asking so many questions about LHC design, why don't you read how the LHC is designed yourself?
LHC design report is a bit aged and many things have been improved since then but I guess it's a good place to start with.
If you feel it's not detailed enough, you can also look into baseline documentation.
Hey thanks!

pcatom
LHCPortal Guru
Posts: 68
Joined: Fri Nov 27, 2009 8:33 pm
Location: Geneva

Re: Are all interfaces to the ring optically coupled?

Post by pcatom » Sat May 28, 2011 8:36 pm

Does anyone know if all the interfaces to the LHC ring are optically coupled?

Because of the enormous changing magnetic field, I'd like to know how they deal with induced pick up. It appears that one of the PLCs had to be replaced that controls the cryogenics and perhaps a voltage spike killed it. I'm also surprised that the PLCs are not doubled up to offer some sort of redundancy in the case of a failure, given the major delay they cause as a consequence and their relative low cost.
In any big installation emc is an issue and had been carefully built into the design The magnetic fields of the machine are not the major problem - more mundane things like motors and pumps cause bigger headaches.

Redundancy is used wherever possible - but it is not always possible. The magnetic bearings of the cold compressors are delicate things and any interruption in the control of them could be a problem. Switching from 1 PLC to another seamlessly is not easy ...

The reason for the plc failure is not known (yet).

jmc2000
Posts: 160
Joined: Sun May 08, 2011 2:51 pm

Re: Are all interfaces to the ring optically coupled?

Post by jmc2000 » Sun May 29, 2011 8:43 pm

pcatom wrote:Redundancy is used wherever possible - but it is not always possible. The magnetic bearings of the cold compressors are delicate things and any interruption in the control of them could be a problem. Switching from 1 PLC to another seamlessly is not easy ...

The reason for the plc failure is not known (yet).
How expensive are Power converters? A power converter fault in S2 dumped today's beam so I can only think they're too expensive or unlikely to fail to duplicate.

adam_jeff
Posts: 49
Joined: Thu Nov 18, 2010 3:07 pm

Re: Are all interfaces to the ring optically coupled?

Post by adam_jeff » Mon May 30, 2011 9:31 am

As pcatom said, redundancy is not quite as simple as 'let's buy two of everything'. There is also the question of finding space for 2 systems, having suitable alarms to know when the primary system is failing, whether the backup can kick in fast enough to prevent a dump...

User avatar
morgad
Posts: 48
Joined: Sat Nov 21, 2009 8:43 am
Location: Devon, UK
Contact:

Re: Are all interfaces to the ring optically coupled?

Post by morgad » Mon May 30, 2011 8:54 pm

relative low cost ...

If i am reading the figures at the bottom of todays cyro report correctly. The cost of upgrading the 65 PLCs to get reducndancy is 1.5 Million swiss franks + 2 man-years of labour for cabling and code re-write.

best regards
Dave

jmc2000
Posts: 160
Joined: Sun May 08, 2011 2:51 pm

Re: Are all interfaces to the ring optically coupled?

Post by jmc2000 » Tue May 31, 2011 8:17 pm

morgad wrote:relative low cost ...

If i am reading the figures at the bottom of todays cyro report correctly. The cost of upgrading the 65 PLCs to get reducndancy is 1.5 Million swiss franks + 2 man-years of labour for cabling and code re-write.

best regards
Dave
[spluttering cough] 1.5 million SFs for material alone to add another 65PLCs?

According to the report, the MTBF is acceptable so they don't have to do anything. The report doesn't say anything about the total cost of failures in the PLC chain to the cryos over the total lifetime of the LHC.

User avatar
PhilG
Posts: 135
Joined: Sun Dec 06, 2009 9:58 pm
Contact:

Re: Are all interfaces to the ring optically coupled?

Post by PhilG » Wed Jun 01, 2011 5:27 am

How can you measure the cost of downtime? The LHC does not produce a product that can be given a value in dollars.The loss is in terms of scientific data.

In direct money terms the LHC probably saves loads of cash when it goes down because they save on electricity. That is a high cost greater than the costs of replacing most individual parts that fail.

In any case there are many ways the beams can be lost but they can quickly be brought back with a new peak luminosity. The overall loss should be minimal provided the dumps are not too frequent. When they are running at maximum luminosity without constant changes we will see how well it runs.

Since they are likely to beat their target luminosity by some factor this year it is not likely that they could make a case for spending extra money on better reliability.

jmc2000
Posts: 160
Joined: Sun May 08, 2011 2:51 pm

Re: Are all interfaces to the ring optically coupled?

Post by jmc2000 » Wed Jun 01, 2011 1:40 pm

PhilG wrote:How can you measure the cost of downtime?
You can't measure it exactly, but you can certainly approximate it and so it's worth doing just to see if there are losses and where they're going. Part is going to be in people's time being paid for less efficiently while the beam is down, which is going to depend upon the cause of the beam down.

Is it worth replacing a faulty PLC that causes the cryogenics to inoperate for over 24 hours every month?

Is it worth having a full time PLC programmer, or should they contract out?

Is it worth hiring a company to carry out a cost-benefit analysis of running costs against down time?

It's not possible to do anything exactly and approximations still make a task wothwhiile doing in the long run.

Post Reply