1 / 57

1v4

1v4. 1. Case Study: The Beam Interlock System Electrical and Physical Architecture Component Locations Design for Durability Scope of the Tests (CIBD and CIBU). 2. Power Supply Tests and Results Detailed Description of the Power Supply Test methods Test Results.

wan
Download Presentation

1v4

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. 1v4

  2. 1. Case Study: The Beam Interlock System • Electrical and Physical Architecture • Component Locations • Design for Durability • Scope of the Tests (CIBD and CIBU) 2. Power Supply Tests and Results • Detailed Description of the Power Supply • Test methods • Test Results 3. User Interface Tests and Results • Detailed Description of the User Interface • Test methods • Test Results 4. PLC Interlock Systems and Conclusions • Powering / Warm Magnet Interlocks • Conclusions TE/MPE/MI - CERN Machine Interlocks

  3. 1. Case Study: The Beam Interlock System • Electrical and Physical Architecture • Component Locations • Design for Durability • Scope of the Tests (CIBD and CIBU) 2. Power Supply Tests and Results • Detailed Description of the Power Supply • Test methods • Test Results 3. User Interface Tests and Results • Detailed Description of the User Interface • Test methods • Test Results 4. PLC Interlock Systems and Conclusions • Powering / Warm Magnet Interlocks • Conclusions TE/MPE/MI - CERN Machine Interlocks

  4. Beam Interlock System Locations Designed to protect CERN high energy accelerators = SPS / LHC / INJ / EXT TE/MPE/MI - CERN Machine Interlocks

  5. LHC Function BIS ~200 connections to User Systems distributed over 28kms Both-Beam LHC has 2 Beams Some User Systems give simultaneous permit Others give independent permit Beam-1 Beam-2 TE/MPE/MI - CERN Machine Interlocks

  6. Signals Safe = FMECA Redundant Available = Power Supplies TE/MPE/MI - CERN Machine Interlocks

  7. Electrical Architecture USER SYSTEM rack “some radiation” Worst “UJxx”-type In INTERLOCK rack “negligible radiation” Typical background at Sea Level Test / Repair before each Mission - accessible 2U Chassis VMEChassis TE/MPE/MI - CERN Machine Interlocks

  8. Dependable Design • Critical Function separate (fewer critical circuits) • Failure Modes Analysis (weaknesses designed out) • Xilinx CPLD used elsewhere (EPC / PIC) • 5V and redundant PSU (Highly Available) • Mature Electronics (Not risky!) • Electro-Magnetic Compatibility (key issue from start) • Internally Reviewed 2006 • External Review in late 2009 • = robust designreliability and rad tolerance seem linked! • -running ½ LHC system in SPS for nearly 4 years  TE/MPE/MI - CERN Machine Interlocks

  9. Motivations for the Tests The Beam Interlock Controller (x17 in LHC) -few in surface buildings -most in zones considered ‘negligible radiation’ -one in control room -VME Chassis can be moved to surface if ever there is a problem -any problem will result in safe failure by design VME is the basis = Not going to Test The User Interface (x130 in LHC) -Goes where the User System is “UJxx” We have to Test The User Interface Power Supply (x260 in LHC) -Goes where the User Interface is “UJxx” We have to Test TE/MPE/MI - CERN Machine Interlocks

  10. Typical Hardware User Interface BIC (Front) TT40 BIC (Rear) TT40 TE/MPE/MI - CERN Machine Interlocks

  11. 1. Case Study: The Beam Interlock System • Electrical and Physical Architecture • Component Locations • Design for Durability • Scope of the Tests (CIBD and CIBU) 2. Power Supply Tests and Results • Detailed Description of the Power Supply • Test methods • Test Results 3. User Interface Tests and Results • Detailed Description of the User Interface • Test methods • Test Results 4. PLC Interlock Systems and Conclusions • Powering / Warm Magnet Interlocks • Conclusions TE/MPE/MI - CERN Machine Interlocks

  12. CIBD Specification Tracopower TXL-025-25S 5V, 5A, 25W • Device housed in shielded conductive case • Reliability analysis = redundant PSU = good Availability • Decoupling diode added • +3% trim to accommodate Vdiode • 400mA Fuse and filter added on input • EMC Tests done with CIBU/CIBD configuration • Very good results • https://edms.cern.ch/document/762174/ • Not the cheapest • excellent reliability = better than predicted • 4 failures to date = 10x better than datasheet TE/MPE/MI - CERN Machine Interlocks

  13. CIBD – 5A 5V PSU TE/MPE/MI - CERN Machine Interlocks

  14. CIBD – 5A 5V PSU ~12cm TE/MPE/MI - CERN Machine Interlocks

  15. CIBD – 5A 5V PSU PROSPERO 235U core and a 238U reflector NEUTRONS Energy = 1MeV Fluence = 1E12 n/cm2 =Several 10’s Years LHC <1% Voltage change And CIBU OK down to 4.5V (guaranteed by design) PSU well protected - PSU 25W, we need <5W Degradation is expected & planned for TE/MPE/MI - CERN Machine Interlocks

  16. 1. Case Study: The Beam Interlock System • Electrical and Physical Architecture • Component Locations • Design for Durability • Scope of the Tests (CIBD and CIBU) 2. Power Supply Tests and Results • Detailed Description of the Power Supply • Test methods • Test Results 3. User Interface Tests and Results • Detailed Description of the User Interface • Test methods • Test Results 4. PLC Interlock Systems and Conclusions • Powering / Warm Magnet Interlocks • Conclusions TE/MPE/MI - CERN Machine Interlocks

  17. User Interface Specification Device housed in conductive case with Burndy Connectors -great EMC -good reliability from connectors (many others already use these) Completely 5V -good reliability (more resistance to perturbation) Classic 7400 Series -mature technology -wide support Current Loop inputs -accommodate many different types of User System -DESY input circuit (M. Werner) RS485 Outputs -good integrity -mature Technology 9500 Xilinx CPLD for TEST and MONITOR -Manchester frames (Used General Machine Timing as a basis) -excellent dependability shown so far -not susceptible to program corruption 100% remotely testable 100% remotely monitorable TE/MPE/MI - CERN Machine Interlocks

  18. User Interface TE/MPE/MI - CERN Machine Interlocks

  19. User Interface Beam-2 Beam-1 TE/MPE/MI - CERN Machine Interlocks

  20. Test Areas TE/MPE/MI - CERN Machine Interlocks

  21. Test Areas Beam-2 Beam-1 TE/MPE/MI - CERN Machine Interlocks

  22. Test Areas TE/MPE/MI - CERN Machine Interlocks

  23. Test Areas Beam-2 Permit A Beam-1 Permit A Beam-2 Permit B Beam-1 Permit B TE/MPE/MI - CERN Machine Interlocks

  24. Test Areas All use the SAME electronic circuit! It’s repeated four times Beam-2 Permit A Beam-1 Permit A Beam-2 Permit B Beam-1 Permit B TE/MPE/MI - CERN Machine Interlocks

  25. USER PERMIT Paths The mission critical function for the CIBU; -relay USER_PERMITcorrectly Nothing else! All the BLACK lines are critical! TE/MPE/MI - CERN Machine Interlocks

  26. USER PERMIT Paths The mission critical function for the CIBU; -relay USER_PERMITcorrectly TE/MPE/MI - CERN Machine Interlocks

  27. The Irradiated Zones • USER_PERMIT Optocoupler • USER PERMIT Current Regulation Circuit • USER PERMIT RS485 Driver • USER PERMIT Schmidt Trigger • The WHOLE CIBU Test 3 Test 1 Test 4 Test 2 TE/MPE/MI - CERN Machine Interlocks

  28. Irradiation Parameters Irradiation Parameters 60MeV Protons – OPTIS at PSI 15MeV/cm2/mg in SiO2 – can cause SEE in plastic packages N.B. 1 x 1011protons cm-2 equivalent to 140 Gy “Representative of LHC Environment” – F. Faccio, M. Muhtinen TE/MPE/MI - CERN Machine Interlocks

  29. The Irradiated Zones TE/MPE/MI - CERN Machine Interlocks

  30. Test 1 – The Optocoupler Received a TOTAL DOSE of 62 Gy = 4.8 x 1010 pcm-2 10 Glitches observed – observation window open only ¼ of test - 40 assumed. Cross Section = 8.3 x 10-10for Optocoupler Glitches TE/MPE/MI - CERN Machine Interlocks

  31. Test 1 – The Optocoupler usual transition times were observed during the test ~3.6us delay for TRUE to FALSE ~0.6us delay for FALSE to TRUE TE/MPE/MI - CERN Machine Interlocks

  32. Test 2 – The Current Regulator Received a TOTAL DOSE of about 120 Gy NO Glitches observed Current regulation drifted by about 4% TE/MPE/MI - CERN Machine Interlocks

  33. Test 3 – RS485 Driver Received a TOTAL DOSE of 496 Gy NO Glitches observed, No FALSE Transitions FAILED SAFE at 496 Gy, power cycle restored operation TE/MPE/MI - CERN Machine Interlocks

  34. Test 4 – Schmidt Trigger Received a TOTAL DOSE of 1000 Gy (then experiment halted) NO Glitches observed, NO FALSE transitions TE/MPE/MI - CERN Machine Interlocks

  35. Test 5 – Whole PCB Two PCBs tested ~140-160Gy - CPLD stopped sending data frames TE/MPE/MI - CERN Machine Interlocks

  36. Overall Current Consumption Three PCBs tested at various points User Interface ID 1 = 180 Gy TE/MPE/MI - CERN Machine Interlocks

  37. Overall Current Consumption Three PCBs tested at various points User Interface ID 682 = 1650 Gy TE/MPE/MI - CERN Machine Interlocks

  38. Overall Current Consumption Three PCBs tested at various points User Interface ID 2 = 160 Gy TE/MPE/MI - CERN Machine Interlocks

  39. Beam Interlocks - Conclusions • Critical Path of Beam Interlock System • - 8.3 x 10-10 Cross-Section for Glitches = use 1.6us GLITCH FILTER! • Good radiation tolerance • CPLD failed first at 150Gy – only for MONITORING • BUT once we got it back here – reprogrammed it and it worked! • Will monitor Glitch counters to indicate potential problems • Had plans for RAD HARD CIBU but on hold • Needs motivation, time and money! ($) • VME is weak? move BICs to ‘surface’ ($$$) • Critical Matrix CPLD is same as in CIBU (3.3V variant) • VME PSU – redundant! • Can research a RAD HARD BIC (not including VME) • Needs motivation, time and lots of money ($$$$$) TE/MPE/MI - CERN Machine Interlocks

  40. 1. Case Study: The Beam Interlock System • Electrical and Physical Architecture • Component Locations • Design for Durability • Scope of the Tests (CIBD and CIBU) 2. Power Supply Tests and Results • Detailed Description of the Power Supply • Test methods • Test Results 3. User Interface Tests and Results • Detailed Description of the User Interface • Test methods • Test Results 4. PLC Interlock Systems and Conclusions • Powering / Warm Magnet Interlocks • Conclusions TE/MPE/MI - CERN Machine Interlocks

  41. PLC Systems Multitude ofPowering Interlock Systemsdeployed in CERN accelerators for protection of nc (WIC) and sc magnets (PIC) - Installed in LHC, SPS, TI2/TI8, CNGS, LEIR, Booster, Linac,… - based on a variety of industrial PLCs /IOs and additional electronics • Depending on the specific requirements, use of faster SIEMENS 319series or slower (safety) series 300 • By default, PLCs located in locations considered ‘safe’, generic PLC code • Fail safe design, for increased safety most designs incorporate a redundant path using different technology • Remote I/O modules collect interlock signals close to clients (magnets, power converters, QPS) and transmit via Profibus to PLC • For all components located in / close to radiation environments, dedicated radiation test campaigns conducted to validate TE/MPE/MI - CERN Machine Interlocks

  42. Example - Warm magnet Interlock Controller (WIC) Warm magnet Interlock Controller (WIC) Via Ethernet PVSS Supervision & local Touch-screen Power Converter Status info Power Permit Surface / Tunnel area considered “safe“ Tunnel / Radiation environment Profibus Remote I/O crate Thermoswitches Water Flow Red button… Several thermo-switches @ 60°C Magnet 1 Magnet 2 TE/MPE/MI - CERN Machine Interlocks

  43. PLC Systems – Radiation Test campaigns TCC2 TT60 MDSV 660106 Test Rack Many tests on IO and PSU in many places since 2002 In close collaboration with T. Wijnands Results shown at last RADWG workshop Additional Tests in CNGS this year Villingen Louvain TE/MPE/MI - CERN Machine Interlocks

  44. Closing remarks … Rules of Thumb … • Use simple remote logic, keep the complicated electronics away from the radiation • Use Fail-Safe everywhere • safety systems = redundant signal paths, • perhaps less availability = but more dependability • Make equipment auto-restarting • adequate monitoring/diagnostic tools • Implement remote reset (soft vs hard) • dedicated radiation tests of critical paths • Record batch & lot numbers of critical path devices • Use past experience wherever possible = proven components • CNGS this year = some new tests “Better the devil you know” TE/MPE/MI - CERN Machine Interlocks

  45. FIN TE/MPE/MI - CERN Machine Interlocks

  46. CIBU : User Interface Locations IR1 IR2 IR3 IR4 IR5 IR6 IR7 IR8 other SR1 SR2 SR3 SR4 SR5 SR6 SR7 UA83 CCR US151 UA23 UJ33 UA43 UJ56 UA63 UJ76 UA87 USA151 UA27 UA47 USC55 UA67 TZ76 UX85 USA152 SX4 RR53 US65 RR73 US851 RR13 CR4 RR57 US651 RR77 RR17 US451 UJ14 UJ16 “critical areas” BIC : Beam Interlock Controller Locations IR1 IR2 IR3 IR4 IR5 IR6 IR7 IR8 other US15 SR2 SR3 UA43 UJ56 UA63 SR7 SR8 CCC UA23 UJ33 UA47 UCS55 UA67 TZ76 UA83 UA27 UA87

  47. Tracopower TXL-025-25S INPUT VOLTAGE fails or glitches = OUTPUT VOLTAGE maintained Mean Time Between Failures = several years Industrial Device = heavily tested, good conformity Short Circuit and Over Voltage Protected Three Year Guarantee TE/MPE/MI - CERN Machine Interlocks

  48. USER_PERMIT SEE Glitch Filter USER_PERMIT BEAM_PERMIT If USER_PERMIT FALSE for <=1.6us then it’s IGNORED! If USER_PERMIT FALSE for >1.6us then it’s ACCEPTED! = NO BEAM ABORT = BEAM ABORT TE/MPE/MI - CERN Machine Interlocks

  49. reference http://indico.cern.ch/conferenceDisplay.py?confId=20366 http://doc.cern.ch//archive/electronic/cern/others/LHB/public/lhcb-2003-049.pdf http://doc.cern.ch//archive/cernrep/2007/2007-001/p547.pdf http://doc.cern.ch//archive/electronic/cern/others/LHC/Note/project-note-385.pdf TE/MPE/MI - CERN Machine Interlocks

  50. Closing remarks • -it was a LOT of effort to get the tests organised • -Thanks to Thijs Wijnands / Christophe Martin et al. for all the great work • -additional tests won’t need as much effort • afterLHC beam commissioning started – we could do some more! • If we feel it’s necessary: • Optical Transmission Board… (CIBO) • Fibre Extended User Interface (CIBFU/CIBFC) • VME Equipment (find the weaknesses?) • We should plan some beam time next year. NOT a priority for LHC • http://indico.cern.ch/conferenceDisplay.py?confId=56796 • “Better the devil you know” TE/MPE/MI - CERN Machine Interlocks

More Related