CNGS Horn Experience issues since NBI 2010 Ans

CNGS Horn Experience (issues since NBI 2010) Ans Pardons (with input from Jean-Marc Cravero) CERN Ans Pardons, CERN NBI 2012 1

Outline • Introduction – – CNGS Horns Memorable Moments Electrical Properties Power Convertor Circuit • CNGS horns operation issues 2010 -2012 – Exit Inductances – Thyristor – Capacitor Banks • Summary Ans Pardons, CERN 2

Introduction: CNGS horns CERN Gran Sasso • 150/180 k. A, 0. 3 Hz avg. (20 Hz inst. ) • Water-cooled • Design lifetime: 20 106 double pulses Ans Pardons, CERN 3

Introduction: CNGS horns POWER CONVERTORS High Voltage cables (1 km, 12 per horn) 7. 7 k. V, 9. 4 k. A (horn) 6. 3 k. V, 5. 6 k. A(refl) ce a f r u he s t n O nd: u o r erg d n U TRANSFORMERS X 16(H) x 32 (R) HORN/REFLECTOR 150/180 k. A Water cooling & Demineralisation (20 m) Ans Pardons, CERN 4

Introduction: Memorable Moments Double pulses 1, 2 E+07 1, 0 E+07 Water conductivity issues 8, 0 E+06 Electronics radiation damage Reflector outlet water leak 6, 0 E+06 Broken stripline cable 4, 0 E+06 Capacitors Inductances& measured thyristors broke 2, 0 E+06 0, 0 E+00 янв-06 июл-06 Ans Pardons, CERN дек-06 июн-07 дек-07 июн-08 дек-08 июн-09 дек-09 июн-10 дек-10 июн-11 дек-11 июн-12 5

Introduction: Memorable Moments • 2006: water outlet leak – Badly designed brazed & machined ceramic assembly – All brazed ceramics in horns replaced • 2007: broken stripline cable – Insufficiently restrained & brazing-weakened cables – Replaced with restrained & (more) rigid solution of Agplated copper plates Ans Pardons, CERN 6

Introduction: Electrical Properties Ans Pardons, CERN 7

Converter Circuit Diagram 1 st 2 nd Ans Pardons, CERN 8

Exit Inductance Failure: mid 2011 • History – Recuperated from WANF, no documentation at the time – Found later from manufacturer’s documentation: Under-dimensioned for CNGS – Replaced with new, correctly dimensioned inductances all OK • Failure details Mechanical fatigue Insulating foil degradation Short circuit at the exit of the power converters Ans Pardons, CERN 9

Thyristor Issues: Mid 2011 • At each inductance failure Thyristor stack for pulse 1 – pulse 2 – The output of the converter is in short-circuit, leading to the destruction of the thyristor stacks • Additional observation – Sometimes pulse 1&2 fired at the same time due to the EM susceptibility of the thyristor firing board – Problem solved with filter added on the board and additional EM shielding Thyristor firing board Inductance For pulse 1 - Pulse 2 10

Capacitor Banks Issue: Early 2012 • History – New capacitors installed in 2006, 24 capacitor cans per horn – Designed for 20 million pulses at 10 k. V • Performance – During machine stop early 2012, measured capacitance value found to be at ˜ 50% (horn) and 90% (reflector) of nominal – At this moment, only 10 million pulses and max. 7 k. V life time issue Ans Pardons, CERN 11

Capacitor Banks Issue: Early 2012 • Solution – For 1 st half of 2012: 7 spare cans (original design) were installed horn up to 84% ok until new – In June 2012: 24 new cans installed in the horn converter (new design from manufacturer) – No degradation observed since • Reason for premature wear-? ? ? – Unknown (awaiting tests results from manufacturer) Ans Pardons, CERN 12

Summary • 2006 -2010: Many mechanical issues, lots of repair & re-design work done, many lessons learnt (and in parallel: enhancement of other systems) This seems to have paid off: no mechanical or cooling issues in 2010 -2012 • 2010 -2012: Electrical issues found and solved • Lessons learnt: – Double check when recuperating existing equipment Do look the gift horse in the mouth! – Capacitor banks have to be measured on a regular basis to detect life time issues (early). – EM compatibility is always a tricky issue when dealing with k. A current pulses Ans Pardons, CERN 13