Present Status and Future Plans for the MKE Kicker Magnets

1
Present Status and Future Plans for the MKE
Kicker Magnets

• Acknowledgements some slides are adapted from
  previous presentations by Fritz Caspers, Enrique
  Gaxiola, Tom Kroyer Jan Uythoven.

M.J. BARNES, AB/BT
2
Transmission Line Kicker Magnets

• The kicker magnets installed at LSS4 LSS6 are
  travelling-wave type magnets, each consisting of
  7 ferrite cells
• 30 year old, recuperated, equipment
• Transition piece between vacuum tank and kicker
  magnet
• Tank length 2174 mm
• Magnet length 1700mm.

MKE magnet in the clean room
3
Cross-section of MKE Magnet(prior to 2003 at
LSS4 and 2005/2006 at LSS6)

• Ferrite C core constructed from three ferrites.
  
• In general, ferrite used is type 8C11 from
  Ferroxcube.
• Rectangular aperture for the beam.

Ferrite
Rectangular aperture
4
LSS4

• LSS4 is used for extraction to CNGS/LHC
• LSS4 has 5 magnets, magnetically in series, each
  powered by its own PFN and terminated in a
  matched resistor
• Three of the magnets have large (L) apertures
  (147.7mm x 35mm)
• Two of the magnets have small (S) apertures
  (135mm x 32mm).

5
LSS4 Parameters
CNGS LHC (Protons)
Energy 400 450 GeV/c
Deflection 0.54 0.48 mrad
Nominal PFN Voltage 50 51.2 kV
Nominal Impedance 10 10 O
Available rise-time 1.1 10 µs
Measured rise-time (1?99) 1.1 1.1 µs
Usable flattop length 10.5 7.9 µs
Flattop ripple ? 2 ? 1
E. Gaxiola et al, Upgrade And Tests Of The SPS
Fast Extraction Kicker System for LHC And CNGS,
EPAC2004.
6
LSS6

• LSS6 is used for extraction to LHC
• As of 2006 LSS6 has 3 magnets, magnetically and
  electrically in series, powered by a single PFN
  and terminated in a short-circuit
• Two of the magnets have large (L) apertures
  (147.7mm x 35mm)
• One magnet has small (S) aperture (135mm x
  32mm).

7
LSS6 Parameters

• Before 2006, LSS6 consisted of 4 series magnets
  terminated in 10O. In 2005, proof of principle
  tests were carried out, in lab, for a
  short-circuit system
• Rise and fall time increased
• Reduced PFN voltage for same deflection
• BUT voltage reversal.

LHC (protons)
Energy 450 GeV/c
Deflection 0.429 mrad
Nominal PFN Voltage 33.1 kV
Nominal Impedance 10 O
Available rise-time 8.8 µs
Measured rise-time 8 µs
Usable flattop length 7.9 µs
Flattop ripple ? 1
8
Beam Induced Heating

• 2004 Need to reduce both beam impedance and
  losses.
• The MKE (SPS extraction) kickers LSS4 LSS6 are
  heated by the beam due to their beam coupling
  impedance. Ferrite heating is caused by coupling
  between beam and real part of ferrite impedance.
• High intensity beam can result in high power
  deposition in ferrite.

9
Past Developments

• Kicker magnets at LSS4 (2003) LLS6 (2005/2006)
  equipped with
• High thermal conductivity Aluminium-Nitride
  (AlN180) plates
• Water cooling
• PT100 temperature sensor.

10
Measured Probe Temperature

• Design Space simulations were carried out to
  determine the relationship between the highest
  ferrite temperature and probe (PT100) temperature
  (with water cooling).
• Validated by machine measurements for 8C11
  ferrite kick strength diminishes above 80C
  measured, which corresponds to 125C highest
  ferrite temperature, i.e. Curie temperature of
  ferrite (AB-Note-2004-005 BT, section 5).

11
Summary of situation at 2006

• Above about 120?C the 8C11 ferrites loose their
  magnetic properties, this has been measured
  during the scrubbing runs.
• Note 4E2 ferrite has a Curie Temperature of
  400?C.
• Additional risk is structural damage above about
  150?C.
• For normal operation the beam is interlocked at
  70 ?C measured (125?C on the ferrites)
• For scrubbing runs the beam is interlocked
  above 90?C measured, which is about 140?C on the
  ferrites.
• The water cooling system allows the magnet to be
  operational with approximately double the beam
  power deposition in the ferrite (AB-Note-2004-005
  BT, section 5).

12
Beam Coupling Impedance

• Beam coupling impedance can be reduced using
  conductive stripes (serigraphy), i.e. interleaved
  comb structure, directly printed onto the
  ferrite blocks and having a reliable contact to
  the metallic HV plates at either side
• Capacitive coupling between stripes (stripes
  carry beam image current).

13
Longitudinal Measurements (1)

• Comparison between the two extreme cases
• MKE-L8, without any serigraphy on 8C11 ferrites
• MKE-L10, serigraphy on all seven 8C11 ferrite
  cells.
• Significant reduction of longitudinal impedance
  from serigraphy.
• Low frequency resonance directly linked to
  geometry of serigraphy.

Note Impedance based on 2.2m length (1.7m
actual) hence should be increased by a factor of
1.3
L8
L8
L10
L10
14
Dissipated Power

• Calculated using beam spectrum measured during
  2004 SPS scrubbing run (data from J. Uythoven),
  but theoretical beam coupling impedance
• 272 LHC bunches in SPS at 450 GeV
• Comparison between fully shielded MKE-L10 and
  MKE-L8 (no shielding at all).

Serigraphy (painted stripes) reduce calculated
power deposition by a factor of gt4, for LHC beam.
15
Longitudinal Measurements (2)

• In a comprehensive measurement campaign data for
  all types of MKE magnets was collected.
• Conclusions
• High Curie temperature ferrite (4E2) displays
  similar Real Impedance as 8C11 (Ferroxcube)
  ferrite.
• S-Type L-Type magnets, without serigraphy,
  display similar Real Impedance.

16
CNGS Power Depositions

• Calculated power deposition based on
• CNGS beam spectra measurements made by G.
  Arduini and T. Bohl (4.5 s period) see
  Note-2004-39
• 2.5x1013 protons per pulse
• A total cycle duration of 6 s.

• From Jan Uythovens presentation to APC, 10
  December 2004 (real part of beam impedance
• 2 x 210 W/m ? Thottest-equilibrium107?C
• Replace theoretical beam impedance of SPS kicker
  by measured data, for MKE-L8 (see slide 13),
  scaled by 2.2/1.7
• 2 x 178 W/m ? Thottest-equilibrium91?C (based
  on 26?C tunnel)
• Replace theoretical beam impedance of SPS kicker
  by measured data, for MKE-L10 (see slide 13),
  scaled by 2.2/1.7
• 2 x 25 W/m ? Thottest-equilibrium35?C (based
  on 26?C tunnel)
• Therefore shielding stripes on MKE-L10 kicker
  reduce beam induced power deposition,
  attributable to nominal CNGS beam, by a factor of
  7 c.f. MKE-L8.
• For 1.7x1013 protons per pulse 3 cycles per
  39.6s, replace theoretical beam impedance of SPS
  kicker by measured data, for MKE-L10 (see
  slide 13), scaled by 2.2/1.7
• 75 W/m ? Thottest-equilibrium40?C (based on 26?C
  tunnel)

17
Measured Temperatures (2007)

• LHC type beam. No serigraphy (yet) at MKE4.

18
Measured Temperatures (2007)
19
Measured Temperatures (2007)
CNGS type beam. MKE6. Serigraphy on L10 results
in a factor of 6.4 (9/1.4) lower temperature rise
than L9 (factor of 7 expected see slide 16).
20
HV Issues Pulse Shape for MKE6
Terminating the series magnets by a short
circuit the magnet peak voltage is reduced in
absolute value (33kV PFN 16.5kV magnet
voltage). However, because of the reflection from
the short circuit, there is a full negative
voltage (-16.5 kV) on the magnet. If we consider
peak to peak, the voltage is 2x16.5 33 kV. The
magnets are designed for 30kV.
21
HV Breakdown in MKE6

• The MKE6 magnets were tested one by one in the
  lab only S6 (adjacent to short circuit) was
  therefore tested with the actual waveform shape.
• After re-installation of magnets (2006-2007 shut
  down), some conditioning (breakdown) problems
  with magnet L10 (closest, electrically, to PFN).
  But
• L10 magnet is exposed to the longest duration
  and dV/dt of both positive and negative voltage
• Before 2008, the DC conditioning was only made
  with positive polarity. During the past shut down
  (2007-2008) we have also carried out conditioning
  with negative DC so far it seems that such
  conditioning has had good effect i.e. no HV
  problems with L10
• Last year, we had little time to condition in
  MKE6. Effort has been made, during 2007-2008
  shutdown, to take every opportunity to do
  conditioning, and progressive improvement has
  been observed.
• Slightly negative effect due to the stripes is
  not excluded, but we have no evidence so far.

22
Summary

• Aluminium Nitride plates and cooling system
  allows the magnet to be operational with
  approximately double the power deposition in the
  ferrite
• Serigraphy (painted stripes) reduces predicted
  power deposition, in ferrite, by a factor of gt4,
  for LHC beam this is consistent with temperature
  measurements made during October 2007 scrubbing
  run (slide 18).
• Serigraphy (painted stripes) reduces predicted
  power deposition , in ferrite, by a factor of 7,
  for CNGS beam this is consistent with
  temperature measurements made during October 2007
  (slide 19).

23
Transverse Impedance

• Information re Transverse Impedance, and
  measurement techniques, can be found in
• Tom Kroyers presentation Wire Measurements on
  the MKE Extraction Kicker Magnets APC meeting
  10/11/2006.

Shielding may increase transverse impedance at
100MHz, but reduces transverse impedance above
300MHz.
24
Future Plans

• All 9 MKE magnets will be equipped (eventually)
  with serigraphy
• 1 (and 2/7 !) of 5, installed, L-Type magnets
  equipped with serigraphy
• 0 of 3, installed, S-Type magnets equipped with
  serigraphy
• We have 1 spare L-Type 1 spare S-Type
  magnet (without serigraphy).
• Spare Aluminium-Nitride plates ordered (expected
  delivery end of May 2008)
• 8C11 ferrite in stock (that can be prepared
  serigraphed)
• L-Type 7 top, 13 middle 8 lower. Therefore one
  complete magnets worth
• S-Type 7 top, 10 middle 7 lower (including 4
  radioactive ferrites). Therefore one complete
  magnets worth
• Ferroxcube to manufacture and return following
  L-Type 8C11 ferrites 6 top, 9 lower (date of
  expected return not yet known)
• Plan is to replace one S and one L magnet at each
  SPS shutdown (using the normal operation year to
  convert the 2 spare magnets for the next
  shutdown) -- 4 shutdowns required!!
• BUT potential problem of converting 2 spare
  magnets during operation if problem is
  encountered with an installed magnet, there may
  not be an available spare
• If faster deployment is needed, we need to
  investigate the possibility to equip another 2
  magnets during the shutdown. In which case a
  shutdown of at least 3 months is necessary (work
  includes dismantling, machining of radioactive
  ferrite (radii), serigraphy (by another CERN
  service), mounting, vacuum tests, HV test and
  conditioning in lab, installation, vacuum, DC and
  pulsed conditioning in SPS).

25
Bibliography

• T. Bohl, CNGS Beam in the SPS Beam Spectra,
  Note-2004-39
• F. Caspers Impedance Measurement of the SPS MKE
  Kicker by means of the Coaxial Wire Method,
  PS/RF/Note 2000-004
• F. Caspers, A Retrofit Technique for Kicker
  Beam-Coupling Impedance Reduction,
  CERN-AB-2004-048
• F. Caspers et al, The Fast Extraction Kicker
  System in SPS LSS6, EPAC 2006
• E. Gaxiola, Upgrade and Tests of the SPS Fast
  Extraction Kicker System for LHC and CNGS,
  PAC2004
• E. Gaxiola et al, Performance Of The CERN SPS
  Fast Extraction for the CNGS Facility, PAC2005
• E. Gaxiola et al, Experience with Kicker Beam
  Coupling Reduction Techniques, PAC2005 E.
  Gaxiola, SPS Extraction Kicker Performance with
  Impedance Reduction Measures, http//ab-div.web.c
  ern.ch/ab-div/Meetings/APC/2006/apc061110/EG-APC-1
  0-11-2006.pdf
• T. Kroyer, Wire Measurements on the MKE
  Extraction Kicker Magnets, http//ab-div.web.cern
  .ch/ab-div/Meetings/APC/2006/apc061110/TK-APC-10-1
  1-2006.pdf
• T. Kroyer et al, Longitudinal and Transverse
  Wire Measurements for the Evaluation of Impedance
  Reduction Measures on the MKE Extraction
  Kickers, AB-Note-2007-028
• M. Timmins et al, SPS Extraction Kicker Magnet
  Cooling Design, AB-Note-2004-005 BT (Rev.2),
  TS-Note-2004-001 DEC (Rev. 2)
• J. Uythoven, MKE Heating and Measured Power
  Spectra CNGS BEAMS , http//ab-div.web.cern.ch/a
  b-div/Meetings/APC/2004/apc041210/uythoven.pdf
• J. Uythoven et al, Beam Induced Heating of the
  SPS Fast Pulsed Magnets, EPAC2004