Folie 1

Diagnostic of Coating Results: Microwave Measurements F. Caspers, S. Federmann, E. Mahner, B. Salvant, D. Seebacher1ContentsIntroductionEcloud and coatingMeasurement possibilitiesModulation of signals

Realisation of the experimentProblems and Solutions

Latest results and preliminary conclusions

Summary and outlook2IntroductionHigh intensity beams may build up ecloud

Undesired effect Reduces luminosity, beam stability

Mitigation: Coating of beampipe reduction/supression of ecloud expected

testing required3Measurement ConfigurationsPick up electrodes to monitor ecloud (talk by C. Yin Vallgren)Local measurement only

Pressure measurements (talk by M. Taborelli)

Measurement of phase modulation of a microwave (MW) signal due to ecloud 4MW Transmission Measurements - TheoryMeasurement of phase modulation (PM) of MW signal could give information of integrated ecloud density:

is proportional to the electron cloud density:

Expect a phase shift of = 2.3 10-3 rad (order of magnitude)

wp. Plasma frequencywInjected frequency (f = 2.68 GHz)wc. Cutoff frequency (f = 1.23 GHz)Llength of transmission path (6.5 m)

ne typical electron cloud density 1012 5ModulationConsider a continuous, pure sinusoidal wave (CW):

Amplitude ModulationPhase Modulation6AM Modulationwhere:AC. constant determining overall signal amplitudea modulation index [0|a|1]m(t)... normalized modulation signalC.. frequency of carrier

If:

Taken from: R. Witte, Spectrum and Network Measurements, 1991 7AM ModulationTime domain:

Frequency domain:C C + m C - m Result in time domain8PM ModulationIn case of sinusoidal (and narrowband) modulation:

.... modulation index

180 Phase shift!9PM ModulationTime domain:

Frequency domain:(narrowband FM)C C + m C - m 10Result in time domainModulationCombination of modulation:

Unequal height of sidebands!This is the practical situation in SPS MW experiment (BA5) and we have to separate the undesired AM signal from the desired PM signal by suitable instruments of the Vector Spectrum Analyzer (VSA)11Experimental RealisationExcite preferably the TE10 mode, couple to Magnetic field due to space constraints in the pumping port regions

Not much influence of beam signals since TEM like mode of beam considerably decays on the side of the beampipe

E field of the TE10 modeH field of the TE10 mode

12Experimental RealisationCoupling antennas (loops) on left and right side of beampipe in the pumping ports

13First experimental setup

VSA Vector Spectrum AnalyzerBP Band passDC .. Direct current bypass 14Problems with this SetupMain problem: Intermodulation distortion (IMD) caused by: (43.3 kHz = SPS revolution frequency)~43.3 kHz on power supply~43.3 kHz induced on cables down to tunnel beam harmonics (around 2.68 GHz)PM to AM conversion caused by different sideband attenuation in the hardware transfer function (HTF)

Change of CW signal amplitude and phase over the magnetic cycle (just of academic interest since we stay on the flat bottom)15New SetupTo get rid of this effect:

Concentrated on only one section: uncoated/coated

Installed high pass filters on surface as well as new amplifiers and capacitors in the tunnel16Observation with New SetupIMD still presentChange of carrier found to be effect of deformations in beam pipe tests on stand alone magnet confirmed this

Evolution of carrier amplitude over cycle without beamdBmTime [s]17Latest ChangesInstalled DC blocks (corner frequency ~100 MHz) in tunnel

Reduced hardware on surface (only DC blocks at the end of the cable used)18Preliminary resultsBeam 1 batch: PM signal in uncoated magnet 10 dB above noise

Uncoated magnetCoated magnet19Preliminary results

Uncoated magnet, 3 batches20Preliminary results

Coated magnet, 3 batches21Preliminary ResultsBeam 3 batches:Signal in uncoated magnet increases about 3 dB with each injectionOn average signal 13 dB over noise

Uncoated magnetCoated magnet22Preliminary ResultsBeam 3 batches, 10% less intensity:Signal in uncoated magnet increases about 3 dB with each injectionOn average signal 10 dB over noise

Uncoated magnetCoated magnet23What quantity of the ecloud is measured?The carrier is modulated by the ecloud (shape relative unknown) in time domain. The modulation repeats as the ecloud every revolutionWhat we seen on our instrument is the peak value of the fundamental wave of the modulation (ecloud)Measured value has to be multiplied with an, shape dependent and not precisely known factor to get the peak value in time domain!

FundamentalHarmonicsOffset24Preliminary ResultsWe see a demodulated signal of about 10 dB above noise(1 batch) or about 14 dB above noise (3 batches) for the demodulated PM signal on the coated section

For a beam with 90% nominal intensity we obtained a reduction o modulation signal with respect to nominal beam by 3 dB for 3 batches in both cases

Increase of 3 dB of modulation sideband with each injection

For 2 batches we should see an increase in modulation signal strength by 3 dB with respect to 1 batch and about 5 dB for 3 batches

25Summary and OutlookLatest measurements have delivered promising results (no modulation seen in the coated section and a clear signal in uncoated beam pipe for the same beam and instrument settings)

Need to exclude any potential sources of error and signal contamination

Tests foreseen in SPS with 75 ns beam

Ecloud testbench in PS 84 where we have a clearing electrode and MW transmission to be used for comparison26Thanks for your attention!27