Microwave diagnostics for electron cloud detection on the SPS F. Caspers, E. Mahner, T. Kroyer B. Henrist, J.M. Jimenez Thanks to the SPSU study team members.

Microwave diagnostics for electron cloud detection on the SPS

F. Caspers, E. Mahner, T. KroyerB. Henrist, J.M. Jimenez

Thanks to the SPSU study team membersFor helpful discussions

Outline

Layout Basics and history First results during scrubbing run

Conclusion and outlook

ECM 08 Nov 2008 CERN

2Nov. 20th 2008

Microwave diagnostics in the SPS ECM 2008 Caspers, Mahner, Kroyer, Henrist, JimenezF. Caspers, E. Mahner, T. Kroyer

Basic layout for microwave transmission setup (SPS 2008)

Principle idea:“When electromagnetic waves are

transmittedthrough a not too dense plasma, theyexperience a phase shift and possibly a

smallattenuation”

Antenna1 Antenna3

14.4 m 33.1 m

Antenna2 (excitation)

dipolequad

F. Caspers, T. Kroyer

2008 is similar to the 2004 setup

3Nov. 20th 2008

Microwave diagnostics in the SPS ECM 2008 Caspers, Mahner, Kroyer, Henrist, Jimenez

Some basics and history (1)

4Nov. 20th 2008


Some basics and history (1a)This general expression for the phase shift in a low density plasma filled waveguide could suggest that it would be ideal to work at or very close to waveguide cutoff…Well in the end we are interestested in a good signal to noise RATIO and if we have too much attenuation when working close to cutoff , it does not help a lot.Also we must keep in mind that we dont like steep slopes in the HTf (hardware transfer function) in order to avoid HTF related AM/PM conversion in other words the upper and the lower modulation sideband should pass with the same attenuation otherwise the difference between upper and lower sideband would look like a phasemodulation.Thus avoid steep slopes !

5Nov. 20th 2008



6Nov. 20th 2008



7Nov. 20th 2008



8Nov. 20th 2008



9Nov. 20th 2008



The presently used setup in BA5 is rather similar, except that we have dedicated button like PU‘s and kickers with a circulator after the hybrid and rather narrow-band filters (40 MHz) with steep slopes.

The filters are centered at 2.5 GHz. We used this time as receiver a conventional spectrum analyzer and a vector spectrum analyzer.

10Nov. 20th 2008


Preliminary measurement results in 2008 - scrubbing run (1)

SPS super-cycle 17656

Display from a conventional spectrum analzyer.

The phase modulation is in the small peaks, which are +/- 44 kHz away from the center.

The beam induced signal are the two bigger peaks next to the center line.

11Nov. 20th 2008



Display from a conventional spectrum analzyer.

The phase modulation is in the small peaks, which are +/- 44 kHz away from the center.

The beam induced signal are the two bigger peaks next to the center line.

In this display one can see the true width of the carrier, the modulation sideband and the beam signals.

SPS super-cycle 17651

12Nov. 20th 2008



SPS super-cycle 17568-69

Time axis is from top to bottom, intensity is color coded. CW signal on

13Nov. 20th 2008




CW signal off-on-off-on

The VSA has only 80 dB SFD (spurious free dynamic range) due to ADC limitations!

14Nov. 20th 2008




Shift of CW frequency

15Nov. 20th 2008




Higher harmonics of the modulation sidebands since the modulation is rather rectangular than sinewave

16Nov. 20th 2008



SPS super-cycle 17552Expanded view of one modulation sideband vs time.

17Nov. 20th 2008



SPS super-cycle 17641Intensity of the modulation sideband vs time in logarithmic scale

18Nov. 20th 2008



SPS super-cycle 17650Intensity of the modulation sideband vs time in linear scale

19Nov. 20th 2008


Preliminary measurement results in 2008 –AM contamination (1)

SPS super-cycle 21196, July 9th 2008

How about AM contamination?

Are we really sure that what we see on the green trace (upper half) is pure FM or PM

At least in this case its pure AM since the demodulation function of the VSA tells us that there is NO FM or PM signal where we would expect it.(lower trace)

20Nov. 20th 2008




Here one can see a small assymetry in height of the modulation lines in the conventional SPA display which is ALWAYS a hint for superposition of AM and FM or PMAND we also see a very small FM –PM Linie in the lower half of the display where we expect it.

21Nov. 20th 2008




Another exampleUpper half:Conventional spectrumLower halfPM demodulation

This shot was taken after acceleration at flattop where usually we observed the strongest PM signals

2 batches

22Nov. 20th 2008




Here we have in the upper half the conventional spectrum and in the lower half the PM demodulated partObviously we cannot see anything during accerlation and a quantitative evaluation from this colour coded plot is not really possible.

23Nov. 20th 2008


LHC beam in SPS, 25 ns bunch spacing, 72 bunches, 1 batch


PS-type shielded button pickup signals

Yellow trace stripline in stainless steel chamber #1Blue trace pickup in chamber #1 containing enamel clearing electrodeMagenta trace pickup in stainless steel chamber #2 w/o enamel, for referenceNote, that there is some (unexpected) low frequency ringing (EMI) in the blue and magenta traceWe will try to understand its origin and get it removed in the future

Note that depite EMI contamination one can clearly see (light blue trace) the buildup of the e-cloud related current over the duration of the batch

24Nov. 20th 2008


Conclusion and Outlook

The intermodulation problem observed in 2003 (experiments in BA2) has been considerably reduced using a proper RF setup with highly selective filters at the frontend, both, for excitation and on the pickup side. However the application of the phase demodulation function (using a vector spectrum analyzer) is strongly recommended for the elimination of residual AM contamination due to faint saturation effects [gain compression] (more than 50 dB below the carrier) caused by beam induced signals.

First electron cloud transmission experiments performed during the 2008 SPS scrubbing run in June have shown very promising results in BA5. The presence of electron clouds was identified simultaneously with button pickups.

Microwave transmission over a full LHC arc at cryogenic temperature has been already successfully demonstrated around 7 GHz. The technique discussed here can be directly applied for integral electron cloud diagnostics over a full arc in the LHC. For this purpose the installed in situ reflectometer couplers at the end of each arc could be used.

Many thanks to Christian Carli for lending us the real-time spectrum analyzer and contributing during data taking.

25Nov. 20th 2008


Conclusion and Outlook (additional comments July 14th)

As for potential AM/PM conversion due to steep slopes in the HTF (hardware transfer function) one can test the hardware transmission with some know AM level on the generator (and of course known spurious PM) and check at the downstream end of the transmission for changes (in particular additional PM or FM).

For the present case we see from the data shown (well, not all shown this morning) that we had up to 20 dB high pure AM sidebands and no PM indication.

For further reduction of intermodulation risk via beam induced signals the passive front end filter BW (both for the driver and pickup amplifiers) can be reduced to say 2x frev. (here we had 40 MHz with frev in the SPS = 43 kHz).

If additional dynamic range is required the carrier can be suppressed with a high Q passive element notch filter before the received signal is going into the first PU amplifier.

Instead of using the VSA (vector spectrum analzyer) one could also apply a VNA (vector network analyzer) which has a suffiently large IF bandwidth (at least 100 KHz for the SPS) and operate it in the CW FFT mode (may require the use of macros).There might be a higher instantaneous dynamic range with the VNA as compared to the VSA (80 dB limited by the ADCs)

26Nov. 20th 2008


Related recent results from other labs (1)

27Nov. 20th 2008


Related recent results from other labs (2) [ILCDR July2008]

Microwave diagnostics for electron cloud detection on the SPS F. Caspers, E. Mahner, T. Kroyer B. Henrist, J.M. Jimenez Thanks to the SPSU study team members.

Documents

sps ecm

microwave diagnostics

setup slide

outlook ecm

cern slide

kroyer basic layout

jimenez thanks

layout basics