----- Original Message ----- From: Frank Zimmermann To: Francesco Ruggiero ; Gianluigi Arduini ; Jose.Miguel.Jimenez@cern.ch ; Karel.Cornelis@cern.ch ; Daniel Schulte ; Fritz Caspers ; Tom.Kroyer@cern.ch ; Thomas.Bohl@cern.ch ; Elena.Shaposhnikova@cern.ch ; Joachim.Tuckmantel@cern.ch ; Noel.Hilleret@cern.ch ; Adriana.Rossi@cern.ch ; Trevor.Linnecar@cern.ch ; Hermann.Schmickler@cern.ch Sent: Thursday, January 23, 2003 9:35 PM Subject: Summary of Brainstorming Meeting on SPS Electron-Cloud Microwave Studies 23.01.2003 Summary of Brainstorming Meeting on SPS Electron-Cloud Microwave Studies 23.01.2003 Present: Fritz Caspers, Francesco Ruggiero, Tom Kroyer, Joachim Tuckmantel, Elena Shaposhnikova, Daniel Schulte, Karel Cornelis, Frank Zimmermann Excused: Thomas Bohl, Noel Hilleret, Gianluigi Arduini ----------------------------------------------------------------------------------- Tom Kroyer presented several slides describing the proposed experiment (see attached power point file). The idea to measure the integral electron cloud density originated in the context of the LHC reflectometer project, whose primary objective is the detection of obstacles in the LHC. Noel Hilleret then proposed to perform a test experiment in the SPS. Standard LHC button BPMs do not work well above 0.5 GHz, and for the microwave study special pick ups are needed. The proposed type are rf button with a conical coaxial transition. They can operate at frequencies up to 10 GHz or higher. A smaller prototype of such a device is successfully in operation at the CTF. Measurements in the SPS should be performed at frequencies between about 2 and 3 GHz. The idea is to operate with the lowest H mode, which can propagate through the dipole chambers, but stay well below the cutoff of E modes that couple to the beam and also would lead to mode mixing and enhanced attenuation. For the frequency range of interest the expected attenuation is 50 dB over 1 km. The dynamic range of the spectrum analyzer is larger than 100 dB. Treating the electron cloud as a plasma with a typical density of 10^12/m^3, the expected phase shift is -25 to -17 degree. A similar effect is well known from the ionosphere where an electron plasma is responsible for phase shifts in the GPS systems. The peak ionospheric electron density is comparable (10^12 m^-3) to that in our accelerators. In a side remark, Fritz pointed out that these electrons and the instabilities they induce are the reason why star-wars weapons based on charged particle beams cannot work. Of course, we were hoping that the effect on our storage rings will turn out to be more benign. Further, the phase shift in the atmosphere is quoted as 1 m delay over 200 km at 1.5 GHz. This number agrees within a factor of 2 with the analytical estimate. The exact number might depend on details of the density profile of the ionosphere, which we have no information of. The principle of the electron density measurement is that the cloud is not static but its density is modulated at multiples of the revolution frequency 43 kHz. The phase modulation at this frequency will introduce sidebands around the primary rf carrier frequency. The latter must be adjusted so that beam-related signals do not overlap with the frequency of the sidebands, and the latter can be detected. The microwaves can be sent either in beam direction or against the beam direction, and these two measurements might give complementary informations. In addition to density measurements, there is the possibility to study the effect of the cyclotron resonance. This occurs at 28 GHz/Tesla times B (the field B is 0.117 T at injection into the SPS). There should be an enhanced absorption at this frequency. However, if this will be visible is not clear, since the H mode couples only weakly to this resonance (el. field lines are parallel to the static magnetic field B). So it might be easier to observe the cyclotron resonance at higher frequencies and higher magnetic fields. However, Fritz pointed out that even if the absorption is not directly measurable, we may still 'listen' to the beam at the cyclotron frequency. The experiments so far are all fully parasitic. If we wanted to carefully study the cyclotron resonance at different frequency values (different fields) dedicated beam time would be required. We discussed the optimum locations for the rf buttons. One requirement is that there should be spare cables, another less stringent one might be that the radiation should not be too high. The proposed locations are BA3 and BA4, where according to Joachim cables likely are available. It is thought that most of the electron cloud detectors are in BA5, so that BA5 and BA4 might be a promising alternative to BA4 and BA3 (if cables can be found). In addition to the buttons and cables, a drive amplifier and a pre-amplifier should be installed in the tunnel. Further, outside the tunnel, a synthesizer and spectrum analyzer are needed at the sending and receiving ends, respectively. A 10 MHz frequency reference signal would be nice at both ends, but is not essential. A resonant circuit may reduce the required power level. Summarizing the proposed electron cloud experiments. There are three main applications: - measurement of the integral cloud density - effect of cyclotron resonance - possibilty of shaking the electrons to suppress cloud build up or to destroy the coherence which causes the instabilities In principle the obstacle finding could be tested in the SPS as well, by closing sector valves, as proposed by Karel. However, that might overconstrain the choice of location. The rf pick ups could also be useful for other beam diagnostics. There may further exist a remote possibility to acquire some informations about the frequency dependent machine impedance. Karel asked if one could measure the cloud density over shorter distances, e.g. to study electron cloud near septa and kickers. There is no guarantee that this will work (weak signals may be marked by beam-related noise), but, at least in principle, in the absence of beam noise Fritz is able to measure nanodegree modulations. As for the practical implementation, hardware could be built either by BI or by the Vacuum group. The vacuum group might be better, because they in any case will need to do the final installation. Vacuum flanges (CF36 or CF40) which allow for later connections of the pick ups should be installed with a high priority (within a few days?). The buttons can be produced in parallel and mounted later within a short time. Installation could be done by the vacuum group and be confirmed with Alain Spinks (now in AB group). Gianluigi Arduini and Miguel Jimenez will be contacted. This discussion was a dry run for two presentations next week. Daniel Schulte will describe the proposed experiment at the Daresbury workshop in the UK, and Fritz will present it to the US LHC collaboration meeting on beam instrumentation. One week later, Tom Kroyer will give a presentation to the LTC.