Abstract:

Active alignment algorithms for linear colliders such as the Compact LInear Collider (CLIC), require measurements of beam positions inside accelerating structures and quadrupoles. The talk gives an overview of how to achieve the required accuracy on the beam position measurements using single cell cavities, undamped and damped accelerating structures. Simulations and CTF II beam experiments are presented.

Hans welcomed Prof. Mnich and Prof. Böhm from the Technical University (RWTH) Aachen and thanked them for the supervision of Jan's Ph.D. thesis.

After motivation for the beam position measurement at CLIC, Jan described the position monitoring by resonant cavities with the advantages of the possible use of the accelerating structures as BPMs and without gain equalization.
He showed the design of a 30 GHz single cell resonant cavity BPM used at CTFII, where the TM₁₁₀ mode is sensed through waveguides. This cavity was used in the probe beam of CTFII and the beam position varied inside the BPM by an upstream corrector magnet. The BPM RF signal was downconverted by a local oscillator to f_IF=120MHz and analyzed on an oscilloscope. Both amplitude and phase of the signal were calculated and show the expected behaviour, the mean amplitude deviation is 8.4 µm rms, the mean phase deviation is 1.1° (but with the measurement split into two phase states). The  phase was found to be valuable because of its strong position dependency in the cavity centre.
This design is not usable for CLIC where one needs dipole mode damping and better common mode rejection. A possible design for CLIC was presented that achieved strong damping and further common mode rejection by a different coupling geometry.
An undamped 30 GHz accelerating structure was used to measure simultaneously beam position and angle with beam position and angular resolution of 6.3 µm and 7.5 µrad, respectively.
The heavily damped short SICA prototype 3 GHz structure was modified to sense the dipole mode. The structure and the coupler were modelled by GdfidL and compared to HFSS. The frequency response agrees for several peaks in the spectrum, though there are some differences in the shape. These first simulations with GdfidL are promising. The signal from the damping waveguide was also measured and compared to the simulations. The peak frequencies agree reasonably well but quality factors are significantly different. The signal shows a very clear position dependence with a mean deviation of 5.7 µm rms from an expected straight line fit in the complex plane.

- A question was raised if noise subtraction of the 0 position raw signal could improve the resolution. This would strongly depend on the source of the noise which is not clear.
- The 3 GHz and 30 GHz structures show about the same resolution. This could be due to the last part of the mixer channel which is identical, beam jitter or the similar beam size for the measurement. This point is to be better understood.
- A comment was made that this BPM could have a possible use for stochastic cooling.
- It was asked which part of the mechanical design properties is the most critical. The slot position for the coupling slot is important since an error breaks the symmetry and leads to coupling for the fundamental mode.

- Concerning the Q factor differences for the SICA prototype, it was added that one peak in the spectrum is there only due to a manufacturing error discovered later that might not be so well understood. In addition, there is an uncertainty in the SiC load description in the simulation.