The beam Position Monitor (BPM) system that will be used in the superconducting heavy ion synchrotron SIS100 consists of the following components: i) pick-up (PU) installed in the cryostat with the chassis being a part of the beam pipe ii) signal feed–throughs that are leading the signal out of cryostat iii) impedance matching transformers (described later on in the text) iv) low noise low input–impedance amplifiers [1] and v) Libera digitizer that consist of four channel ADC (with 14 bit resolution and 125 MSa/s) and integrated FPGA [2]. The requirements for the BPM system are specified in Ref. [3]. The development of the BPM system was focused on the further investigation of electrical and mechanical features of the PU and the test of the impedance matching components. In parallel an optimization of the position evaluation algorithm was performed and tested on signals stored during one run of SIS18. The results of optimization of the PU are reported in Ref. [4]. The mechanical test of the most crucial PU components interconnections will be carried out in the liquid helium in the new test bench under construction. The first operation of the test bench is foreseen in spring 2008. This paper presents the results of the tests performed for the coupling transformers and optimization of the position algorithm.

In this paper new approaches for BPM (Beam Position Monitor) measurements are described, which are needed in hadron accelerators with strongly varying beam parameters, such as intensity, accelerating frequency and bunch length. After the data collection and offline evaluation in 2005, first FPGA implementations of algorithms were completed in 2006 and tested at SIS18 and CERN PS. Main aspect of the first tests was the proof of concept in terms of online calculation feasibility. This includes online calculation of the needed integration windows as well as the baseline restoration algorithms. The realization of the hardware and the data handling are discussed. Least squares techniques were used for parametric fitting to gain bunch signal properties which can be used to monitor beam position.

In the present contribution results of a simulation of linear-cut Beam Position Monitors (BPMs) based on a design using a metal coated ceramics are compared for two different geometries. The investigated BPMs will be used in the FAIR facility. The simulations were performed using CST Microwave Studio. The main goals of the design optimization were pick-up sensitivity and linearity of the position determination. High position sensitivity can be achieved by reduction of plate-to-plate cross talk caused by coupling capacities. In case of ceramic based BPMs insertion of an additional guard ring into the gap between the active plates leads to an increased sensitivity by about a factor of two.

The first steps of research and design made for the BPM (beam position monitor) system for the FAIR project, with partly use for the SIS-18 upgrade, will be described. It includes the evaluation software parts as well as the hardware part for the direct digitization of the position pick-up signals and, in addition, the first design steps for the cryogenic detector itself. Design of the BPM detector using Finite Element Methods There are a lot of requirements in the design of the BPM detector for the FAIR project, the most important are listed here: • The BPM hast to fit in the cryogenic modules planned e.g. for SIS100, see Fig. 1 as an example for the straight section. • The shape of the BPM should be elliptical as the beam pipe for impedance matching. • The materials have to be compliant to the UHV and cryogenic requirements of the machine. • The target value for the resolution of the BPM is 0.1mm, therefore the mechanical stability should be better than 50μm (because of additional error terms coming from thee electronics and data evaluation, see below). Figure 1: Layout of a cryogenic module for the straight section of SIS100 in the FAIR project Based on studies done for the “shoe-box type“ BPM used in the synchrotron of the HICAT facility [1] the crosstalk between the plates in one plane was investigated, which should be as low as possible. The different geometrical configurations and the calculated values are shown here: Metal coated Ceramics Metal plates No guard ring, 1mm gap -5.1 dB -7.9 dB No guard ring, 2mm gap -8.1 dB -10.8 dB With guard ring -20.8 dB -22.5 dB From these results one would deduce that a design based on metal plates containing a guard ring would be the best solution, but the mechanical stability would be a big problem due to the amount of parts, which have to be aligned within 50μm under cryogenic conditions. Therefore the other solution based on metal coated ceramics tubes will be investigated, a first sketch can be seen in Fig. 2. Figure 2: First layout of a BPM pick-up for SIS100 As the simulation tool “CST Microwave Studio” (CSTMW, version 5) is used, all calculations are performed using the transient solver. The work is on-going in 2006. Digitisation of BPM signals and evaluation

In this paper we describe new approaches for BPM (Beam Position Monitor) measurements, dedicated to hadron accelerators which have strongly varying beam parameters, such as intensity, accelerating frequency and bunch length. Following the adjustment of the signal dynamic, direct digitalization and treatment of digitized data, we should reach a BPM resolution of 0.1mm. Interchangeability between accelerators should be provided, which results in almost autonomous data treatment algorithms, free of external status and timing signalling. This should ensure the usability of the system in other bunched accelerator rings. Different operation modes are intended, allowing online storage of beam position data over full acceleration cycles as well as storage of beam waveforms in regions of acceleration that are of special interest e.g. transition, kicking, bunch gymnastics. First results of realised hardware/software combinations will be described and discussed.

New, digital BPM techniques needed in hadron machines, accelerating beams with fast varying frequencies, are to be presented. The role of analog electronics is reduced to signal amplification and attenuation as well as bandwidth limitation. This paper explores approaches for the position evaluation of acquired signals, suggesting systems for "free running" estimation as well as machine timing de- pendent methods. For accurate determining of the trans- versal bunch position, a good integration window estima- tion is needed. Two filtering methods will be introduced for this purpose, median and FFT filtering, both methods detecting peaks at bunch signal starting and ending points. Parallel to those a digital PLL approach is discussed in (1). *

For most beam parameters the stored current is monitored by a commercially available dc-transformer offering a dynamic range from 1 μA to 20 A maximum. Its bandwidth of 10 kHz allows determination of the beam lifetime with ms time resolution. But it is known, that these types of transformers have severe problems by a high current bunched beam passage [1]. Even though an improved version is available, no test with beams has been performed so far and scalability from other machines might be doubtful due to the resonance nature of the severe distortion. Therefore investigations of a different type of sensitive magnetic field sensor in collaboration with the University Kassel (Germany) have been started. The idea consists in taking the scheme of a clamp-on ammeter and using a high-effective magnetic sensor in the core gap. Usually Hall-sensors are installed in such measurement devices, but due to the larger core diameter needed in beam diagnostics and the high dynamics of the beam currents to be monitored other sensors have to be investigated like AMR (Anisotrope MagnetoResistance), GMR (Giant MagnetoResistance) or GMI (Giant MagnetoImpendance) and tailored for this application [2]. The scheme of such a device and a first simulation of the magnetic flux concentrator are shown in Fig. 1. First results show that the GMI effect would be the most interesting candidate for the sensor, but no commercial element is available until today [3]. For GMR the situation looks better [4] and the further development will concentrate on this sensor type.

New, digital BPM techniques needed in hadron machines, accelerating beams with fast varying frequencies, are to be presented. The role of analog electronics is reduced to signal amplification and attenuation as well as bandwidth limitation. This paper explores approaches for the position evaluation of acquired signals, suggesting systems for free running estimation as well as machine timing dependent methods. For accurate determining of the transversal bunch position, a good integration window estimation is needed. Two filtering methods will be introduced for this purpose, median and FFT filtering, both methods detecting peaks at bunch signal starting and ending points. Parallel to those a digital PLL approach is discussed in [1]. *