Prototype of beam position monitor (BPM) for next linear colliders developed and constructed by Branch of the Budker Institute of Nuclear Physics have been tested in the BNL Accelerator Test Facility using a 45 MeV, 0.5 nC single bunch beam. The test set-up consisted of three BPMs, which were mounted on three precision movers with 0.3 m m resolution in both (horizontal and vertical) directions for displacement calibration. The detection electronics allowed to take and process data pulse to pulse independently in horizontal and vertical positions in each BPM. Tests BPMs and detection electronics in lab showed that the potential resolution of the BPM system on the BNL ATF beam was less then 0.1 micron. But jitter at ATF BNL appeared to be about 25 m m that is why we had to reduce the output signals from BPMs by 30 dB to conform dynamic range of detection electronics and jitter. So the limit on resolution of 1.9 m m has been obtained directly with ATF beam in these condition. Several tests were made with 10 dB attenuation. And the first results of these data analysis show that in this case resolution was 0.2 m m.

Description of the experiment and obtained results are presented in this paper.

For preservation of the beam emittance on next-generation of linear colliders the high accuracy of aligning of magnetic elements and accelerating structures is required.

So the sensitivity of BPMs for linear colliders have to be better than 0.1 m m [1]. The simplest and effective microwave BPM is circular cavity, excited in TM₁₁₀-mode by an off-axis beam [1,3,7,8].

The measured amplitude of transverse mode is proportional to the beam offset and bunch charge: .

Where q means the beam charge,
M=sin(kh/2)/(kh/2) - beam transit time factor;
-space factor;
r ' -normalized transverse shunt impedance;
Q - loaded quality factor.

Phase of the oscillations depends on beam offset direction.

One of the main problems is the large amplitude of the fundamental and others symmetrical modes exciting in resonant cavity by beam independently of it's offset. These modes must be dumped. Thermal noise and noise of electronics determine another limit of resolution. Calculations and laboratory tests of BPM prototype and detection electronics showed that the potential resolution is better than 0.1 m m (for BNL ATF beam).

The aim of the experiment described in this paper was testing the complete BPM system in a real beam of accelerated particles. The experiment was made in the BNL ATF using it's 45± 1 MeV, 0.25¸ 0.5 nC single bunch beam with longitudinal size 5 - 10 ps [6].

2. EXPERIMENTAL SET-UP

Figure 1 represents the setup of the experiment. Three prototypes of BPM cavities were made for this experiment (1,2,3). Each BPM was placed on the high precision electromagnetic mover (5) that allows to move cavity in both directions on ± 1 mm with step 0,3 mm. All magnetic elements were magnetically shielded to

Figure 1. Experimental set-up.

exclude the influence on beam. Construction of movers (Fig. 2) provides position measurement of BPM relatively of table support (6) with 0.03 mm accuracy. All three movers were placed on support table and aligned using standard ATF procedure. To be able to move freely, the BPMs were connected to the vacuum channel through bellows (4). Before the experiment all movers were calibrated.

Each BPM has two outputs for horizontal and vertical signal coupled with detection electronics by 4m RF-cables because electronics were placed outside of the tunnel.

Figure 2. Electromagnetic mover.

Before test started the TM₁₁₀-mode resonant frequency of each BPM had been measured and tuned. For fine tuning thermal heating was used. Each BPM had individual heater with feedback temperature stabilization, which automatically kept the frequency at desired value.

Figure 3 represents the detection electronics used in the experiment. For measurements in horizontal (X) and vertical (Y) directions several type of electronics were used. In horizontal direction only beam pass filters, crystal detectors and video amplifiers were used (amplitude method) and in Y-direction the circuits for phase measurements were applied (phase method). These circuits comprise a RF-switch, a bandpass filter, a mixer, a phase detector, an intermediate frequency amplifier, and a video amplifier for each monitor. Laboratory tests showed that the electronics have sensitivity about 10^-12 W and dynamic range about 40 dB [4].

Afterwards the signal is transmitted from RF electronics to the analog-digital converter (ADC) located in CAMAC. Precision mover is connected to CAMAC, too. The above circuit allows to find the coordinates for each single bunch of a beam.

Using 2856 MHz signal from ATF reference line as reference signal for mixing down and phase detecting is important feature of this circuit. It allows using permanent (not pulse) RF signal that have information about "phase of beam".

During May 1997 the first stage of the experiment has successfully been accomplished. The experimental equipment has been installed and connected to the ATF vacuum system. The first results have been obtained. Figure 4 shows the beam-excited spectrum in BPM. The spectra for two different beam offsets are displayed in

Figure 5. The experiment showed a good agreement with the simulation results. Fig. 6 presents the output signal from the amplitude detector in X- and Y-directions.

Figure 5. BPM spectra for two different offsets of the beam

Figure 6. BPM output signals for off-sets in X- and Y- directions

Before data taking began, the beam was placed as near to BPMs centers as possible by ATF equipment. For that procedure amplitude method in all measuring channels was used. After that Y-channels in each BPM were connected to the circuits for phase method. Then all BPMs were moved several hundred microns off center in a known direction to provide a clear displacement signal for tuning of detection circuit for phase measurements.

Figure 7. Amplitude method sensitivity measuring.

During amplitude measurements mechanical offsets of BPMs from line were observed and fixed. Figure 7 presents the results of measurements of amplitude method. Measuring channels X_up, X_middle had one type of crystal detector, channel X_down had more better crystal detector with external DC offset. This picture showed that amplitude method could be used for not very precise beam position measurements in very wide beam displacement range.

Amplitude measurements and first attempt to use phase method showed that beam has a high displacement and angle jitter. And dynamic range of detection electronics is not enough for measuring in such conditions without saturation. That is why output signals from BPMs have been attenuated on 30 dB before transmitting to detection electronics for phase measurements, and all presented results have been made for attenuated signal.

On figure 8 correlation between three BPMs are presented. During measurements all BPMs have been moving on 0.34 micron per pulse.

Figure 9 shows the measured data for BPMs sensitivity calculation. All BPMs have been shifted approximately on ± 200 microns by movers in known direction. Using this data and data about noise of electronics we obtained sensitivity and expected resolution about 0.9 micron for best measuring channel and 1.7 micron for worse.

Figure 8. Phase method. Jitter measurement.

Figure 9. Phase method sensitivity measuring.

Figure 10 explains how to verify the intrinsic BPM resolution using three BPMs independently from beam jitter [5]. Using obtained sensitivity we calculated vertical beam position in each BPM for data that had bean obtained without saturation of electronics. And obtained resolution is 1.9 micron.

Figure 10. Using three BPMs for determine intrinsic BPM resolution.

Several tests were made with 10 dB attenuation of BPMs output signal. But in this case electronics dynamic range was only about ± 15 mm that less then jitter, and

Figure 11. Determination of the BPM resolution.

electronics worked with saturation. That is why the analysis of these data is difficult and it is under way yet. The first results are shown on the figure 11 and resolution in this case is about 0.2 mm.

Complete resonant BPM system has been tested at ATF BNL using real beam, which had position, angle, intensity and energy jitter. In these conditions the resolution of 0.2 m m was obtained. Three independent BPMs were used to exclude position and angle jitter. For absolute BPMs calibration precision movers were applied.

Achieved resolution isn't the limit for these BPMs. It's value was determined by experiment condition.

Dynamic range of used in the test electronics wasn't enough for measurements with real beam jitter. Therefore BPM output signal was attenuated.

Time of laser pulse in RF gun was not absolutely stable relatively of accelerating RF signal, which was used as the reference signal for phase measurements. That is why "beam phase" has jitter relatively reference oscillations and it is the cause of additional jitter noise in the output signal of phase detection electronics. And this jitter couldn't be excluded using three BPMs.

Beam trajectory through BPMs was not absolutely rectilinear because the magnetic field has influence on moving charged particles. And even in the constant magnetic field (for instance the earth field) trajectory deflection from line isn't abiding. It depends on beam energy. So energy jitter in accelerator cause mistakes in resolution determination using three BPMs.

So to improve observing BPM system resolution we suppose increase dynamic range of detection electronics for conformation with real beam jitter, add to BPM system reference cavity for measurements of phase shift between reference line and real phase of the beam, to take in to account the influence of magnetic field on beam trajectory through BPMs.

We would like to thank the ATF staff for their help and hospitality during our stay at BNL. We would like especially to thank Ilan Ben-Zvi and X. J. Wang for making the opportunity of the experiment and smoothly coordinating our communication and activities.