Single Pass Stripline Beam Position Monitor Design , Fabrication and Commissioning

To monitor the position of the electron beam during transport from the Booster Synchrotron to the Storage Ring at the Australian Synchrotron, a stripline Beam Position Monitor (BPM) has been designed, fabricated and installed in-house. The design was based on an existing stripline in the Booster and modified for the transfer line with a particular emphasis on ensuring the line impedance is properly matched to the detector system. The initial bench tests of a prototype stripline showed that the fabrication of the four individual striplines in the BPM was made precisely, each with a measured standing wave ratio (SWR) of 1.8 at 500 MHz. Further optimization for impedance matching will be done for new stripline BPMs. The linearity and gain factor was measured with the detector system. The detector system that digitizes the signals is an Instrumentation Technologies Brilliance Single Pass [1]. The results show an error of 1 mm at an offset (from the electrical centre) of 10 mm when a linear gain factor is assumed and an RMS noise of ~150 um that decreases to < 10 um with increasing signal intensity. The results were under our requirements for the transport line. The commissioning results of the stripline will also be presented showing a strong signal for an electron beam with an estimated integrated charge of ~50 nC with a position stability of 28 um (horizontal) and 75 um (vertical).


Introduction
To monitor the position of the electron beam during transport from the Booster Synchrotron to the Storage Ring (BTS) at the Australian Synchrotron, a stripline Beam Position Monitor (BPM) has been designed, fabricated and installed in-house.The BPM uses four capacitive pick-ups that generate a voltage signal when an electron beam travels past.These pick-ups consist of four striplines arranged LQ D ³VTXDUH´ configuration to calculate the transverse location of the passage of the electron beam relative to the striplines.

Design And Fabrication
An existing design for a quarter-ZDYH VWULSOLQH ZDV modified to accommodate new constraints in the BTS.The length of the stripline was chosen to optimise the signal received by the BPM electronics at 500 MHz.The mechanical design is shown in Figure 1 with the arrangement of the four pick-ups.The calculation of the transverse position is given by where k x and k y are gain/conversion factors that give X and Y in meters and A, B, C and D are the peak voltages generated by the electron beam as it travels past the stripline.
In addition to designing the appropriate length of the stripline pickup, the characteristic impedance of the striplines must also be designed to match the electronics ) to minimise signal loss and distortions.The equation used to model the impedance is [2], where h is the height of the strip from the ground plane (chamber wall), w the width of the strip, r is the dielectric constant (in this case of air), Z 0 the characteristic impedance at f = 0 Hz and e LV WKH (XOHU ¶V constant.In this particular design w = 15.7 mm and h = 3.0 mm giving Z 0 = 48 :.This was an error in the initial design, the ideal would have been letting h = 3.17 mm which gives Z 0 = 50 :.The assumption used for the modeling of the impedance is that r and Z 0 are frequency independent.Frequency dependent equations were used EPJ Web of Conferences nm, by the average of the measured gradients (2.155).This gives k x = k y = 21,550,000 nm.Assuming this constant gain, the data in Fig 6 indicate that at beam offsets of 10 mm in either plane the expected position error is ~1 mm.If necessary this data can be used in a look up table to linearise the measured Libera positions.
Another important factor is the noise floor of the system.At the lowest signal intensity that can be detected by the Libera the RMS noise on the position was PHDVXUHG WR EH P JDLQ VHWWLQJ RI -53 and maximum ADC count of 1600 out of 32k).In the middle of the range WKH 506 QRLVH ZDV P JDLQ VHWWLQJ RI -20 and maximum ADC count of 15,500).The BPM was installed and commissioned with the Libera unit and the measured calibration factors.

Initial Commissioning Results
The stripline BPM was installed in the BTS and tested with a 150 ns bunch train.The raw ADC data shown in Figure 7 shows the bunch train of 150 ns (20 samples) followed by ~240 ns of ringing (30 samples).Finally a second pulse is also observed at an offset of 960 ns (120 samples).If the second pulse is a reflection of the original signal, the time of 960 ns (cable specification of 3.97 ns/m) corresponds to the signal travelling a total of 240 meters.The cable laid between the Libera and BPM is too short for the pulse to be internal reflections.The delay is also too long for the pulse to be a second train of electron bunches.The duration the extraction kicker is ³RQ´ at the start of the BTS defines the time window for the electrons to enter the BTS.The width of the kicker is only 200 ns wide so it cannot be a second train of electrons.Fortunately the secondary pulse can be excluded from the calculations so in the meantime it is not of concern.m.This meets the present requirement for beam stability in the BTS.The results here match the results at other light sources [3].Based on these measurements it is expected that the electronics will be sensitive enough to measure the beam position for charges as low as 0.5 nC with a resolution of at least P The BPM will be used to monitor drifts in the beam position to ensure maximum transfer efficiencies are maintained.

Discussion and Future Direction
The results from the RF measurements were not satisfactory with measured impedances well short of the UHTXLUHG .Simulations will be needed to find the optimal configuration to have a matching system [4] as well as improvements in the measurement technique to identify and characterise the impedances [5].
This BPM is the first of three that will be made and installed in the transfer lines at the Australian Synchrotron.The knowledge and experience will be use the improve the current design and eventually work towards designing resonant striplines [6] that can reach VLQJOH EXQFK UHVROXWLRQ RI WKH RUGHU P IRU FKDUJHV DV low as 10 pC.

Fig. 7 .
Fig. 7.Raw ADC counts from the Libera for the four pick-ups showing ringing after the bunch passes and a secondary pulse 960 ns after the first pulse passes.The measured beam stability during one operation period ZDV VKRZQ WR EH P where the estimated integrated charge of the electron beam was ~50 nC.With a strong signal the expected resolution of the electronics is < m.This meets the present requirement for beam