Progress towards high-resolution phase-space-density measurements using tomography and slice-emittance
Vitaly Yakimenko, for the BNL ATF team.
In the initial period of this program the main emphasis was placed on the improvement of the various components of the accelerator. The goal is to achieve the necessary level of stability and accuracy of the control for tomographic recovery. Those improvements were immediately noticed and appreciated by other experiments at ATF.
Electron beam transport analysis and control. The data of quadrupole lenses settings and beam size measurements on multiple beam profile monitors over the entire Post Linac transport line were combined to achieve a good understanding of the transport line. This gives us confidence in our ability to "blindly" control phase advance (phase advance was not measured directly). The same technique was successfully extended to the STELLA beam line (ATF Beam Line 1). It allows to "blindly" tune an electron beam (without sufficient diagnostics due to limitation in the interaction region) through a 500 um limiting aperture. Both horizontal and vertical phase advances were controlled during the measurements. Figure 1 shows the excellent agreement that has been demonstrated between fitted beam envelope and the measured beam sizes.
Figure 1. Calculated and measured beam sizes along the transport line (dotted line) and measured beam sizes on beam profile monitors (x’s). Red and blue correspond to the vertical and horizontal planes, respectively.
RF phase control. A considerable afford was made to improve RF stability. This is a crucial element for the complicated and time consuming phase-space measurement. Various correlations were measured to localize problem areas. A dedicated computer code was written to keep track of the drift in the different components. As a result of these investigations, temperature stabilization boxes were installed by Mark Montemagno for the gun and linac low level RF electronics. The daily phase drift was reduced by feed back correction from 30-50 degrees to approximately 10. The RF gun phase feedback program was finalized and is part of the ATF's routine operations. Effective (=observed by user) long-term (=hours) RF phase drifts was reduced to 1-2 degrees.
Figure 2. Measured Gun phase drift without Feed Back correction before the installation of the temperature stabilization in the low-level RF.
Figure 3. Measured Gun phase drift without Feed Back correction after installing temperature stabilization for low-level RF.
With the feed-back on, this figure (naturally) becomes very flat. As noted above, under feedback one has to measure the drift by using results from experiments or effects such as the charge stability from the cathode (which is phase dependent). These measurements indicate 1 to 2 degrees phase drift over many hours when the feedback is enabled.
Mathcad Software for the tomography measurement. The first version of the software was written in Visual C++. The main code was able to change the current of the quadrupole magnets, insert/retract popin monitors, control the video switcher, request measurements from the frame grabber and analyze the measured data. It comprised 1500 lines of code. The main disadvantage of that programming environment was having hidden physics under massive programming. Recently Bob Malone developed an ATF library that allows communicating between any Mathcad program and the ATF control database. The main advantage of this approach is the great visibility of the physics in the program. Following that, the Visual C++ version of the program was transcribed into the Mathcad environment. It now comprises three separate ONE page programs: The first controls the trajectory and minimizes the quadrupole magnet steering. The second is used to perform phase rotation in both vertical and horizontal planes and force the frame grabber to save images. The third is used to analyze previously saved images. The enhanced transparency of the program immediately elucidated a previously unseen problem. For example, a beam profile monitor image imperfection, such as the image of a fiducial mark excites of high frequency components in the filtered projections and causes a large numerical noise in the recovered phase space. Special digital filtering is under development.
Figure 4. A real intensity image of the electron beam as measured on the phosphor screen. The fiducial marker appears on the image as line with less intensity (1). It translates into small disturbance on the Y projection (2) and large fluctuation in the "filtered" projection (3).
Figure 5. The recovered vertical phase space density distribution with digital noise from high frequencies in the filtered projections. The level of the noise is approximately 10-20 % compared to the maximum intensity.
Orbit correction - quadrupole magnet steering minimization. A special program was written to minimize steering effects from quadrupole magnets on beam line. Analysis of the collected data demonstrated that a re-survey of the quadrupoles is necessary to solve that problem. We plan to re-survey the quadrupole magnets on that line. The steering leads to electron beam position change on the High Energy Slit and disrupts the precise selection of different longitudinal slices of the beam. We plan to use quadrupoles on beam line 1 to do phase rotation for slices selected by the HES as another solution to this problem.