伊藤 功 氏 - 東京大学物性研究所 軌道放射物性研究施設
Mr. Ito Isao - Synchrotron Radiation Laboratory, Institute of Solid State Physics, University of Tokyo
Researchers at the University of Tokyo constructed the high-intensity soft X-ray beam line for cutting-edge materials science in the large-scale synchrotron radiation facility, SPring-8 (Figure 1).
They installed a polarization-controlled undulator in the beam line to switch the synchrotron radiation polarization (Figure 2). The undulator is a piece of equipment that generates high-intensity synchrotron radiation by repeatedly bending the electron beam orbits.
The polarization-controlled undulator consists of four undulators generating horizontal polarization and four undulators generating vertical polarization. The undulators are positioned alternately with electromagnetic phase shifters between them. The phase shifters have three polarized electromagnets positioned so the poles alternate. When the phase shifter is excited, the periodically alternating magnetic fields are generated. When the electron beam passes through the alternating magnetic field, it follows the bump orbit. This bump orbit generates phase differences between horizontal and vertical polarization. By controlling this phase difference and superposing the horizontal and vertical polarization, it is possible to generate linear polarization and left/right circular polarization.
To generate high-quality synchrotron radiation with the polarization-controlled undulator, the phase shifter must meet several conditions including:
- The phase shifter’s magnetic field should have a consistently stable output so it does not degrade the quality of the synchrotron radiation, and warp or fluctuate the whole electron beam orbit.
- The magnetic field incurred on the electron beam during passing through the phase shifter should be near zero so the orbit of the electron beam does not shift too much before and after the bump orbits. In other words, integral quantity of the magnetic field from the input to the output of the phase shifter (in what follows, we call integral magnetic field [G･cm]) should be as close to zero as possible.
To meet these conditions, we designed and manufactured a phase shifter prototype (Figure 3).
For the performance evaluation regarding the requirements for this phase shifter, we measured DC magnetic field. Before developing the system based on NI hardware and software, we took measurements with the DC magnetic field measurement system constructed with a Hall probe and a 3D movement device (Figure 4).
This system can measure the magnetic field distribution by slowly moving the Hall probe into the phase shifter using the 3D movement device, and performing the integration of the distance moved. However it was not enough to evaluate condition (1) because this system performs the value measurement for only one point to evaluate the stability and reproducibility of the magnetic field. Furthermore, because this system measure the DC magnetic field distribution by slowly moving the Hall probe from the input to the output of the phase shifter, it takes approximately three hours to take the measurement, and the effect of the temperature dependence in the power supplies and the measurement device is indispensable for evaluating condition (2). Therefore, we developed the flip coil DC integral magnetic field measurement system to measure the DC integral magnetic field quickly and with high precision.
Figure 5 shows the principal drawing of the DC integral magnetic field measurement taken by the flip coil.
As shown in Figure 5, when a coil (number of turns = N, length = L, width = W) rotates inside the DC magnetic field (B0), induction voltage (V) is generated in the coil as shown in the following equation:
Furthermore, when the induction voltage is integrated by the rotation time, integral magnetic field (B0L) is solved by the following equation:
Because this method performs the measurement within the flip coil rotation time, it is not affected by the temperature dependence in the power supplies and the measurement devices.
The following requirements were necessary for the flip coil DC integral magnetic field measurement system:
- Control coil currents with sufficient resolution, stability, and reproducibility to generate an integral magnetic field as close to zero as possible (ideally, 1 mA or less).
- Perform data acquisition with sufficient resolution to measure the integral magnetic field as close to zero as possible (ideally, 10 G cm or less).
- Have synchronization and certainty between the stepping motor rotation control and the integral magnetic field measurement to measure the induction voltage by rotating the flip coil.
- Give the user the ability to create a control program in accordance with the test contents.
Considering these requirements, we chose the NI LabVIEW Real-Time Module and an NI PXI system to construct the flip coil DC integral magnetic field measurement system.
Figure 6 shows the external view and block diagram of the flip coil DC integral magnetic field measurement system.
The flip coil has a 600 mm by 5 mm glass epoxy bobbin, wrapped 10 times with 0.2 mm diameter copper wire. To rotate the coil, we use an Oriental Motor RK566BE stepping motor. Rotation speed is 180 degrees every 0.5–1 s. We use an Omron E6B2-CWZ6C incremental encoder to measure the coil rotation angle.
We used an NI PXI-6123 simultaneous sampling multifunction data acquisition (DAQ) device, an NI PXI-6221 multifunction M Series DAQ device, an NI PXI-6733 high-speed voltage output module, an NI PXI-7330 motion controller, and an NI PXI-8106 embedded controller to construct the flip coil control system. We control these devices using the VI we created in LabVIEW Real-Time.
For the control procedure, we send an external reference signal from the PXI-6733 to the power supply. The power supply runs a current corresponding to the external reference signal to the phase shifter. The current sent to a phase shifter is raised or lowered in a ramp control of 0.1 A every second. When the phase shifter is excited by the DC current, the motion controller rotates the flip coil and the measurement is gathered by synchronizing (20 MHz) the flip coil rotation angle (encoder signal) and the phase shifter DC magnetic field (flip coil induction voltage) with two DAQ devices. Because the flip coil induction voltage is weak, we amplify it with a low-noise preamplifier. We use equations a and b to convert the measured induction voltage to integral magnetic field. The previously mentioned control procedures are executed on the real-time OS. Figure 7 shows the front panel of the flip coil controller VI “Control Flip Coil.vi.”
The PXI-6733 controlling the current can output an analog signal with 0.3 mV resolution, so the current running to the phase shifter can be controlled with 0.3 mA resolution by inputting this analog signal to the power supply’s external reference signal terminal. This sufficiently fulfills requirement A.
The PXI-6123 has a 16-bit analog-to-digital converter (ADC), and it can measure with a minimal dynamic range of ±1.25 V, thus the resolution is 2.5 V/216=40 μV. Using this DAQ module and a low-noise preamplifier that gain of 20dB, we can realize 0.4 μV resolution (40 μV/100). By converting the 0.4 μV resolution to integral magnetic field using equation b, it becomes 2 G x cm. This sufficiently fulfills requirement B for resolution.
Because both the voltage output module and the DAQ module use the same clock speed (20 MHz), we can synchronize the power supply for the phase shifter and measurement of the integral magnetic field with 50 ns precision. Also, because the built-in PXI-8106 controller has a real-time OS, the control program is not delayed by any interruptions and executes in the microsecond level of time definiteness. This sufficiently fulfills requirement C.
We can easily create control programs to match the test contents by using graphical programming and the multitude of functions in LabVIEW. For example, we easily achieved synchronization between the motion controller and two DAQ modules using the NI-DAQmx functions and error wiring. In addition, we easily created the ramp control program using LabVIEW, which raises the current in constant time intervals when a current is run into the phase shifter. This sufficiently fulfilled requirement D.
Effects of the Installation
Figure 8 (left) shows the flip coil induction voltage when approximately 640 G DC integral magnetic field is generated only by a central magnet of the phase shifter. The flip coil is rotated counterclockwise at a speed of 180 degree/0.8 s. Figure 8 (right) shows the induction voltage integrated by the rotation speed. By repeating this measurement five times, the standard deviation of the average value was found to be 0.12230±0.00004 V・s. Because N = 10 and W = 5.2 mm, the DC integral magnetic field from equation b is 10888±4 G・cm.
The flip coil DC integral magnetic field measurement system constructed with LabVIEW and the NI PXI system realized a current ramp control with 0.3 mA resolution and integral magnetic field measurement with 2 G x cm resolution. This made it possible to perform the DC integral magnetic field measurement that previously took approximately three hours, in less than one second. We can now achieve evaluation tests with practical stability and reproducibility using this flip coil DC integral magnetic field measurement system.