CONTROLLING THE MAIN INJECTOR TEST STATION USING LABVIEW
University of South Carolina, Columbia
Project Supervisor: Joe Dey
This paper describes the creation of a virtual instrument used to monitor the Main Injector Cavity Power Amplifier. The system is upgraded from Microsoft C version 6.0 to Window NT LabView version 5.1. The report describes the importance of the test station and the similarities and differences between the old system and the new system.
The objective of the project is to create a virtual instrument using Windows NT LabView to test the Main Injector Cavity power amplifier. The Main Injector Test Station is design to monitor the power amplifier performance and reliability. This system consists of three major subsystems: the physical power amplifier stand, the power supplies of AC/DC voltages and Radio Frequency (RF) drives, and a data acquisition and control system. The present data acquisition and control system is written entirely in C language and is complied using Microsoft C version 6.0.
A recent upgrade is implemented to the data acquisition and control system of the test station. The software now uses the graphical programming environment of LabView. This paper describes the importance of the power amplifier as well as the power amplifier test station. It also describes how the application of LabView replicates the old test station and the similarities and differences between the old system and the new system.
The Main Injector is built to accelerate the beam and inject it into the Tevatron. Beams are sent from the Booster to the Main Injector at a velocity of 8 GeV. Eighteen Main Injector Cavities are used to increase the velocity of the beam from 8 GeV to 150 GeV. On each cavity a new power amplifier is mounted. These power amplifiers provide up to 200 kW of power with 4kW solid state drivers. The power amplifiers consist of a low-level radio frequency drive (LLRF), a solid state amplifier, and an Y-567 power tetrode. The LLRF drive is a low level, 53 MHz signal that is use to drive the test station. The solid state amplifier amplifies the power from100 W up to 4 kW. The Y-567 tetrode power tube amplifies the signal from 4 kW up to 200 W. The power amplifier in essence creates a signal that couples into the cavity through a coupling loop. The signal then travels through the alumina seal into the vacuum. The alumina seal separate the air from the Beam tube vacuum. The signal reaches the gap in the center of the cavity and allows the beam to be accelerated on the beam axis around the Main Injector. A schematic drawing of the Main Ring/Injector Cavity with the power amplifier is shown in Figure 1.
The amplifiers are design to create reliability and reduce the number of failures that cause downtime. The Main Ring amplifiers could only produce 112W. The new Main Injector amplifiers can produce a maximum output power of 200 kW at an anode voltage of 30 kV. If the amplifiers do not run at maximum power, the failure creates downtime. Thus monitoring the power amplifiers is vital to the production of the beam in Main Injector Accelerator.
Figure I. Schematic diagram of the Main Injector Cavity with Main Ring Power Amplifier
The Main Injector Test Station is designed to monitor the power amplifier performance and reliability. The station is divided into three subsystems: the physical power amplifier test stand that includes a power amplifier, the anode modulator, and the water-cooled RF load, the AC/DC voltages power supplies and the RF drives, and the data acquisition and control system. A diagram of the test station is shown in Figure 2.
The Main Injector Test Station will be completed in the Fall of 2000. The Booster and Main Injector Test Station will be identical. Therefore, the Booster Accelerator Test station is being use to monitor and test the amplifiers. The power amplifier is not mounted on the cavity as it is in the Booster or the Main Injector tunnel. Instead, a water-cooled 50W load is place on the amplifier in order to calculate the output power. The power amplifier is connected to a cavity resonator with a resonant frequency of 51 MHz. A 30 kV anode supply feeds into the anode modulator, which is used for programming the anode voltages on the power tube. The RF drive is obtained from a programmable frequency synthesizer.
Figure 2: Diagram of the Main Injector Test Station
The old computer system, the IBM PC/AT, with a digital-to-analog (D/A) and analog-to-digital (A/D) converter cards handles the data acquisition and control. The D/A card is used to program the power tube anode voltage, the cascode grid bias, and the RF drive level. It creates a waveform with a DC voltage of 0 to 10 volts. The duty cycle, the risetime over the period, determines when the waveforms are on and off. The RF level remains constant at approximately 4 volts peak. The anode and cascode levels are adjusted and scaled to produce cascode currents and anode voltages. The A/D card is used for monitoring the operating parameters of the test station. The inputs to the A/D card come from a monitor box affix above the power amplifier. The box contains voltage dividers, current sensors, and RF diode detectors. Several monitors are also built into the power amplifier to observe the RF waveforms. The A/D cards converts the analog signals from the amplifiers into digital signals the computer can understand.
The power amplifier output is accomplished by calorimetry. The water flow through the RF load and the temperature differential between the inlet and outlet are used to calculate the power dissipation. The equation use to calculate the output power (in kilowatts) is shown below, where GPM represents gallons per minute and D°C is the change in temperature between the inlet and outlet.
The system is written entirely using C language and is complied using Microsoft C version 6.0. This language evolved around the 1970’s. Since it’s existence, several languages have been created to make programming easier to execute and comprehend. Also, other parameters have been added to the test station where it would be difficult to alter the program. It is decided that an upgrade of the system is vital to the continuous use and effectiveness of the test station.
When the program is written in C language, it begins execution by initializing the instrumentation system and setting up the data structures. The main loop control is entered after initialization. The loop awaits keyboard input and passes control to the appropriate routines when activity is detected. Control is accomplished through the function keys and numeric keypad. The function keys initiate major test, allow the system to reset, and exists the program. The numeric keypad moves cursor to adjust the RF drive frequency, power tube anode program, and the cascode grid bias program. An example of the operating parameter display during power testing of the PA is shown in Figure 3.
The PC/AT personal computer is replaced with a Pentium II processor. It includes 10 GB hard disk, a 3.5 inch disk drive, and an expandable zip drive that can hold up to 100 MB. 6713 Analog Output, 6071E Analog Input, and PCI-GPIB (General Purpose Interface Bus) cards from National Instruments are installed to allow the computer to interface with the test station. Window NT LabView is installed to handled the data acquisition and control. An 8 channel DSP (Digital Signal Processor) is use to produce analog waveforms in range of ±10V. The 36 channel A/D converter is use to convert the analog signals into digital signals that the computer can understand. The GPIB allow the computer to interface to a Hewlett-Packard 8656B Signal Generator where it sets the frequency (52.8-53.1 MHz) and the amplitude of the RF level.
The Windows NT LabView is a revolutionary graphical programming environment based on G programming language for data acquisition and control, data analysis, and data presentation. It allows engineers to work in a graphical programming environment where they create programs using block diagrams instead of text code. Programs are created to control instruments and acquire signals from sensors where measurements and calculations can be made easily. In LabView, two windows are open. The first window is use to create the block diagrams of the program. The other window also called the front view displays the control use for the program. Both examples are shown in the Appendix.
Figure 3: Example of the operating parameters display during power testing of PA
The programming language that LabView uses is called G. It allows programs to use virtual instrument where their appearance and operation can imitate actual instruments. The program occurs in a sequential order. The program begins by setting the frequency and level to the signal generator. Once the signal generator is set, the waveforms for the RF level, the anode program, and the cascode grid bias program are generated. Adjusting a slider to set the duty cycle and DC voltage creates the waveforms. The program reads the data from each signal and displays the result. The program exits by pressing the stop button. The waveforms produce by the LabView program is shown in Figure 6.
Comparing the Systems
The old system and the new system perform similar tasks with a few changes. Both systems generate waveforms for the power tube anode program, the RF gate, and the data acquisition triggers. The triggers are use to average the data from the waveform. The programs also read (A/D) the inputs from the parameters of the amplifier and display the measured result.
The difference between the old system and new system is the old waveforms were generated by hardware. The new waveforms are generated by a DSP. The new system also allows the computer to interface (GPIB) with a signal generator to set the RF drive frequency. The old system updates the data at 10 Hz. The new system updates the data at 15 Hz. Finally, in the old system, the duty cycle of the waveform can only be at 2%, 25%, 50%, and 100%. The new system allows the duty cycle to be adjusted from 2% to 100% with a slider.
The goals of the project are met. The new system replace with LabView replicates the old system that uses the C language. The Power Amplifier Test Station now has a flexible data acquisition and control system. The program can easily be altered for future upgrade.
It has been an honor to be the first person from South Carolina to participate in the SIST program. I would like to thank Dianne Ingram, Elliot McCrory, Mr. Davenport, and the other SIST committee members for given me the opportunity to gain the experience I needed in my field of study by selecting me for the program. I also will like to give an enormous thanks to my supervisor, Joe Dey. Without his knowledge, supervision, and patience, my project will not have turn out as successful as it did. I will like to thank Mitch Adamus and Jeff Ruffin who showed me how the Test Station operates. This was very important to my project. Finally, I would like to thank the SIST interns who have made this summer enjoyable for me. This will be an experience I will never forget.
 Reid, J. and H. Miller. “A 200 kW Power Amplifier and Solid State Driver for the Fermilab Main Injector.” IEEE Particle Accelerator Conference. 1995, pg1544.
 Dey, J. and D. Wildman. “Higher Order Modes of the Main Ring Cavity at Fermilab.” IEEE Particle Accelerator Conference. 1995, pg.1675.
 Champion, Mark S. “A New Data Acquisition and Control System for the Power Amplifier Test Station.” IEEE Particle Accelerator Conference. May 1991, pg.1511
 “LabView.” National Instrument Corporate Headquarters. Austin, TX, 1998.
Figure 4: Example of a Front Page in LabView
Figure5: Example of the Block Diagram in LabView
Figure 6: Waveforms generated in LabView