Electron Cooling Rookie Book
Table of Contents:
Magnets – to be written
Power Supplies – to be written
The ECool BPM System
Scrapers – to be written
Flying wire – to be written
Loss monitors – to be written
Pepper pot – to be written
YAG – to be written
MultiWire – to be written
Optical Transition Radiation (OTR) Monitors in ECool
Vacuum System
Water Cooling System
Operation of the Electron Beam
Electron cooling is a process used to reduce the size, divergence and energy spread of a stored charged-particle beam without removing particles from the beam. What this means is the phase space occupied by the stored particles is compressed.
The main idea of electron cooling is a heat exchange through Coulomb scattering between hot antiprotons and a cold electron beam when they are mixed.
A stored Pbar beam is overlapped with a nearly monochromatic and parallel electron beam. The velocity of the electrons is made equal to the average velocity of the stored Pbars. In the beam frame, mixing a hot Pbar gas with a gas of cold electrons results in the thermal energy being transferred from circulating Pbars to freshly produced electrons.
 |
The electron beam provides cooling primarily in the longitudinal plane in addition to the stochastic system that cools more effectively in the transverse planes.
FermiLab bought a commercially-made Pelletron from a company out of Middleton, Wisconsin, National Electrostatics Corporation (NEC). It is known as the 6URe, a 6 million volt, gas-insulated, electrostatic, single-ended type accelerator, and is designed to accelerate electrons. More information can be found on Pelletron machines from this company at www.Pelletron.com.
There are four major components of the Pelletron: 1) a pressure vessel, 2) a high voltage insulating support structure (the column), 3) a charging system to generate the high potential and 4) evacuated acceleration tubes through which the electron beam passes.
1) The Pelletron is housed in a specially designed and certified steel pressure vessel (the tank) that is 144” in diameter and 324” long. Its purpose is to contain the SF6 insulating gas. The nominal pressure inside the tank is ~75 psig, and nominal temperature is ~31 deg C. The tank has several ports built in: an access port on the lower end to allow servicing of components inside; SF6 drain ports; ports for the generating voltmeter (GVM), capacitive pick-off (CPO) and service platform support cables; and ports for electrical feedthrus, beam tubes, and safety sensors.
The SF6 recirculation system is at the base of the tank and serves four functions: cooling, drying, filtering, and removing breakdown products of the SF6. A recirculation blower pulls the warm SF6 gas into the cooling skid located on the platform where the gas is cooled, dried and filtered before returning it to the tank. For servicing the Pelletron, the SF6 gas is removed from the tank to an outside vessel by a transfer skid located on the floor just outside the cave area in MI-31.
2) The support column is a set of hollow aluminum disks, so-called separation boxes, supported by 6 insulating posts (see fig. A). Each post is an assembly of metal and ceramic disks bonded together. In addition to providing mechanical support for all elements, the posts are used to hold electrostatic guarding rings (hoops), which cover the inner space of the column and provide an even distribution of electric fields. The voltage between hoops is distributed by a chain of resistors mounted at one of the posts.
The separation boxes contain electrical generators, powered by a rotating shaft; power supplies of lenses and correctors; and controlling electronics. A larger separation box in the middle of the Pelletron (#3) also has two ion pumps with their power supplies.
Figure A: A standard part of the Pelletron column taken at the time of assembly (the deceleration tube is not mounted).
3) The high voltage terminal is mounted at the top of the column. It houses the terminal charging system, electron gun, collector, and electronics. A cylindrical shell, or dome, surrounds the terminal structure. Electrical power for the terminal is supplied by a drive shaft system consisting of a motor mounted on the base bulkhead plate to drive alternators in the terminal via an insulating shaft. High voltage is generated by means of a mechanical charging system (the chain) consisting of steel pellets (from which the name “Pelletron” comes from) connected by nylon insulating links rotating around two pulleys, one at the ground end driven by an electric motor, and one in the HV terminal. A 7.5 horsepower motor in the bottom of the tank provides power to turn drive sheaves of the chain.
Negative charge is induced on the pellets at the ground end and carried by the pellets to the HV terminal; positive charge is induced on the pellets in the terminal and carried to ground for a total charging capacity of more than 300ua. Table T1 lists some of the charging system parameters.
Table T1: Charging System Parameters
Motor revolution frequency: 19.3 Rev/sec
Motor Power: 5.6 kW
Actual Load: ~0.5 kW
Nominal chain speed: 18.5 m/sec
Chain central diameter around sheaves: 30.5 cm
Distance between sheave centers: 3.68 m
Chain revolution frequency: 1.8 Hz
Length of a single pellet: 32 mm
Gap between pellets: 5.3 mm
Maximum voltage of charging PS: 60 kV
Charging efficiency in good condition: 7.3 µA/kV
(based on two chains with charging at the ground and at the terminal)
This picture shows the charging system. It should be noted that our Pelletron uses a negative configuration.
The terminal voltage control system is a closed loop voltage control system. The device
which accomplishes this is known as the Terminal Potential Stabilizer (TPS). There are several devices that are used as inputs for the TPS:
- The Generating Volt Meter (GVM) is a precision device which generates a signal proportional to the voltage on the accelerator terminal. It is located on the inside wall of the tank and consists of a motor which turns a rotating vane in front of 8 static sector plates. When voltage is applied to the accelerator terminal an AC voltage is induced on the plates. This signal is amplified, rectified, and read out on a digital display on the TPS.
- The Capacitive Pick-Off (CPO) is used to detect high frequency changes in the terminal potential. The CPO plate is located on the inside wall of the tank, read out using a scope.
- Terminal voltage is stabilized by a corona current flowing from needles located in the terminal shell (the so-called Corona Probe). This current is a function of the needle potential with respect to the terminal surface around it. To control the terminal voltage, the needle is connected to a high voltage triode. Correction signal, proportional to the difference between GVM reading and the voltage set by R:TPSTRV is applied to the triode grid. This in turn regulates the current drawn from the terminal by the corona probe.
4) Two identical tube assemblies are located in the column: the acceleration tube transports the injected negative electrons from the terminal to ground; the deceleration tube transports the electrons back from ground to the terminal collector. There are six sections in each tube. Ends of each section are connected electrically and mechanically to separation boxes. Sections are composed of 1’ long modules (see Figure B), which consists of a bonded set of ceramic cylinders and titanium disks with bonded titanium vacuum flanges at the ends. The standard module has 21 insulating gaps with voltage distributed along the module by a string of 0.5 Gohm resistors. Electrodes attached to the titanium disks determine the electrostatic fields in vacuum; the minimum aperture in the tube is 1”.
Figure B: Sketch of a standard module of the Pelletron tube.
Electron cooling of 8-GeV anti-protons is achieved using a continuous electron beam of ~4.3 MV at a current of 0.1 – 0.5 amps passing through a 20-m section in which the two beams interact. In principle the electron beam could be dumped following this interaction but at 0.5 A the beam power is 2.15 MW! Discarding the electron beam at this point would require a large dump to handle the energy and X-rays, and a huge and costly power supply to produce the beam. To over come these issues the electron beam is returned to the Pelletron so that almost all of this power can be recovered. Upon returning to the Pelletron the electron beam enters a “decelerating” tube identical to the accelerating tube. As the electrons pass through this tube, they see an increasing negative potential and are slowed down. They reach the collector at the HV terminal with a low kinetic energy of ~3keVdetermined by the potential between the collector and gun cathode.
This conversion of energy is as simple as a roller coaster starting at one high point and going to the next. At the terminal the electrons start at a negative high potential. As they pass through the “accelerating” tube their potential energy is converted to kinetic energy and they have maximum kinetic energy when they leave the column and are at ground potential. They then coast through the transport line, do their work, and reach the next tube. At this “decelerating” tube they are going back to the negative potential from which they started and thus exchange kinetic energy for potential energy. At the terminal they have lost all or most of their kinetic energy, come to a stop and return as free electrons to the terminal. At the same time other electrons behind them are going through the same process; this is a DC or modulated DC beam, and with small fluctuations the system is stable.
The Electron Cooling beam line is made up of several distinctly-named sections. The electron beam first enters the Acceleration line which runs from the gun through the bottom of the tank and ends before the beam enters the first 90 degree dipole bend magnet (R:DYS1A) under the tank. (The 90 degree dipole bend magnets are made up of two 45 degree bend magnets, labeled A and B.)
The Supply line runs from the first 90 degree dipole bend magnet (R:DYS1A) located under the tank, goes through the MI-31 Stub area and ends at the next 90 degree dipole bend magnet (R:DXS6B) located in the Main Injector Tunnel. The ECool beam line merges with the Recycler beam pipe near the location R:H307.
In the short Adjustment (Before Cooling) line, located in the Main Injector tunnel, electrons are focused and steered to provide a parallel beam in the cooling section. This line is between the 90 degree dipole bend magnet R:DXS6B and the first cooling section solenoid.
The 20 meter long Cooling Section is located in the Main Injector tunnel and is composed of 10 identical solenoid modules.
The 180 degree bend region is located in the Main Injector tunnel and runs from the end of the Cooling Section to just after the 180 degree dipole bend magnet R:DYQ3B. The 180 degree dipole bend magnet consists of two 90 degree dipole bend magnets: R:DYQ3A and R:DYQ3B; the ECool and Recycler vacuum pipes are separated inside the magnet R:DYQ3A near location R:V305.
The Return line is located in the Main Injector tunnel and begins right after the180 degree dipole bend magnet (R:DYQ3B) and ends at the next 90 degree dipole bend magnet (R:DXT1A). The Return line is between the Recycler and Main Injector, so is very vulnerable to Main Injector losses and ramp cycles.
The Transfer line begins in the Main Injector with the 90 degree dipole bend magnet (R:DXT1A), goes through the MI-31 stub area and ends under the tank at the last 90 degree dipole bend magnet (R:DYT6B).
The Deceleration line begins under the tank in MI-31 after the last 90 degree dipole bend magnet (R:DYT6B) and ends inside the tank at the collector.
There are two operating modes for the Electron beam line. When the electron beam starts at the gun, passes through all of the separate sections and returns to the collector, this is called Full Line mode. In the second mode, called U-Bend mode, the electron beam begins at the gun, travels through the Acceleration line, but instead of being bent into the Supply line, the first 90 degree dipole bend magnet (R:DYS1A & B) is turned off and the beam goes straight towards the ground until it reaches the 180 degree dipole magnet known as the U-Bend magnet. After passing through the U-Bend magnet, the beam returns to the Deceleration line and ends in the collector. The major difference between the two modes is that for Full Line mode the Main Injector Safety System must be enabled. In U-Bend mode, the electron beam never leaves MI-31, so this mode can be run regardless of the status of the Main Injector. To switch from full line mode to U-Bend mode, a key must be turned on the interlock chassis at MI-31.
A diagram of the ECool beam line can be found on the homepage: HTTP://www-ecool.fnal.gov and clicking on the box labeled ‘Schematic’ on the left-hand side, or go directly to the page: http://www-ecool.fnal.gov/EcoolDevices2-23-05.pdf .
The Electron Cooling beam line lattice differs from the rest of the Accelerator Div. beam lines. It uses solenoids as the primary focusing element. Dipole bend magnets are used to bend the beam around the 90 and 180 degree bend regions. Small dipole corrector magnets are used to move the beam horizontally or vertically in between the solenoids. There are also quadrupole magnets, but these are not used at this time.
The naming convention used by ECool is different than the other beam lines in the Accelerator Div. The naming of the devices is based on the positions of the solenoids. The naming begins with a solenoid and everything after it is based on that solenoid until another solenoid is reached. As an example, in the Acceleration line, the first solenoid outside the tank is R:SPA06. Every device after that will be R:xxA06x until you get to the solenoid R:SPA07. After that the devices will be R:xxA07x until you get to the U-Bend magnet. This is generally true throughout the ECool beam line. The names given to the devices are based on what type of device it is. Following is a guideline to how the devices are named.
The control system used by the Pelletron came with it from NEC and is called AccelNet.
AccelNet stands for ‘Accelerator Network control system’ and is designed specifically for the control of electrostatic particle accelerator systems. AccelNet runs on PC hardware under the Linux operating system.
There is a dedicated PC located in rack 103 in the MI-31 control room that runs the Pelletron AccelNet control system. Accelnet controls everything inside the tank. The readbacks come out of the Pelletron tank via a fiber optic light link and go into the Pelletron CAMAC crate located in rack 216 in MI-31. They are then accessed through the AccelNet PC. These readbacks are also interfaced with the Accelerator Division CAMAC system through a VME front end node called ‘ECOOL’ (located in the computer room across from the Main Control Room). Through node ECOOL, the readbacks from inside the Pelletron tank can be viewed and controlled via ACNET parameter pages. It should be noted that when the shaft is off, there is no power going into the Pelletron tank, thus no readbacks coming out. If either the AccelNet PC or node ECool develop a problem (i.e. hang-up, crash, etc.), readbacks will stop updating or will report back errors on parameter pages. If this happens, the ECool on-call person should be contacted. If the problem is with node ECool, a simple reboot may solve the problem. But if the problem is with the AccelNet PC, it is more complicated since problems with the PC can cause secondary problems with node ECool.
Readbacks for all other devices associated with the ECool beam lines, diagnostics, etc. come through Acceleration Division controls VMEs and IRMs located in MI-31.
The Electron Cooling beam line is made up of five types of magnets: correctors, dipoles, quadrupoles, and two types of solenoids.
Correctors
To be written
Dipoles
To be written
Quadrupoles
To be written
Solenoids
To be written
To be written
The Cooling section is the region of the ECool beam line where the Pbars that are spinning around in the Recycler interact with the electrons and are cooled longitudinally. The cooling section is made up of two sections: the Before-Cooling section and the Cooling section.
The Before-Cooling section is very short. It begins after the 90 degree dipole bend magnet DXS6A/B and ends at the first Cooling section solenoid SPC00. This section is very important because it is where the angle and position of the electron beam are determined prior to entering the cooling section. The Before-Cooling section is comprised of two solenoids (SPB01 and SPB02), two sets of dipole correctors (X & Y), and two sets of BPMs (X & Y).
The Cooling section is where the longitudinal cooling of the Pbars takes place. It begins with the first solenoid station SPC00 and ends with the last solenoid station SPC90 for a total of 10 solenoid stations. Each solenoid station is just over 2 m long and contains an ion pump, a loss monitor, a scraper, a BPM set, and a cooling solenoid. Inside each cooling solenoid are correctors, labeled 0 through 9, in both the X and y planes. These correctors are set to very specific values based on what the magnetic field is in the cooling section and should not be changed. These solenoids provide strong focusing with small transverse angles. The magnetic field in the cooling section can be set as high as 190 gauss, but we run with a field of 105 gauss for electrons, with an energy of 4.32 MeV. The radius of the beam through the cooling section is 3-5 mm, with rms angles of 0.2 mrad. The Cooling Section solenoids are LCW cooled with MI LCW.
The ECool BPM System
The BPM system used by ECool is a very unique system. It was essential to have a system that would very accurately measure the relative position of the electron and Pbar beams to within +/- 100μm of each other in the cooling section. In order to separate the signals from each beam, the electron beam is modulated at a frequency of 32 kHz. The digital signal receiver (DSR) processing electronics has the ability to discriminate between different signal frequencies by down-converting an input frequency to base band using a numerically controlled oscillator. The circulating Pbar beam is detected at the revolution frequency and the electron beam is detected at 32 kHz.
The operational BPM system located at MI-31 consists of two VXI crates. The operational system node names are ECBPM and ECBPMA. All devices coming through these two operational nodes will be R:* devices. The device nomenclature is as follows: the first letter R:B* denotes that it is a BPM; the next letter, R:BX* or R:BY* tells which plane the BPM is showing data for, X or Y; the next letter tells which ECool beam line the BPM is in, R:BXA* for acceleration line, R:BXS* for supply line, etc.; the next two digits tell at which solenoid station each BPM is at in each beam line, R:BYR01* is a BPM showing the Y-plane in the return line at the first solenoid station; the last letter tells what type of BPM read back you are receiving, R:BXC10I is the intensity read back, R:BXC10S is the slow, low bandwidth BPM position data in mm, and R:BXC10F is the fast, high bandwidth BPM position data in mm. Each BPM also has an associated clearing voltage denoted by R:BC***, for example R:BCA05. The developmental system node name is ECBPMD. All devices coming through this developmental node will be Z:* devices.
The console application for interfacing with the ECool BPMs can be found on ACNET parameter page E50, or can be accessed by opening up any machine BPM page, like R39, clicking between the yellow diamonds at the top of the page and choosing E-COOLRR, for the operational BPM system. The developmental system will have the name E-COOLDV and is only used by LLRF experts.
The BPM display program will display with a menu bar shown at the top and 4 sub-windows as shown in fig. 1. There are five entries on the menu bar: DATA DISPLAY, PLOT SETUP, ARCHIVE, TIMING and UTILITY. Each menu item has a corresponding sub-window shown under the menu bar except the UTILITY item. At the bottom of the window, there is a message window.
The basic function of each sub-window is similar to other machine BPM display pages. There are some differences that are ECool BPM specific however:
- DATA DISPLAY – enables user to select which beam (electron or circulating Pbars) to probe, the type of filtering to use for the position data to display and the operational mode to use in the display. This will be described further in the next section.
- PLOT SETUP – enables the user to select the position data orientation (horizontal or vertical) to display, the plot destination and scaling for position and intensity.
- ARCHIVE – enables the user to save the generated plot in a file.
- TIMING – enables the user to change the frequency used to detect either the electron or circulating Pbar beam.
- UTILITY – enables the user to perform ECool-specific configuration. This window is initially hidden and displayed only when the UTILITY menu is selected; selecting this sub-window hides all other sub-windows.
Fig. 1 ECool operational BPM program
Data Display sub-window – Operational Modes
There are four beam position measurement modes:
1. Electron beam position measurement mode;
2. Circulating beam position measurement mode (for measuring Pbars);
3. Switched beam position measurement mode (Both);
4. Electron pulsed beam position measurement mode.
The user can invoke any of the above modes from the DATA DISPLAY sub-window. Within this sub-window there are two pop-up menus: BEAM and DATA. The BEAM menu selects the beam type to measure; electron, circulating Pbars, or both, which corresponds to Switched beam position measurement mode. Electron pulsed beam mode will be discussed later. The DATA menu option, SLOW/HiRes or FAST/LoRes, selects the filtering to apply to the position data. The SLOW position data is generated from a narrow band filter (BW < 5 Hz), and is an averaged output of the FAST positions; this reduces the noise in the position data, relative to the FAST position data, at the expense of a longer response time. The user can change the tuning frequency for the corresponding measurement from the TIMING sub-window. In general, the tuning frequency is not necessarily the same as the modulation frequency. The user can move the tuning frequency off the modulation frequency to avoid saturation in the DSR modules. A common indicator of saturation is high beam intensity and unstable fluctuation of beam position measurements. During normal cooling operation, this frequency is automatically set. These timers should not be changed without expert approval.
While in the electron or circular measurements mode, the ECool BPMs provide horizontal and vertical positions for either electrons or circulating Pbars. The timing of the measurement is controlled by the TCLOCK event $EB. This event can be tied to other TCLK events and delayed from these events by a specified amount using the device R:ECDLY, thus allowing the user to accurately time data collection relative to other accelerator events. Once the $EB event is received, the front end collects data from the DSPs for all BPMs and maps that to the ACNET BPM display program (E50). In the switch mode, the BPMs alternate between measuring beam position for both beams. The frequency with which the measurement alternates between electron and circulating Pbars can be changed from the TIMING sub-window.
Once the measurement type, BEAM and DATA has been set, a click of SNAPSHOT in the DISPLAY DATA window invokes the corresponding measurement mode. The selected mode remains until the user selects another mode. The message window will display the progress of the operation, as well as any errors encountered. Messages from different sources appear in the message window with different colors; messages from the ECOOLBPM system will appear in yellow. The instantaneous beam positions measured for all BPMs in the system will be shown in a new pop-up window (see figs. 2 and 3). Continuous beam positions are also available as ACNET devices, which can also be shown on fast time plots. The ranges of operational BPM intensities large enough to give reliable position information are as follows: for DC beam in either Electron or Circular mode need to have a BPM intensity of at least 150; for pulsed beam in pulsed operating mode need to have a BPM intensity of at least 10.
The electron pulsed measurement mode can be invoked by clicking on PULSE. There is a time delay between the electron pulse TClock event trigger and data acquisition due to the digital signal processing. The data acquisition is triggered by the TCLCK event $EB. It is important to check the status of the ECool Variable Pulse Delay ACNET device – R:ECDLY. This device should be on. When R:ECDLY is off, a red asterisk will be shown next to it on an ACNET parameter page.
A click on the UTILITY menu bar item pops up a window that experts use to perform a system calibration.
The ACNET devices for the ECool BPM system can be found on page R117, under DIAGNOSTICS.
Below are the graphics that the user sees when clicking on SNAPSHOT. Figure 2 shows the electron beam trajectory on axis (with no vertical position offset through the cooling section); Figure 3 shows the Pbar trajectory through the cooling section.
Fig. 2 Electron beam trajectory.
Fig. 3 Circulating Pbar beam trajectory through the cooling section.
To be written:
Scrapers
Flying wire
Loss monitors
Pepper pot
YAG
Multiwire
Optical Transition Radiation (OTR) Monitors in Ecool
Optical transition radiation monitors are being used to image the transverse profiles of the 4.3 MeV electron beam used in the electron cooler at Fermilab. The linear response of OTR monitors to beam charge, the high spatial resolution and the fact that the radiation is prompt has several advantages over more traditional imaging devices. The transition radiation is produced by the charged particles as they traverse the boundary between media with different dielectric constants, for example a metallic or dielectric foil in vacuum. For relativistic electron beams, transition radiation can be measured at optical wavelengths (OTR) with readily available equipment and imaging techniques. The imaging system used in the FNAL set-up consists of digital CCD cameras connected to computers via IEEE 1394 fire-wire interfaces. This provides the operator with real-time beam images and tools for image analysis and measurements of the beam’s dynamic properties as a function of the optics of the accelerator. The OTR diagnostic systems used at the cooler are designed to be routinely used to optimize the beam transport and to measure the transverse beam size and shape with a resolution down to 25 µm. Very thin (5 µm) Aluminum foils are being used to reduce the background signal due to beam scattering and bremsstrahlung radiation.
OTR Beam Profile Monitor Set-up
The OTR beam profile monitors are composed of several elements; a stepper motor actuator that inserts the OTR foil into the path of the electron beam, an optical lens with CCD camera to image the beam, and a Labview based data acquisition and processing system to deliver images and beam parameters to operators. The OTR foil is inserted into the beam at an angle of 38.9 degrees to the beam direction. This corresponds to the angle θ ≈1/βγ, where the intensity of the light is a maximum. A schematic layout of the system is shown in Figure 1 along with a picture of the OTR foil mounted on its frame. To transform images to real-world coordinates four 75 µm tungsten wires are mounted over the radiation screens to form a rectangular 10mm x 10mm grid.
Figure 1.
Six OTR stations exist in the ECool facility, four of which are located in the MI-30 tunnel at locations R03, R05 and R06, and 2 are located in MI-31 pit under the acceleration and deceleration sides of the machine at locations A07 and D07 respectively. The OTR at location A07 is used only when the electron beam is running in U-Bend mode; all other OTRs are used in full line mode. These devices are primarily used during studies in pulsed mode and are not used during regular operations.
Vacuum System
The ECool vacuum system is the same system that is used throughout the Main Injector with the exception of the fast acting valve system. All of the readings come through a VME node called ECVAC, located in relay rack 202 in MI-31. This rack also contains the vacuum CIA crate and the ion gauge power supplies. Rack 201 contains all of the ion pump power supplies.
Inside the tank, the vacuum is kept at roughly 10-9 Torr. There are ion pumps located on the gun, at deck level 3, and on the collector. There are also manual vacuum valves located under the tank at locations A06, A07, and D07. When there is a full discharge of the high voltage inside the tank, the vacuum will rise very rapidly. It can reach as high as 10-5 Torr. When this occurs, the electron beam cannot be run until the vacuum recovers. The recovery will take anywhere from 15 minutes to hours, depending on how bad the voltage discharge was. If the vacuum pressure rises too high, the fast acting valve system will close the valves connecting the Ecool beamline with the Recycler.
In the beamline sections, vacuum is kept roughly at10-10 Torr. There are ion pumps at every solenoid location. There are manual vacuum valves located at the following locations: S02, Q04, and T05.
The Pelletron uses sulfur hexafluoride (SF6) inside the tank because of its properties as an electrical insulator. The SF6 is pressurized to ~75 psig. A system of fast acting beam valves has been installed in the event of a catastrophic vacuum failure that would cause the SF6 to flood the beam pipe and possibly contaminate the Recycler. In the event of such a failure, the SF6 would reach the Recycler beam pipe and contaminate the entire Recycler vacuum system in just over 5 seconds. The fast acting valve system also protects the Recycler vacuum system against high pressure vacuum bursts resulting from full discharges in the Pelletron that could rupture vacuum components.
Two fast-acting valves are located on the MI-31 side and two on the Recycler side. The function of the first set of valves, one in the Supply line and one in the Transfer line, is to minimize the amount of SF6 gas that will reach the Recycler. The second set of fast valves, located on either side of the cooling section in the Recycler, will stop the flow of gas from reaching the rest of the Recycler vacuum system. The fast valves are activated by what are known as VAT controllers, which are triggered by the VAT cold cathode gauges when a pressure of 5 × 10-4 Torr is reached. The valves will close in 32 ms. The valves on the MI-31 side (BVS04F and BVT03F) are controlled by the cold cathode gauges and controllers on that side. The Recycler fast valves (BV304F and BV308F) are controlled by the gauges and controllers on the Recycler side. A trip on the MI-31 side will not affect the Recycler valves. Likewise, the Recycler fast valves will both close if either VAT gauge on the 90 degree bend magnets in the Recycler read greater than 5 × 10-4 Torr.
The four fast acting valves are not completely leak tight. Thus, each one is backed by a standard pneumatic valve, which is activated through the CIA crate vacuum interlocks. They take a few seconds to close. These valves will close if three or more of the four selected ion pumps on either side of the valve trip off. In the MI tunnel, these valves are named BV304S and BV308S. In MI-31 they are BVS04S and BVT03S. The ion pump permits will close the fast and slow valves. To get the permits back, first a reset and on command must be given to the slow valves, then to the fast valves.
Two vacuum burst disks are located on the 90 degree bend magnets under the Pelletron. In the event of vacuum system over-pressurization, the disks will rupture, relieving the pressure in the system.
Water cooling system
The Pelletron uses three different water systems for cooling. The first is the cooling water skid located in MI-31 next to the elevator mechanical room. This water cooling skid is a closed loop system that was originally filled with LCW from the Main Injector system. If it needs LCW added, it would be taken from the MI system. This skid utilizes a Freon heat exchanger that is located behind the MI-31 building inside the fenced area. This fenced area is classified as a High Radiation Area when the Pelletron is running and RSO approval is needed to access it. There are only a few devices that are cooled from this skid. They are split between two systems, the Silo magnet water system and the SF6 Heat Exchanger water skid. The Silo magnet water system cools:
- The U-Bend magnet under the Pelletron tank and its associated power supply, located under the stairs next to the MCC;
- The solenoid doublets SDS01 and SDT06 inside the two 90 degree dipole bend magnets located under the tank (the dipole magnets themselves are air-cooled);
- Two of the four Walker solenoid magnets SPA07 and SPD07 (the other two Walker solenoids, SPA06 and SPD06 are inside the tank and are SF6 cooled).
The SF6 Heat exchanger water skid cools:
- The SF6 recirculation skid exchanger located on the platform next to the Pelletron tank.
The Acnet parameters that control these two water systems can be found on R117 Utilities, subpages 13 and 14. There is also a document that describes operation of these systems: Electron Cooling (SF6 & Pelletron) Water System Information.
The second water system used by ECool is called the Collector Cooling system. It is located inside the Pelletron tank on top of the collector, and is used to cool the collector. It is a closed loop system that uses deionized (DI) water. To fill this system, the Pelletron tank must be accessed, so the SF6 would have to be transferred to the holding tank outside. The device that monitors this is R:COLFLO. It should be noted that when the rotating shaft is off, there will be no readbacks for this device, so one can only tell if this system has a problem when the shaft is running.
The third water system used by ECool is the Main Injector LCW system. This is used to cool the solenoids in the cooling section. There are 10 solenoids in the cooling section that use MI LCW. Their temperatures are monitored by thermocouples named R:TCC00, TCC10, TCC20, and so on to TCC90.
Operation of the ECool Electron Beam
There are several programs that are used to help make the operation of the ECool beam easier and more efficient. The most useful of these are the sequencer aggregates. We have designed the ECool sequencer so that all that is needed to set up the ECool beam for cooling is to run certain aggregates. The ECool sequencer is on ACNET page E48 (see fig. 1). The main sequencer aggregates that are used by Operators are listed below:
- Pelletron OFF short access (to MI-31),
- Turn on from short access,
- Turn beam on for ecooling,
- Shift beam in Cooling Section,
- Regulate ebeam current,
- Energy adjustment
- Recover from full discharge
- Check vacuum valves open.
There are also a number of aggregates that are used by experts. Operators should only manipulate the ECool beam and Pelletron through the sequencers unless otherwise instructed by experts.
Fig. 1, ECool Sequencer
It should be noted that there is a device that is the equivalent of a master switch, R:PELSEQ. Turn this device OFF and the electron beam will turn off. It can then be turned back on through the normal sequencer once the controls problems have been cleared.
ECool also uses a number of JAVA programs that have some parameters that are controlled through the sequencer aggregates. Two Autotune programs are used to keep the beam fixed at the entrance to the cooling section and the collector. These Autotune programs monitor the BPM positions and adjust correctors at certain locations to maintain the correct trajectory. The first one is the Acceleration side Autotune (see fig. 2); it keeps the beam fixed going into the cooling section. The second is the Deceleration side Autotune (see fig. 3); and it keeps the beam centered going into the collector.
Fig. 2, Acceleration side Autotune
Fig. 3, Deceleration side Autotune
A dedicated program is used to monitor the machine’s critical parameters and keep the beam on and running at predetermined set points. This program is called the “Crash Recovery Finite State Machine”. This program compares the conditions that are present when the beam is running to a nominal set of parameters (found on ACNET page R117 Operation subpage 9, see fig. 4) and will turn the beam off if it deviates from these nominal parameters.
Fig. 4, Nominal parameters set into the Crash Recovery FSM
The parameters that are checked include: beam current (controlled by PARM00 and PARM02), high voltage (PARM01), and vacuum (PARM09 and PARM10). This program is also designed to turn the Pelletron off in a safe manner after a specified number of trips have occurred (PARM03 and PARM04). When this happens, it should be investigated to see why so many trips have occurred, as well as what the trips are, before turning back on. The program that would assist in diagnosing these trips is called the Pelletron Mode Controller. The Pelletron Mode Controller program monitors the loss monitors located throughout the ECool beam lines, as well as certain voltage readings inside the tank.
The Pelletron Mode Controller has four subpages that could be helpful to the Operators in diagnosing trips. The first page, the Box Status shows which input caused the trip. It should be noted that not all inputs that can trip off the Pelletron come through the Mode Controller. Fig. 5a shows what it looks like during normal running; Fig 5b shows what it looks like when the Pelletron is turned off and the tank has been opened.
The first line, IRM Trip Channel, will tell what the IRM channel is that has caused the trip. It will latch on the first occurrence and will not clear until a reset has been sent to the Acnet device R:PELTRP. It will be either a loss monitor or a voltage readback from inside the tank.
The second line, IRM Error Channel, displays the number of the channel whose reading went outside the limits set up for that channel. The IRM Trip Channel latches the IRM Error Channel.
The third line, Pelletron Klixon Interlock, monitors the temperature inside the tank with Acnet device R:T4LVL as the maximum temperature and R:TNKT4 as the current temperature. If R:TNKT4 goes above R:T4LVL the mode controller will turn the beam off.
The fourth line, Beam Permit, monitors the critical device controller R:PELCDC.
The fifth line, IRM permit, monitors the IRM and will turn off if there is either a tripped channel or error.
The sixth line, IRM Heartbeat, monitors the IRM itself to make sure the signal from the IRM is present.
The seventh line, Fast CPO, monitors the fast signal coming out of the CPO (Capacitive Pick-Off) and will trip if this signal is too large.
The eighth line, Master Permit, is controlled by the device R:PELTRP. This Acnet device will display a Hex number of the IRM Trip Channel. The mode controller will not allow beam to be turned back on until R:PELTRP has been cleared and reads back 0.
The ninth line, Sequencer Permit, is controlled by Acnet device R:PELSEQ. If this device is turned off, the beam cannot be started. This device is the software version of a master switch.
The tenth line, Motor Generator, monitors the motor control center in MI-31 and will turn the beam off if there is a problem.
The eleventh and last line, Fast Cathode, monitors the cathode for fast spikes. It is enabled/disabled using the Acnet device R:FCTRIP. We normally run with this enabled.
Fig. 5a, Operating normally
Fig. 5b, Pelletron is turned off and tank has been opened.
The second page is called the Sequence page (see fig. 6). This page tells you what mode the Pelletron is using, DC Beam or Pulsed Beam, and which template is loaded into the mode controller. In Fig. 6, you can see that the Mode is ‘DC Beam’, and the template loaded is ‘Return line 8V’. This also shows what is in this template by channel number.
Fig. 6, Pelletron mode controller, Sequence page
The next page is the Template page (see fig. 7). This is the page where the template is set up. The channel assignments are set up here, as well as the limits for each channel. A password is required to make changes to this page, and should only be changed under expert supervision.
Fig. 7, Pelletron Mode Controller Template page
The fourth page is the Status Page (see fig. 8).This is the page that will show the individual channels with the trip limits in graph form. When there is a trip, go to this page and click once on the box labeled ‘IRM Front End’. This will go out to the IRM and retrieve the most recent data. The IRM buffer will stop when there is a trip and won’t start collecting data again until the device R:PELTRP is reset. To see what the tripped channel was doing at the time of the trip, click on the box next to the word ‘Select’ labeled in Fig. 5 as 00)Cooling Long Corrector 1I. This will pop up a box with all of the channels and simply click on the channel you wish to see. It will be displayed in the window.
Thanks to A. Shemyakin, S. Nagaitsev, L. Prost, K. Carlson, A. Warner, C. Schmidt, G. Saewert, C. Gattuso, P. Joiremann and M. Sutherland for sharing their knowledge of the Fermilab Electron Cooling Project.
S. Nagaitsev, Introduction to Electron Cooling.
S. Nagaitsev, Electron cooling Project at Fermilab.
National Electrostatics Corp., FermiLab 5URe-2 Pelletron System.
AD Mech. Support Dept., Engineering Specification 1302-ES-296 422, Recycler/Electron Cooling, SF6 Vacuum Protection System.
P. Joiremann, Electron Cooling BPM User’s Guide
A. Warner, et al, OTR Measurements and Modeling of the Electron Beam Optics at the E-cooling Facility.