12 INCH CYCLOTRON CONTROL AND DAQ SYSTEM
Perhaps the system of most complexity is the CONTROL and DATA AQUISITION SYSTEM (DAQ for short). This is essentially the nervous system of the cyclotron. It consists of the control computer, which as the name suggests, controls all aspects of the cyclotron's opertation. In addition to controlling the entire system, the computer monitors many components and writes their values to a "DATA FILE." The computer connects to most instruments via General Purpose Interface Bus - GPIB (also know as IEEE488 or HPIB - as HP developed the standard) and RS-232 (standard serial communications). The GPIB buss is a cable that connects at PCI card in the PC to each instruments GPIB port, each instrument has a unique GPIB address number.
Fig.1a Screen capture of "Big Board" and the operators control display in normal operation.
Fig.1b Screen capture of "Big Board" and the operators control display under emergency shutdown.
The control rack was laid-out with the control operator in mind. Two flat panel LCD screens display all the statistics and values during operation. The top LCD screen is reserved as a display-only board, known as the "Big Board" (stealing it's name from Stanley Kubrick's movie "Dr. Strangelove"). The Big Board has two types of displays, annunciators and a time-domain chart recorder. First, the annunciators follow the submarine "Christmas Tree" style: GREEN = OK, YELLOW = A hazadous condition exists, and RED = A Failure or Dangerous Situation exists. A quick glance at the annunciator portion of the screen is all that is needed to ensure everything is working within safe limits. Currently listed annunciators are: water flow, hydrogren flow, magnet coil temperatures, mechanical pump temperature, difussion pump temperature, RF Power On, Magnet Power On, RF cooling system. The second portion of the Big Board is time-based chart recorder with a 60 second span. The vertical scale of the chart recorder can be selected from a list of parameters. A few of the parameters are: Cyclotron Beam Current, Magnet Current, RF Power (FWD and REF), etc. This allows the operator to make adjustments at the magnet or other far distances while watching the response on the chart recorder. The chart recorder is also useful when when demonstrating the operation to visitors, they can stand at a relaxed distance and still watch the data update.
Fig.2 View of control rack from the operators position.
The bottom of the two displays is the operator's control panel or Operator's Screen. Being very densely populated, this screen intakes and displays all the desired input values, i.e. magnet current, this screen also contains all the "boolean" controls, i.e. pump switches, all RF parameters are controlled, and all the data file managment is handled here. In short, the instruments have virtual controls on the computer screen. Because of the complexity, most of these controls are automated in the code, but have manual overides if the operators so desires. All of the data displayed on the Operators Screen is saved to a file for further analysis later. This file provides a nice log to refer to when there is a failure of some type - this extensive record makes the pathology much easier.
For those interested at a deeper level in the DAQ Software, we have provided the "blue print" of the coding.
Features of the control rack outside of the computer are also designed for the operator's ease. Equipment that does not need any or much attention is placed the furthest away, such as the broadband solid state RF amplifier. While equipment that requires some dithering, such as the filament bias supply is at an arms reach. The magnet power supplies were awarded the bottom position as they were the heaviest.
Fig.3 Rear view of control rack - note fan out of all signal cables at top-right.
For the many low-voltage signal applications, i.e. pump relays, thermocouples, current shunt voltage measurements, we use two HP Data Aquistion and Control Units. They each consist of a HP3497A mainframe that is customised to our application by installing needed cards. This is a truly versatile system, each card providing many channels of analog-out voltage cards, analog signal measurement cards, digital I/O, low current relays, and Type K thermocouple inputs. Each channel has a unique address allowing specific commands to appropiately read or write over the GPIB network. One of HP3497A's is mounted at the magnet table and primarily reads temperatures and controls the Mass Flow Controller, the other HP3497A is mounted at the control rack and interfaces all other low voltage signals. The cables leaving the various DAQ/control cards are routed up to a set of terminal strips in the rear of the control rack, where connections are easily made and removed as needed.
Fig.4 Rear of a HP3497A. Dozens of low voltage and digital I/O signal cables enter and are converted to GPIB (right).
If the DAQ system is nervous system of the cyclotron, then the cable harness connecting the control rack to the magnet table is to be likened to the spine. This harness is called the "umbilical cord" and carries the cabling for the DC magnet power, the RF power, beam dianostic signals, the low voltage signals to and fro, an RS-232 link, and a GPIB link. To make the system as flexible as possible, the umbilical cord fans out to a series of terminal blocks at both ends. As components are installed and removed, the cabling does not have to be disturbed, so far there has only once been a need to increase the cable capacity. Since then ample cabling has allowed for many unforseen improvements. A benifit of the fan outs has been seen in troubleshooting: easy access to terminals allows quick DVM or oscilliscope measurements. The "map" defining all of the connections is perhaps most critical document of the cyclotron, and it is mounted inside the rear of the control rack directly next to the fan out (of course a duplicate is in the log book).
Fig.5 Close-up of umbilcal cord's low-voltage portion fan-out at control rack.
Fig.6 Umbilical cord from the control rack fans out at the rear of the magnet table.
Efficient utilization of space was a key part in the design of the 12-inch cyclotron. The magnet table was needed to raise the cyclotron's working height to a comfortable level, however, a resulting feature was the space underneath the table to mount equipment. Because it was easier to run data cables than bulky AC and sensor cables, it was decided early on that the entire vacuum system (from pumps to gauge controllers) would reside under the table, and as a consequence, all of the thermal telemetry would also reside there. Four 19-inch racks were built into the table for ease of mounting equipment - two in the front and two in the rear. Having the vacuum monitoring system located at the magnet table reduced power and space demands at the control rack - already a tight fit.
Fig.7 A front fish-eye view of controls mounted at the magnet table.
Behind the clear and easy accessable magnet table controllers resides a densely packed network of signal cables, RF cables, power cable, cooling lines, several pumps, vacuum lines, pressure lines, and a tank of HV cooling oil. During initial cabling we could sit underneath the table and have our work in front of us. Now at full capacity acessess is considerably reduced. There is alway room for more if need be.
Fig.8 A view underneath the table reveals a dense network of AC and DC power, RF power and signal, and low voltage signals cables.
Return to home .