Accelerators – Reports 2


Technical Developments

Beam phase measurements

New phase measuring equipment is planned for the separated-sector cyclotron. The first of twenty new phase probes has been installed, near the extraction radius in the south valley vacuum chamber. The probe has been used to investigate the possibility of using a commercial lock-in amplifier, with amplitude and phase measuring capabilities, and cancellation of the pick-up signal, without beam, with a harmonic generated from the dee voltage, before vectorial subtraction of the resulting signal from the beam signal.

Beam phase measurements have up to now been done by displaying the amplified beam signal from a phase probe from one of the two multi-head probes on an oscilloscope, triggered by a signal derived from the dee voltage, and noting the time while the probe is driven from the injection to the extraction radius. Although the accuracy of the method is sufficient it is time consuming and the probe support also intercepts beam, which means that the beam phase in the cyclotron is not known during operation. New, non-interceptive phase probes with greater sensitivity are therefore planned.

In the layout of the test setup for phase measurements, shown in Fig. 1, the combined signals from the top and bottom plates of a phase probe are amplified and filtered before addition to an harmonic of the dee voltage generated by an Analog Devices direct digital synthesizer chip AD9952 mounted on a development board. Two signals with independent phase and amplitude adjustment by computer control are available. One signal is used as a reference for the lock-in amplifier, and the other for cancellation of the pick-up signal, by adjustment of the amplitude and phase of the signal from the synthesizer, to minimize the voltage displayed on the lock-in amplifier, without beam. Accurate phase and amplitude measurements are therefore not required for cancellation of the pick-up signal, except that such measurements are useful to calculate the approximate phase and amplitude settings of the AD9952 synthesizer before final adjustment, in order to speed up the process. The harmonic of the dee voltage can also be selected by computer control. The results of some beam phase measurements with the fixed phase probe at extraction, by cancellation of the pickup signal only and without subtraction of the pick-up signal from the signal with beam, are shown in Table 1.


Figure 1: Layout of the test setup for beam phase measurements.

Top of Page


Beam current





















Table 1: Beam phase measured at the second harmonic with the non-interceptive probe at extraction


The Berlin HMI ECRIS


The ECR ion source and the low-energy beam line of the Hahn-Meitner Institute (HMI) in Berlin, Germany, which were donated to iThemba LABS, were dismantled and shipped to South Africa. This source (see Figure 2 below) consists of a water-cooled plasma chamber (length 18cm, diameter 7cm) surrounded by FeNdB permanent magnets that produce a hexapole field of 1T (at the wall of the chamber) for radial plasma confinement. Two solenoid coils produce an axial field that confines the plasma axially. The field on the axis typically varies from 0.4 to 1.1 T. Microwave power is coupled to the source via a wave guide. The generator can deliver up to 2 kW of microwave power at a frequency of 14.5 GHz. The source is designed to operate with oven- or sputter target techniques, but was used in Berlin mainly for producing beams from hydrogen, nitrogen, and noble gases. Typical values of intensities and charge states for argon are: 25, 6, and 3 eµA for Ar11+, Ar13+, and Ar14+ respectively. These values demonstrate the advantages of this source compared with the existing ECR source at iThemba LABS.


Because of the different beam heights at iThemba LABS and the HMI a new support structure was designed and constructed. The source with its extraction system, the solenoid lens, and the first diagnostic chamber are placed on it. The 900-bending magnet is mounted on a separate stand. Connection to the existing Q-beam line will be done with the second diagnostic chamber, and an Einzel lens positioned in the entrance of the bending magnet BQ1. Each diagnostic chamber contains horizontal and vertical pairs of slits and a Faraday cup. The ion source and all beam transport components were aligned using an optical telescope system. Installation of the necessary infrastructure, including water cooling, compressed air, and electrical connections, has been started, and it is expected that the source and beam lines will soon be evacuated. The first beam experiments are not expected to take place before the end of the year. A new floor for the power supplies feeding the source coils and the solenoid has to be built on top of the existing ECR vault. Drawings have been finalized and construction is scheduled for the shutdown period in July 2007. The additional AC power required to feed the power supplies will be provided from the new transformer, which was ordered to increase the power capacity from 4 to 6 MW.



Figure 2.  The ECR ion source from Berlin on the left of the stand

Top of Page


New ECR ion source

The existing 23 year old ECR ion source is in the process of being replaced by a new modern room temperature source, which is a replica of the CERN ion source that was built by the CEA in Grenoble. Various companies in Europe constructed parts of the source. The greatest beam current for different charged states of this source compares favourably with the 28 GHz superconducting VENUS source at LBL, Berkeley.

The source will be coupled to 14 GHz and 18 GHz microwave generators. Provision is made for two ovens. The axial field can be varied between 0 .5 T and 1.2 T and the radial field will have a value of 1.4 T using FeNdB permanent magnets. The source will deliver a beam current of 60 eµA Xe 30+ ions.

The coils and permanent magnets have already been delivered. The RF generators and the mechanical parts will be delivered within the next month. Figure 3 shows the two RF generators at the construction site. The source that is presently in use will be dismantled after the successful commissioning of the two new sources.



Figure 3. The two RF generators being tested at the construction site.


Top of Page

Halo monitors

Air-filled ionization chambers, similar to those in use at the Paul Scherrer Institute in Switzerland, have been installed at 2 metre intervals in the high-energy beam lines leading to the radioisotope production vaults, to detect stray-beam leaving the beam pipe. Each detector consists of two printed-circuit boards separated by 10 mm and with opposing copper-plated surfaces. The ring-shaped copper surfaces have inner and outer diameters of 55.8 mm and 85.8 mm respectively. On one board the copper ring is divided into quadrants to indicate the position of the stray beam. The boards have been designed such that they can be clamped onto the outside of the beam pipes without removing a pipe section. To detect stray particles the quadrant currents induced by a 1 kV bias voltage are measured using integration techniques, and displayed with a computer program called Halo. The system can measure 48 channels concurrently with an accuracy of a few pA. At the centre of the electronic system is a rabbit microprocessor that controls the components. The computer program communicates over the network with the microprocessor, using TCP/IP. The program displays the data graphically and controls the current range. The colour of the graphs changes when certain important conditions are met. Data from 8 stray-beam detectors are displayed. Using a client program the visual interface can be seen from any computer on the network. The stray-beam detectors now form part of the safety interlocking system for protection of beam pipes and peripheral equipment. Stray-beam detectors, using PIN diodes, have been installed in the beam lines leading to the experimental areas.


Beam statistics

The performance of the cyclotrons for the last 12 years is shown in Table 2 below. Beam on target time has remained roughly constant. The scheduled beam time for the past calendar year was nearly 80%. The RF systems have usually been the main source of interruptions; however they were overtaken this year by power failures, which caused 5.87% of the lost time. This was closely followed by water leaks from corroded channels in the extraction components of the SSC. These leaks occurred from time to time (4 times in 2006), resulting in 3.97% of lost time. Other equipment failures, including RF components and chillers, caused 7.55% of lost time.


Year Beam availability during the year: % of scheduled beam time for:
% of total time % of scheduled* time Energy changes Interruptions
1995 72.95 82.91 5.67 8.46
1996 69.69 78.21 9.30 8.92
1997 67.63 77.31 11.02 10.60
1998 66.93 75.55 13.20 9.73
1999 69.12 78.82 9.99 10.81
2000 58.51 73.07 9.36 15.50
2001 66.13 78.70 6.30 12.61
2002 72.29 82.69 7.50 7.28
2003 70.93 82.79 6.87 8.08
2004 72.0 84.9 6.7 5.9
2005 71.3 83.6 5.5 6.4
2006 66.1 80.3 5.5 7.9
Scheduled time is total calendar time minus scheduled maintenance time and the days during which the laboratory is officially closed during December.

Table 2: Beam delivery statistics, 1995 to 2006


Top of Page