Machine Operation and Development

Figure 1. Mean Time Between Failures (blue bars, left axis) and Uptime (red curve, right axis) since the beginning of operations.

The annual operating statistics are shown in Figure 1. The overall mean time between failures (MTBF) continued to be good at 98.4 hours, and each of the five operating runs during the year achieved individually a MTBF in excess of 72 hours, which is the target minimum. The 97% uptime was however below the target of 98% for many reasons including various new faults which consequently took longer to rectify, as well as an increased amount of scheduled downtime.

Using Diamond as a test bed for Diamond-II commissioning

In addition to supporting Diamond operations, machine development periods are frequently used for testing the procedures will be used for commissioning Diamond-II. These studies can provide valuable insight into the effectiveness of different methods and highlight real-world effects that are typically absent or neglected in modelling codes. This can be anything from the impact of hysteresis in magnetic fields, charge-dependence of diagnostic equipment or leakage of the RF cavity fields into the adjacent BPM buttons.

The first method tested was to maximise beam transmission of the injected beam from the booster into the storage ring. The acceptance of Diamond-II is anticipated to be substantially smaller than Diamond, so accurate setting of the dipole steerer magnets to allow beam transmission around the ring forms a vital first step. Distinguishing real signal from the electron beam above the measurement noise floor and adjusting the trajectory of the incoming beam are significant obstacles that must be overcome.

Figure 2: Comparison of orbit response matrices for the Diamond storage ring measured at 10 mA stored beam (left) and from shot-by-shot measurements and 0.5 mA injected beam current (right).

As well as adjusting the electron beam steering, corrections to the quadrupole focussing strengths must also be carried out. Standard techniques exist to perform this task, however, in case of difficulties, alternative methods are being explored to achieve this goal. The first is to quantify how the electron beam transverse oscillation frequency (betatron tune) changes as a function of quadrupole strength, the second is to measure how the electron beam trajectory varies around the ring in response to a change in dipole steering magnet strength. These methods are complementary to each other but each face individual challenges and are subject to shot-to-shot fluctuations and measurement noise. Fig. 2 shows clearly the additional noise that is introduced from shot-by-shot measurements compared to measurements with a stored beam. Adapting the fitting routines to overcome the increased measurement noise and to provide a preliminary correction to the machine optics is currently under study. These investigations provide essential input into the procedures that are being developed for the future commissioning of Diamond-II.

Figure 5. Output from the klystron (left) and input to the bunchers and accelerating structures (right) following the SLED cavity. The RF pulse width from the klystron is 5 μs long while the higher power pulse going to the linac structures is 1 μs long.

Successful operation of the linac at 100 MeV has been demonstrated on several occasions with a single klystron.

Figure 6. Replacement analogue-to-digital controller card

Overcoming obsolescence and improving power supply reliability

The number of beam trips due to power supply faults remains very low, however an increasing number of operational problems with the analogue-to-digital (ADC) controller cards used in the power supply control system, as well as obsolescence, has required that an alternative be developed. The replacement cards (Fig. 6), developed in-house over several years, have higher precision and can be reprogrammed in the future to work with Diamond-II power supplies. Approximately 300 of the 1,300 installed cards were replaced during the year. The remaining 1,000 cards will be replaced during 2024/25.

Figure 7. The location of the new temperature sensors in one accelerator cell.

Measuring the synchrotron tunnel temperature in more detail than ever before

The Diamond electron beam can be as small as 4 μm in size vertically at the X-ray source points. If the electron beam moves, then the resulting X-ray beam will also move. We aim to keep the electron beam position stable to 10% of the beam size. For Diamond-II, these stability requirements are even more stringent. To achieve this, the ambient temperature in the tunnel must be finely controlled so that thermal expansion of components doesn’t impact X-ray beam stability. As part of the Diamond-II preparation work, a new network of over 260 temperature sensors has been installed in the synchrotron tunnel to help us better understand the temperature stability.

Over the past six months a project to install air temperature sensors located every two metres all the way around the synchrotron has been underway. Figure 7 shows the locations of the sensors in one accelerator cell. This project was completed in January 2024, in time to allow us to see the temperature of the storage ring throughout the first user run of the year in unprecedented detail. Before this installation, Diamond only had a handful of air temperature sensors in the tunnel, making it difficult to measure the effectiveness of the tunnel’s air handling system on a local level at the location of the girders and other components. This large-scale installation of additional temperature monitors into the Diamond tunnel can detect small, localised variations in air temperature of just 1/100th of a degree centigrade (see Figure. 8).

Figure 8. The change in temperatures during Run 1 2024 starting from a “cold” machine. The regularly spaced vertical bars correspond to Machine Development days.

Better monitoring of local temperature stability in the storage ring will help to inform future FEA and modelling work done by Engineering, who are working to evaluate thermal stability of components like electron beam position monitors and girders. This data helps to estimate the likely stability of components for Diamond-II, which will ultimately be essential for estimating the electron beam and X-ray beam stability. Better monitoring also enables the potential testing of alternative tunnel cooling approaches that could help provide more stable temperatures.

The system is designed to be flexible and has capacity to increase the number of sensors to investigate areas of interest in more detail. In addition, it is also designed to be compatible with Diamond-II so that it can be re-used after the accelerator upgrade.