“Another feature that makes the LightHouse machine unique is that it has to operate continuously, keeping the electron beam at more than 40 mA, with minimal downtime.
To achieve this, we had to design a robust system consisting of well-proven and dependable components, subsystems and technologies – and a really simple operator screen which enabled robustness that was envisioned from the early stages of the design phase.”
Matej Klun, Key Account Manager at Cosylab
Cosylab’s Solution for LightHouse
At Cosylab, we had to take care of the design of the control system, technologies and components of the LightHouse machine. The system is “high-energy”, capable of damaging key machine components in a few microseconds, thus demanding short reaction times of its machine protection system (MPS) on the order of 350 nanoseconds. We had to monitor roughly 700 inputs which we translated into various control signals for key accelerator components, and react to the inputs.
We also had to be thoughtful in selecting the system’s components. As much as possible needed to be industry standard and off the shelf because the LightHouse system must be “hot swappable” as it has to run without stopping, even if a component fails.
To increase reliability and upgradeability, we designed an FPGA solution based on the National Instruments platform. We also envisioned high availability storage and other tertiary subsystems, aiming to achieve high uptime and trustworthiness of all the device’s elements.
Protecting the accelerator components from damage caused by the megawatt beam is paramount. Still, shutdowns caused by false positive readings would compromise availability. Our machine protection system (MPS) employs a two-out-of-three voting scheme. Signals from the LINAC can go to three separate MPS units, and at least two have to agree on the actual state of the accelerator subsystem. This provides optimum protection of the accelerator hardware and increases availability as we have eliminated a single source of failure in the machine.