From a person's point of view, the beam of electrons driving down the linear accelerator at close to the speed of light would appear as a single thread. "But from the point of view of precision electronics," explains Jachin Spencer, "you might see something, you might see nothing."
To a non-specialist, the word "beam" may be associated with columns of material or light. But on a microsecond time scale, the beam in an accelerator is actually a series of electron-bunches, more like machine gun fire than a shaft of sunshine.
Spencer, a summer student in the SULI program at the Stanford Linear Accelerator Center, is studying the performance of what is called the "fiducial pulse," a pulse of electromagnetic waves used to coordinate the timing of the accelerator. It's the job of the fiducial pulse, Spencer says, to let electronics know "when they can look at the beam, when there's something to see."
Both clock and courier, the fiducial pulse travels just ahead of the electron bunch, and alerts the monitoring devices that line the beam tunnel. Beam position and length monitors, cameras rigged to phosphorescent screens, and other sorts of diagnostics tools use the early warning to shutter their sensors in rhythm with the electron bunches, 30 times a second.
It's no exaggeration to call the fiducial pulse the linchpin of the linac. Without the regularity of its preceding pulse, instruments monitoring the beam would fall out of synch with the bunches, and therefore out of synch with one another.
There are two kinds of triggering signals available and the question of which is preferable is the subject of debate. It's Spencer's project to determine which is best. "My job is to define what better is. They want whichever's better, but they don't know what that means."
Both fiducial signal types consist of very brief pulses added to a sinusoidal wave. Electronics in the linac are sensitive to the voltage difference between this normal wave and the pulse. The fiducial pulse was originally designed to be what's called a "bipolar waveform." However, ever since it was constructed 20 years ago, the linac has been operating with a close cousin, the unipolar waveform.
Because the decision behind the signal change was never officially recorded, the question of which is better for the linac has never been adequately broached, much less answered. The constant operation of the trigger system since it was implemented in 1984 has made it difficult and expensive to schedule actual experiments, thus warranting the use of simulations.
The choice of which waveform to use seems to be a question of power versus clarity. The unipolar wave contributes comparatively more power to the carrier wave by the time it reaches the end of the drive line, but creates a slightly distorted waveform. The bipolar wave preserves the waveform but becomes attenuated by the end of its course. So far, it's unclear whether the unipolar's added power actually helps the triggering system. Spencer says, "The strengths of one make up for the weaknesses of the other...it's almost perfectly balanced that way."
