Due to the physics of synchrotron accelerators, beam at NSRL is delivered in a “spill” structure whereby a few hundred milliseconds of beam-on is followed by a few seconds of beam-off.
This is because each group of particles needs to be injected into the synchrotron, captured, accelerated to their final energy, then extracted down the NSRL beam line. After extraction, the synchrotron magnets must ramp down to their baselines in preparation for the injection of the next group of particles.
Because NSRL shares the Booster synchrotron with other facilities in the BNL Collider-Accelerator complex, the beam-off time is longer than just what would be needed to ramp down then accelerate the next group of particles. Instead, this “cycle” must be repeated several times to accelerate particles for the other programs that are concurrently making use of the Booster. After every cycle has completed, the “supercycle” repeats and each cycle occurs again in the same order.
The diagram below shows the Booster main dipole magnet strength as a function of time for a particular supercycle. 6 short cycles for AGS/RHIC experiments are followed by a single, long cycle for NSRL.
The duration of the supercycle depends on which programs are making use of the Booster synchrotron and how many cycles are being run but typically ranges from 4.2 seconds to 6.6 seconds. NSRL can only make use of a single cycle, so the supercycle length is the same as the NSRL spill period. That is to say, if the supercycle is 6.6 seconds long, a spill of beam can be delivered to NSRL every 6.6 seconds.
The length of the spill itself varies by ion energy and species, but typically ranges from 300-800 milliseconds.
Ideally, beam would be delivered down the NSRL beamline at a perfectly constant rate during the beam-on time. Schematically, this would look like the figure below.
In reality, the spill is not uniformly distributed in time, but has a time structure on many different levels. The figure below shows the time profile of a spill of beam as measured by an ion chamber in the NSRL beamline.
Spikes, steps, and ripple in the beam intensity are due primarily to 60 Hz noise in the power supplies used to drive the accelerator magnets.
The figure below shows the same spill, but zoomed in to better show the effects of the 60 Hz noise.
On a shorter time scale, there is a microstructure in the beam due to the radio frequency cavities used to accelerate the beam in the Booster. These cavities operate at ~2.2 MHz, causing the beam to be bunched in packets of a few nanoseconds long, separated from each other by approximately 400 nanoseconds. The response time of the ion chamber in the beam is not fast enough to allow it to observe this microstructure.
There are RF processes that occur during the extraction sequence which work to debunch particles before they are diverted down NSRL’s beamline and thereby reduce this microstructure. These processes have varying efficacies depending on the ion species and energy of a beam.
Additional Considerations
Under some conditions users might need to make use of beams with other time structures. We have experience modulating the 400 millisecond spill with either a 20 Hz or 60 Hz pulse structure. Other time structures can be developed as needed. Users interested in using time structured beam should contact the NSRL PI.
NSRL also has the capability to extract the entire beam during a single turn of the Booster, also known as “Fast Extracted Beam (FEB)”. This results in a burst of beam delivered over about 800 nanoseconds.
If spill modulation, fast extraction, or any other nonstandard time structure is required for a test, please contact NSRL staff as soon as possible so as to allow ample time for preparation and testing.
This page was last modified:
