Here are some technical details about the raster in Hall A.
The raster is a set of dipoles which deflect the beam at 25 kHz to spread the heat on the target and on the beam dump. Typical deflections are 4 x 4 mm (full width). There are 2 X and 2 Y coils. The X and X are synchronized, as are the Y pair. The ratio of frequencies fx/fy should (in theory) be an irrational number in order to get a Lissajous (TV) pattern. At lower beam energies (< 6 GeV) we can run with 1 X and 1 Y, but or higher energies we double the number of coils to get double the kick.
Upstairs in the counting room there is a function generator, a scope, a manual control, and a relay for turning off the raster. I suggest you don't play with these.
The function generator is an Agilent model 33522A unit near the bottom of rack CH01B05. It drives the entire raster system. There are two outputs, one for X and one for Y. The setting of the function generator is 25.08 and 24.96 KHz with a 5Vptop square wave coming out.
The scope looks at the LEM output, which is a pickoff of the current in the raster. The scope is also in rack CH01B05.
The relay unit is in the NIM crate with all the helicity signals, just below the HAPPEX crate in rach CH01B03. Note, this can turn off the raster ! If either the cable from it is unplugged or if the NIM crate power is turned off, the raster will go off, and then one relies on the FSD to shutdown the beam. We could instead plug the raster straight into the wall outlet, though it would also FSD the beam if that loses power. Why the relay ? The magnets have small currents when the raster amplifiers are on so there was a request, a long time ago, for this relay to be able to turn the current all the way off.
There is also a manual control at the top of rack CH01B05. It should not be used, however, under normal circumstances because the raster is driven by MCC software control.
Need help ? The main people responsible for the raster are : Bill Gunning (hardware) and Michael Johnson (software). Bob Michaels or Chris Cuevas could be called in if the others are not available.
Historically, the main thing that goes wrong is that a power supply fails. There are some spare power supplies. Last time I checked they were in Bill Gunning's work area in the EEL building. Other possible locations are the transportainer inside the hall or in the 2nd floor of the counting house in or near Jack's work area. Bill Gunning is the main person who can service those power supplies, but if he cannot be reached then Bob Michaels should at least be able to swap out a power supply (I've done it a few times before). Other things that have gone wrong are typically related to people "playing" (or having accidents) with the raster-related components in the counting room.
- In Aug 2018, Michael Berkowitz (grad student at Columbia U.) working with Bill Gunning revived the circular raster. His update talk: https://userweb.jlab.org/~rom/raster/Berkowitz_12GeV_Circular_Raster_Aug18.pdf
- Bob's Aug 2016 talk about the raster http://userweb.jlab.org/~rom/RasterUpdate_9Aug2016.pdf
- What caused the wiggles in the raster in recent (2016) running ? The story: https://logbooks.jlab.org/entry/3319834 and also https://logbooks.jlab.org/entry/3318770 ... But it looks like it was solved for Tritium (Sept, 2017) https://logbooks.jlab.org/entry/3484369
- Some other info old and new about the raster http://userweb.jlab.org/~rom/raster
- Running spot++ to measure rastered spot size http://hallaweb.jlab.org/equipment/daq/spot.html
- Raster chapter in OSP https://userweb.jlab.org/~gen/osp/osp_html/node61.html
- Bull's eye scan http://hallaweb.jlab.org/experiment/PVDIS/documents/runplan/How_to_run_Bullseye_scan.html
- BPM calibration https://hallaweb.jlab.org/wiki/index.php/Raster/BPM_calibration
- Good example of detailed beamline checkout procedure https://hallaweb.jlab.org/wiki/index.php/Beamline
- Chen Yan's raster doc https://userweb.jlab.org/~rom/raster/beamline_raster.pdf and his expert howto for maintenance https://userweb.jlab.org/~rom/raster/beamline_fr_manual.pdf