Joint PSP/Taskforce minutes for 02 Jan 2020 ============================ http://beamdocs.fnal.gov/AD-public/DocDB/ShowDocument?docid=7905 (a) Updates from Tan for both PS and Task force (i) Status of PSP projects (1) Penning source being modified. Revisiting arc modulator design. (2) Klystron testing. Two spares have been tested. (3) Flat injection porch. A quick study was done on 11 Dec. Data shows that beam moves as what is expected from B-field changes. (4) Wide bore cavity. To be tested soon. Plan is to install in Booster before shutdown. The number wide bore cavities needed for PIPII has been calculated. Results from these calculations say that at least 16 are required. (5) Collimators. Meeting after V. Kapin comes back from vacation. mechanical engineer to be assigned. (6) Garnet LDRD. Mechanical engineer has been assigned to work on the fixture. (7) GMPS LDRD. Brian will be going to Korea for machine learning workshop. Waiting for Brian Chase's board. (8) Digital paraphase. Studies on 11 Dec are promising. Beam was able to get through transition. Next study will include adjustments for delays and bunch rotation. (ii) The holidays have been very difficult for operations. (1) On 24 Dec, one cap in the KRF6 PFN network shorted to ground. Took 16 hours to find. (2) Power glitch on 01 Jan. Took out BRF22 modulator and Linac QPS303. (iii) Task force (1) Method to reduce 1/2 integeter tested. See today's talk. (2) Magnet girder tests. E4R has been cleaned up for the setting up the girder. (3) 20 Hz infrastructure. Plan is to have meeting soon. (4) Tall aperture gradient extraction magnets. Increase of gap size found to be 5mm required. However, calculation needs to be benchmarked with present operations. (5) 20 Hz cavity test. Report is being written. (6) 50 kV in situ cavity test which is required for PIPII. Plan is to choose 2 cavities and run them at 50 kV until shutdown. (7) BTL to L11 changes. (a) In BTL, a (de)buncher to rotate the bunch to get smaller dp/p. Space charge increases dp/p by 2x. (b) Collimators in BTL are required to clean up tails of beam before injection into Booster. Will they be 2 stage collimators? Instrumentation also required. (b) Kiyomi told us how LRF4 sparking was traced and how it was repaired. (1) LRF4 tripped off on 11/7 from reverse power. It wasn't until 11/15 that the problem was correctly identified and repaired. (2) First suspect was gas barrier. BUT (i) there was no sparking noise (ii) measurement of forward and reverse power did not clearly indicate that the gas barrier was the problem. (iii) Retuning the trumbone traditionally should have helped but that didn't work. (3) We decided to change out the gas barrier first rather than look into the tank. (i) Gas barrier replacement had to done with safety because of asbestoes in the wall that contained the shielding blocks. HAs, photographs documentating blocks were done for the job. RSO coverage as well. (ii) Gas barrier was replaced and ran for 4 hours. But sparking returned. (4) Gradient pickups on the tank were used to isolate where the sparks occurred. (i) Scope data looks like sparking at the low energy end of the tank. (ii) If it was the gas barrier, the collapse of the RF should be in the medium energy part of the tank. (5) Tank was opened up (i) scorch marks were found on the end cap. (ii) The entire tank was examined but there were no other obvious scorch marks. (6) Careful examination of the end cap showed damage of the spacer from sparks. This was probably due to the missing bolt that held the gasket near the bolt. (i) The end cap was cleaned. (ii) Spacer was smoothed out. (iii) A set screw was used to replace the bolt so that it can push on the gasket. (iv) Finger gauges were used to make sure that the gap was tightly bolted down. (6) The repair was successful and beam returned on 11/15 at 14:00 hours. (c) Jeff summarized the 1/2 integer correction study done on 11 Dec. (1) The goal is to move the beam close to the 1/2 integer and see whether the correction will reduce the size of the 1/2 integer. (2) MADX was used to see the effect of moving the beam close to the vertical 1/2 integer. (i) The MADX model clearly shows strong effects on the vertical beta functions. (3) Compensation is achieved for carefully selecting the quads (i) adjacent quads, e.g. QL24 and QL01 are approximately pi apart and so if they are oppositely powered, they contribute in the same direction. (ii) 12 Booster cells again gives approximately pi phase advance. (iii) So the first family (Group A) using 6 quads are -QL24, +QL01, -QL02: adjacent. After 12 cells: +QL12, -QL13, +QL14 (iv) Quads that are 6 cells apart give pi/2 phase advance. Therefore, the orthogonal family (Group B) is -QL06, +QL07, -QL08: adjacent After 12 cells QL18, -QL19, QL20 (4) Using these quad families a tune scan was performed. (i) 75% efficiency was found with Group A = 2.222A and B=-1.238A (ii) Results are close to what MADX had predicted. (iii) Tune was also minimized at the above setting. Again confirming that this is where the best correction is. (iv) Tune scan also showed reduction in size of the vertical 1/2 integer (v) However, the correction did not improve the amount of beam lost when the beam is away from the resonance. (5) Orbits were measured and horizontal and vertical orbits do show excursions at locations with aperture restrictions. (i) shifts were 2-5 mm horizontally and 1-2 mm vertically at a few locations. (6) Tan said that the corrections applied here is probably not the right ones for normal operations. (7) Yuri said that to measure the orbit errors, it is better to be close to nominal tunes because the effect is the largest there. (8) Next steps are (i) to clean up the orbit after compensation. (ii) measure emittances