How to do Target Boiling Test

• How to observe Lumi asymmetry RMS values online:
• using apar account on an adaq machine, type
• $ gopan
• online -f boil
• The panguin should pop up, and you can monitor on the BLumi asymN pages.
• Formal analysis should be done off-line using PAN. Panguin is just to make sure everything works okay at the 1st order.
• Run size: Run at each condition should have at least 15000 pairs, i.e., roughly 17-18 minutes of good beam.

• Short version, to be done during commissioning:
• Run size: Run at each condition should have at least 5000-10000 pairs, i.e., up to 12 minutes of good beam.
• The largest raster size MCC can give us is 6x6 in MCC units, which correspond to 3x2.75mm^2 on spot++.
• The highest current MCC can give us for Nov. 6 owl shift is 85uA.
• raster size (in MCC unit): do 2x2, 4x4, and 6x6 (the largest possible). Other conditions: beam currents 10uA, 20uA and 85uA (the highest current MCC can give us), fan speed 60Hz, for both 20-cm LD2 and carbon target;  This means a total of 2 targets x 3 currents x 3 raster size =18 runs, or about 5 hours.
• If there is more time, repeat for 50uA beam current: 2 target x 1 current x 3 raster size = 6 runs, or another 1.5 hours.
• Record Lumi width (all 8 lumis) for each run. Plot it vs. raster size for each running condition. And post the results in halog right away.
• Note: density fluctuation should decrease steadily at larger raster size.
• Full version: Variables to be studied/"scanned":
• raster size (MCC unit): from 2x2, 4x4, 6x6 (or the largest possible). Other conditions: beam currents 5uA, 10uA, and 85uA (if MCC cannot deliver 85uA, pick the highest current they can give us. If can reach higher current, above 95uA, do one point at 85uA and another at the highest, if between 85 and 95uA, do one point at the highest point), fan speed 60Hz, for both 25-cm LD2 and Al dummy (or carbon?) target;  This means a total of 2 targets x 3 currents x 6 raster size =36 runs. 
• Note: density fluctuation should decrease steadily at larger raster size.
• fan speed: at 40, 50, 60Hz. Other conditions: raster size 6x6, beam currents 5uA, 10uA, and 85uA, for both 25-cm and a solid target. This means a total of 2 targets x 3 currents x 3 fan speeds = 18 runs.  
• Note: density fluctuation should decrease steadily at larger fan speed.
• beam current: 5uA, 10uA, 20uA, 40uA, 60uA, 80uA, 100uA (or 85uA, or the highest the target cooling can take). Other conditions: raster size 6x6, fan speed 60Hz (nominal). For both 25-cm LD2 and a solid target. This means 2 targets x 7 currents = 14 runs. 
• Note: density fluctuation should go up for higher current.
• If all goes well can add the same scan for different raster size or even fan speed.
• for current scan should set Lumi HV to the 100uA value so it won't saturate for the whole procedure

• Online analysis goal: Right after the scan, make the following analysis:
• the 5uA and 10uA tests are used to establish statistical width. The solid target data serve as a base check (solid targets don't boil).
• For each scan, plot the Lumi RMS (all 8 Lumis, and later also the combined) vs. the scanned variable. 
• For beam current scans, use the 5uA and 10uA points to establish the purely statistical width, draw a 1/sqrt{I} fit using these two points. The fit should fall below the width of all higher currents.
• The observed RMS gives sigma_observed, and the 1/sqrt{I} fit gives the expected sigma_statistical.
• The boiling noise is calculated as  sigma_boil = sqrt(sigma_observed² - sigma_statistical²).
• criteria for establishing target condition for production running:
• the boiling noise from the target should be <10^-3 or 1000ppm.

• Other variables: beam spot position?