Commissioning Plan 2017
2017 Hall C Commissioning Plan (draft)
Magnet Field Setting
HMS
- ALWAYS set DIPOLE by NMR coming DOWN from 900 A - still true?
- Set QUADS by initially setting to a current of 200 A above the set currents (or the maximum current of 1100 A for Q1 if set current + 200 > 1100), and then coming DOWN.
SHMS
- Set HB magnet to 1500 A and then go up or down setting HB by NMR
- Set QUADS … - need Q1 mapping data in form we can decide on this
Set DIPOLE by NMR (or with Hall Probe at highest momentum settings, not relevant for this commissioning plan) …
Beam Energy and Equipment Assumption
Hall C Pre-Ops presumably will start at beam energy above 6 GeV (3-pass or 4-pass) as first order one needs to prove the Key Performance Parameters are met. This means also the noble gas Cherenkov is in place. As soon as data is collected for this Cherenkov to validate the detector being operational (i.e., seeing signals fulfilling the Key Performance Parameters), we need to remove this detector and replace by the alternate SHMS vacuum extension, to reduce multiple scattering before the focal plane. This would require removal of some roof blocks and tech assistance. Beyond the >6 GeV checkout for most of the detector and spectrometer optics studies, we will also need about one day of one-pass beam energy, assumed to be in the 2.0-2.2 GeV range, to measure the dispersion matrix elements of the SHMS with a Carbon elastic scan.
Beam Energy > 6 GeV
Check Electronics and Detector Functionality
Set the spectrometer magnets to the initial settings as calculated with the FIELD_12GEV (has a start been made already?) program. This can be done before any optics fine-tuning of the quadrupoles/optics as the initial detector checkout can occur with a defocused run (in fact, it is preferred). Start at a spectrometer angle of 15 degrees and a central momentum of -3.0 GeV/c (assuming >6 GeV beam energy, if this would be one-pass beam energy choose -1.0 GeV/c). Use a current of about 10 μA, the central (Carbon) target of the optics target, and the fast raster with a size of 1 by 1 mm2. Use the large (pion or HMS-100 for HMS) collimators. Check that all electronics signals are well timed. Determine correct thresholds. Verify that all detector channels are counting. At this time checkoff the Initial Detector Checkout Plan (“near-final” for HMS, initial for SHMS to check off Key Performance Parameters) to make sure all detectors are operational. Check carefully that all wire chamber channels are cabled up correctly. Verify all scalers are incrementing. Check for double pulsing and time the wire chamber signals. Determine the scintillator plane efficiency for detecting electrons with the correct discriminator threshold, and make sure that the scintillator signals for detecting hadrons are on scale in the scintillator ADCs.
Beam Checkout with Superharps and Beam Position Monitors
Since the beam line has drastically changed with respect to the 6-GeV beam operations we need to establish the nominal settings for the beam position at the last BPMs before the target. This can be done before or after the initial HMS checkout but should be done before the optics tuning of the SHMS.
Superharp Scan
Make sure there is no target in the beam line, or the central carbon target that can also handle unrastered beam. Take a superharp scan with fast raster off (2 μA current) with superharps IHA3H07A and IHA3H07B. Verify that the beam size is as expected – this was <math>\sigma_x \approx 80 \mu m</math>, <math>\sigma_y \approx 150 \mu m</math> under 6-GeV conditions, as long as it is roughly <math>200 \mu m</math> or better in both directions it should be good enough to proceed.
Last Beam Line Girder Commissioning
At a stable beam condition, using the fast raster with a size of 1 by 1 mm2 and the central carbon target, monitor the values of the three BPM's of the last beam line girder (IPM3H07A-C, or “A, B and C”) and take a short run (~10K). Verify that we have all three BPM's (in Epics readout) of the last beam line girder in the data stream and that the values are consistent with those from the MEDM/TCL screen. Next, use unrastered beam (fast raster off) and the central Carbon target of the optics targets, and 10 μA current. Record superharp scans (they will give us the absolute scales) and check those versus the beam positions given by the BPM's both from the MEDM/TCL screen and from Epics readout. Ask MCC to move the beam horizontally by ±1 and ±2 mm with a far upstream magnet, and record at each setting both superharps and the three BPM's, both from MEDM/TCL and from the data stream after short runs. Have the beam moved back to the nominal central position (horizontally) and ask MCC to move the beam vertically by ±1 and ±2 mm, recording at each setting again the two superharps and all three BPM's. Change the current to 20 μA and repeat the whole sequence. Change the current to 60 μA and repeat. Go back to 10 μA current and take a short run (~10K) with a fast raster with varying size, say 2 by 2, 3 by 3, and 4 by 4 mm2, recording superharp scans after each run (this to calibrate the fast raster size).
Tuning Optics
Reestablish Standard HMS Tune
Use the HMS sieve slits in combination with the Carbon optics target (i.e. the central Carbon target on the optics target assembly hanging off the cryo target system). Use a current of less than 10 μA, and a fast raster size of 1 by 1 mm2, or with 0.5 mm radius. The exact raster shape does not matter: the raster can even be off in this configuration. Measure a short Carbon spectrum, about 250K events. Produce an ntuple, and do the following: select electrons with shower counter and/or Cherenkov cuts, and make a spectrum of x vs y at the nominal focal plane. What you should see is a ``spider" with 5 legs. The non-straightness of the central leg indicates there is an offset in the Z or Y direction. If you don't see a ``spider" or something resembling it one of the polarities of the HMS magnets is set wrong (or the magnet is off). If you see a ``spider" next thing to figure out is what the X, Y, and Z offsets are of your present beam-target interaction point, but at least your tune is fine for detailed electronics/detector checkout.
Verification of Beam-Target Interaction Point
Use the previous run, or if you are in the iteration process run at less than 10 μA current, use the Carbon optics target. Use a fast raster size of 1 by 1 mm2, or with 0.5 mm radius. The exact raster shape does not matter: the raster can even be off in this configuration. Take a short run (100K). Analyze and make an ntuple. Make the following plots: x vs y at the focal plane, xp vs yp at the target (you may have to use some Particle Id. and reconstruction cuts here!), and y at the target. You want to check the following: is the central leg of the ``spider" in x vs y at the focal plane straight? is the reconstructed y position close to y = 0? Is the central sieve slit hole close to (yp,xp) = (0,0)? If the xp position of the central sieve hole is close to 0, you probably are a bit off in vertical beam position after all, fix this. If you are far off (larger than 2 mr), check all your results carefully, did something go wrong in the vertical beam assignment? If the central leg of the ``spider" is close to vertical, you are close to having mid-plane symmetry for the spectrometer. You can vary the horizontal beam position a bit to check this. Note that the present quad alignment is such that the quad system is about 1 mm to the right of the line through the nominal pivot and the spectrometer angle, so the y position at the target can be a little bit negative, and the central leg of the spider can be slightly tilted. If the yp position of the central sieve hole is within 1 or 2 mr of the nominal zero position you are probably fine. The big uncertainty will be whether the targets are actually located at the nominal pivot (z = 0) position. If the target survey says otherwise, you expect (i) the central leg of the ``spider" not to be straight, (ii) the y reconstruction not to be perfect, and (iii) an offset in yp for the central sieve hole. If the three pieces of information are pretty much consistent with the survey assume you are done (note: the HMS sieve is at a distance of 1.66 meter of the target). You can consider checking this by using a hole target, or by rotating the spectrometer to a larger angle, and verifying that indeed things are consistent.
Consistency Check of Beam-Target Interaction Point – Verification of Target-Ladder
If done with the beam line checkout to establish a nominal centered beam on target, and having verified the beam-target interaction point using the well-understood and aligned HMS, one can consider a short run installing a hole target to ensure consistency. This checks that the target ladder position is well known, and that nothing went wrong in either establishing nominal positions for the last BPMs before target or the beam-target interaction centering with HMS.
Verify Quadrupole Settings SHMS
Now that we know the beam-interaction point, we start looking at the SHMS. Use the Carbon optics target and beam raster pattern as before, and the collimator (no sieve slit needed yet). For SHMS, the point-to-point optics at the focal plane will look skewed as the horizontal bend magnet will ruin the regular symmetry of the spectrometer optics, so x and y will be coupled. Thus, for the SHMS you may have to compare simulated focal plane patterns with measurements to be able to judge if quadrupole fields differ from expected although you can do much with just looking at the focal plane pattern. First verify that we have obtained a point-to-point focus with the extrapolated quadrupole settings, by looking at plots of hsxfp vs. hsyfp. The SHMS should have a “tilted” hour glass pattern for a point target source, with the waist of the hourglass at (0,0). If not, vary the quadrupole settings in small steps (try 0.2, 0.5 or 1% steps. Q2 is most sensitive and 0.2% steps should be sufficient for Q2 – the gradient of motion of the hourglass waist is roughly 2 cm per % Q2 change. For Q1 it is about -0.4 cm per % Q1 change), until the golden tune (this is defined as the SHMS quadrupole settings that most closely reproduces the simulated sieve slit patterns at the focal plane) is obtained. Measure a spectrum with high statistics (>>100K) to use for later off-line checks and to continue the second phase of detector checkout – this is needed before starting the sieve slit runs. Check the time-to-distance maps, align the wire chamber positions in software and enable linked stub fitting, check the detector positions, check the timing and calibration constants (shower counter gains, pedestals, timing offsets, pulse height corrections, attenuation lengths, efficiencies, position dependencies). Optimize tracking properties. Make sure that <math>\Theta</math> and <math>\Phi</math> spectra are wide as expected. Construct <math>x</math>, <math>y</math>, <math>\Theta</math>, and <math>\Phi</math> spectra at the nominal focal plane. Does everything look reasonable? Check tracking with one wire chamber set against tracking with two wire chamber sets. Reconstruct target quantities.
Establish Standard SHMS Tune
Use the SHMS sieve slit with the middle vertical column centered in combination with the Carbon optics target. Use a current of less than 20 μA. Preferentially, the fast raster should be off for this series of measurements. However, a small fast raster size of 1 by 1 mm2 should also work. Measure a short Carbon spectrum of perhaps 100K events. Produce an ntuple, and do the following: make a spectrum of x vs y at the nominal focal plane. What you should see is a tilted ``spider" with 9 legs. The center of the spider/hourglass should be at (x,y) = 0 if everything went right. If not, check the beam interaction point and check the quadrupole magnet settings. Because the HB magnet destroys the mid-plane symmetry of the SHMS, you likely will need to compare simulated focal plane patterns with measurements to be able to judge if quadrupole fields are o.k. If you do see a difference you can try to change the quadrupole settings by say 0.1% and measure a new run and produce and ntuple to compare. If the patterns look similar we are done with establishing the Standard SHMS tune beyond the dispersion. We will first do more detailed detector checkout at 3- or 4-pass beam energy towards the Key Performance Parameters. Then, we first likely should go to one-pass beam energy to check the dispersion and then return to >6 GeV beam energy for further optics and systematic understanding checkout.
Detailed Detector Checkout
Note: still need to do editing in this detailed detector checkout section but the first-order fixes were made.
Wire Chamber High-Voltage Plateaus
If not already there, rotate the spectrometers to 15° and set the spectrometer momenta at 3.0 GeV/c (negative polarity, assuming >6 GeV beam energy) to increase the count rates. Plot the counts per wire chamber plane vs. the high voltage, determine the wire chamber high-voltage plateaus for electrons for both SHMS and HMS (as there are new wire chambers there – note that these do not require the checkout versus threshold of the previous HMS wire chambers). Choose correct operating voltages for detecting electrons.
Shower Counter Calibration Run
By the time this starts all changes in the electronics and trigger have been made and the ADC gates for the calorimeter signals have been adjusted. With HMS and SHMS still at angles of 15° and momenta of -3.0 GeV/c, take a run with >100K statistics each. Adjust the delay time for the ADC gates with +/-20 ns and measure again. If the gate timing is optimal, the ``maximum" ADC values should change less than 5%. Adjust the gate timing if not optimal and repeat. If the gate timing is optimal, start gain matching by adjusting the HV's. For optimal HV settings the ADC peaks must be in the range between 80 and 100. Adjust the HV of those PMT's which are out of this range (a ±50 V change results in a change of ~15-20% in amplitude). After this procedure NEVER change the HV settings for the calorimeter PMT's anymore. Now take large statistics runs (>250K each). If time permits, one can consider reversing the polarities of both spectrometers and take again large statistics runs (>250K each). Do not use the Particle Id. trigger in any of these runs!
Calibration Spectra: Carbon
Take a large run (250K) for both HMS and SHMS to check wire chamber time-to-distance maps, align the wire chamber positions in software and enable linked stub fitting, check the detector positions, check the timing and calibration constants (shower counter gains, pedestals, timing offsets, pulse height corrections, attenuation lengths, efficiencies, position dependencies). Optimize tracking properties. Make sure that <math>\Theta</math> and <math>\Phi</math> spectra are wide as expected. Construct <math>x</math>, <math>y</math>, <math>\Theta</math> and <math>\Phi</math> spectra at the nominal focal plane. Does everything look reasonable? Check tracking with one wire chamber against tracking with two wire chamber sets. Reconstruct target quantities. This run can be used for the initial timing/calorimeter/cherenkov calibrations.
Scintillator Calibration Run
The previous run can also be used for timing tests for a “low count rate” situation:
- Make sure that all ADC's, TDC's including the RF TDC are present. Check the timing properties at focal plane (beta resolutions) for this ``low count rate" situation.
- Determine the scintillator plane efficiency as a function of discriminator threshold setting in the electronics.
Then, try to mimic a “high count rate” situation:
- Hodoscope Timing Tests HMS and SHMS --- High Rate. Measure the above run with much higher statistics, say about a factor of five. This you can accomplish by for example using the triple optics target and an increased beam current. Make sure all ADC's, TDC's including the RF TDC are present. Measure at least 250K events for each spectrometer. Check the timing properties at focal plane (beta resolutions).
Defocused Run HMS and SHMS
Here, we want to illuminate the whole detector plane. Adjust the current settings of the HMS Q2 and SHMS Q2 (20% different from default should do the trick). Use the BeO target, and central momentum settings of -3.0 GeV/c (assuming both spectrometers are still at 15o). Make sure that all ADC's, TDC's including the RF TDC are present. Measure a spectrum with high statistics (250K) to use for later off-line checks. Check the time-to-distance maps, align the wire chamber positions in software and enable linked stub fitting, check the detector positions, check the timing and calibration constants (shower counter gains, pedestals, timing offsets, pulse height corrections, attenuation lengths, efficiencies, position dependencies).
End of detailed detector checkout that still needs updating. Detailed detector checkout should at the least have brought us to the Key Performance Parameters, to be shown at this >6 GeV energy. It is assumed we now move to the further SHMS optics tuning at lower (one-pass) beam energy, where we then after this stage return to >6 GeV beam energy and take detailed measurements to determine the optics matrix elements for extended targets and do the further detector checkout related to Particle Identification. The noble-gas Cherenkov should be replaced with the alternate SHMS vacuum extension.
SHMS Tuning - Beam Energy ~ 2 GeV
At this stage the operability of the noble-gas Cherenkov to fulfill the required Key Performance Parameters should have been shown, and the detector should have been removed and replaced with the alternate SHMS vacuum extension.
Quad Tuning
Inelastic Region
Here we will first check if the chosen quadrupole settings at >6 GeV beam energy will keep the focus of the spectrometer (as determined by the waist of the spider/hourglass) fixed for a different momentum/quadrupole field values. Set the spectrometer angle to 15 degrees (or if it was already at seven degrees in preparation of the elastic scan below this is fine) and set the momentum to -1.2 GeV/c. Make sure that the beam remains centered, simply looking at the IPM3H07A-C beam position monitors may suffice already, otherwise use the superharp/BPMs or view at a BeO target, or use HMS. Use the thin C target as before, and the sieve slit, and up to <math>20 \mu A</math> with an unrastered or small rastered (1 by 1 mm2) beam. Measure a short Carbon spectrum of perhaps 100K events. Produce an ntuple, and do the following: make a spectrum of x vs y at the nominal focal plane. What you should see is a tilted ``spider" with 9 legs. The center of the spider/hourglass should still be at (x,y) = 0 if everything went right. If required you can estimate a needed change in Q1 or Q2 from the gradients as given before (2 cm shift for a 1% Q2 change and -0.4 cm shift per % Q1 change). Minimize the variation in the fields until we believe the quadrupole settings are determined to as close to 0.1% as you can get.
C Elastic
Change to an elastic central momentum setting of the SHMS, this should be at around seven degrees for a 2.2 GeV beam energy. At this angle the ratio of ground state to 2+ at 4.4 MeV for carbon-elastic scattering should be about 0.5:1. If SHMS can only be used at larger angles this drastically affects this ratio, like at nine degrees it is more like 0.02:1. Use the thin C target. Verify occasionally that the beam is centered, either with the superharps/BPMS or as viewed on the BeO target. Check that the ground state for carbon ratio of ground state to 2+ is indeed about 0.5:1 close to xfp = 0. Note that if we did everything right the focal plane should be at z = 0 cm. Project to this position and check the two-dimensional plots of xpfp vs xfp, and delta vs xfp to start looking at the energy resolution for the elastic peak (use the SHMS ntuple). Now we start tuning the quadrupole fields to optimize the energy resolution for the elastic peak. It should help that we have comparable strengths of the ground state and first excited 2+ state of Carbon. You may also compare with the simulated focal plane spectra for a delta function in momentum, for variations of Q1 and Q2. Minimize the variation in the quadrupole fields until their settings are roughly determined to better than 0.1% (this likely will not be possible for Q3 which has less effect). If we did the earlier checks of establishing the standard tune at >6 GeV beam energy by optimizing the spider/hourglass right we should already be within a few 0.1% of the right values, but we may need to resolve ambiguities between Q1 and Q2 here. After this, we have established the early SHMS magnet settings and hope it does not change with momentum – if so we are in business!
Dispersion Calibration: C
Install the sieve slit. Shift the elastic peak of Carbon across the focal plane by making 4% changes in the momentum setting (from +25 to -15% going down). Determine the dispersion matrix elements. Verify occasionally that the beam is centered.
Get X and Y Magnifications: Beam Sweep
Use the central momentum setting, the sieve slit, and the thin Carbon target. Simulate a rastered beam by moving the beam up and down, left and right by a few mm (as much as we can have, but probably 2-3 mm suffices). Determine the influence on aberrative matrix elements.
Sieve Slit Measurements with C Optics “Extended” Target
The main goal of this step is just to have a set of sieve slit measurements in the can for the low momentum, as the bulk of the optics matrix element data will come later, this is only for checking if we see any momentum-dependence of the optics. If everything is correct we should not for the SHMS.
Lower the central momentum setting of SHMS to -1.2 GeV/c again. Put the sieve slit in. First verify that the beam is centered on the BeO target. Now we need to mimic an extended target. Presumably, this means measuring two separate runs, for
- The triple optics target, with carbon foils at -10, 0 and +10 cm
- The double optics target, with carbon foils at -5 and +5 cm
Take runs with at least 100 counts for each visible hole. Determine the angular matrix elements. As described, we simulate a 10 cm target length (as viewed by SHMS at a 30 degree scattering angle). Scan over +/- 1 cm (or whatever we used in an earlier step) to verify the effect of beam rastering on the angular matrix elements.
Further SHMS Tuning and Particle Identification Checkout - Beam Energy > 6 GeV
Note: here we assumed a 6.4 GeV beam energy, different energies are fine but we then need to recalculate the kinematics that are picked with heepcheck.
Coincidence Electronics and Particle Identification Checkout
Superharp Scan [1 hr]
Make sure there is no target in the beam line, or the central carbon target that can also handle unrastered beam. Take a superharp scan with fast raster off (2 μA current) with superharps IHA3H07A and IHA3H07B. Verify that the beam size is as expected – this was <math>\sigma_x \approx 80 \mu m</math>, <math>\sigma_y \approx 150 \mu m</math> under 6-GeV conditions, as long as it is roughly <math>200 \mu m</math> or better in both directions it should be good enough to proceed.
SHMS Optics Verification [1 hr]
Start with the SHMS at 15 degrees and at a central momentum of -3.0 GeV/c. Use the central (Carbon) target of the optics target, and the fast raster with a size of 1 by 1 mm2. Use the SHMS sieve slit with the middle vertical column centered in combination with the Carbon optics target. Use a current of less than 20 μA. Preferentially, the fast raster should be off for this series of measurements. However, a small fast raster size of 1 by 1 mm2 should also work. Measure a short Carbon spectrum of perhaps 250K events. Produce an ntuple, and do the following: make a spectrum of x vs y at the nominal focal plane. What you should see is a tilted ``spider" with 9 legs. The center of the spider/hourglass should be at (x,y) = 0 if everything went right. If not, check the beam interaction point and check the quadrupole magnet settings. Because the HB magnet destroys the mid-plane symmetry of the SHMS, you likely will need to compare simulated focal plane patterns with measurements to be able to judge if quadrupole fields are o.k. Hopefully, everything is o.k. and we have established that the SHMS tune is to first order momentum independent.
Set Discriminator Thresholds for (P.Id.) Trigger
Part of these runs can be intertwined with other checkout. We need to have many small runs where we change the discriminator thresholds for the various trigger signals. Note that we obviously have to come back to some of these thresholds at a later stage, as the shower counter related discriminator thresholds will be momentum dependent, so some of this may come back after the following checkout. You can also check the scintillator plane efficiency as a function of discriminator threshold setting in the electronics.
Timing [3 hr]
Rotate HMS and SHMS both to an angle of 27.5°. Change the HMS momentum to -3.609 GeV/c, change the polarity for SHMS and change the momentum to +3.609 GeV/c. This setup should give you 0.2 Hz for the 3% radiation length Carbon target and a 20 μA beam current. You may want to ask for more beam current (50 μA?) and a change in fast raster to 1 by 1 mm2. You can also use the 10 cm LH2 target, which should give you a little above 1 Hz (for 20 μA beam current). Use the large (pion or HMS-100 for HMS) collimators. Check that all coincidence electronics signals are well timed, i.e. check coincidence trigger signals and ADC/TDC timing.
1H(e,e’p) Coincidences [0.5 hr]
If not done so yet, install the 10 cm LH2 target and ask for 50 μA and a fast raster of 1 by 1 mm2. The rate from the 1H(e,e’p) reaction should be about 4 Hz. If you have verified that we see coincidences, measure at least 5K events.
1H or 12C(e, e’ π) [2 hr]
Maintain/set the SHMS polarity to detect positively charged particles. Set the angle of SHMS at 20° and the central momentum setting of SHMS to +2.7 GeV/c, and the angle and central momentum setting of HMS to 16.8° and -3.8 GeV/c (please check). Use either the 10 cm LH2 target or the thick Carbon target (3% radiation length). Make sure we write HMS singles, SHMS singles, AND Coincidences. Measure at least for 20 minutes in this situation. Do we see any evidence for (e, e’ π) coincidences? (we expect about 10K per hour).
1H(e’K) and 1H(e, e’K) [3 hr]
Set the angle of the SHMS at 15°, and keep the central momentum setting of SHMS at +3.6 GeV/c to detect kaons, and the angle and central momentum setting of HMS to 25° and -2.6 GeV/c (please check, maybe better to pick lower-Q2 kinematics, but still not too far from where we are with SHMS and HMS settings, to optimize rates and lower kaon momentum). Use the 10 cm LH2 target and 50 μA beam current. For this run the SHMS aerogel detector should be installed. Make sure we write HMS singles, SHMS singles, AND Coincidences. Measure at least for one hour in this situation. Check whether we can see any evidence for Kaons from Timing w.r.t. RF and the aerogel detector. Do we see any evidence for (e, e’K) coincidences? (we expect about 500 per hour).
