When I fit to the horizontal (tangential) DCOPS offsets using fitmatrix12.gp (note that the ME4-fixed Line 2 fit fails horribly because the ME-4 DCOPS didn't read out) the offsets tend to track pretty well. There's some mm scatter in positions (no surprise), but the overall patterns track each other fairly well, as though we were in fact looking at the effect of bulk rotations of the disks.
If I subtract off the average for each line (station<9) from each line's points I get
When I fit fixing the MAB3 positions the result is a linear transform of the end-point fixed fit positions, as expected. However, since one of the endpoints in Line 2 is missing, I have more fits working when I fix the MAB3 points than when I fix the ME4 points. In the following sets of offsets from fits, the first number is the line, the second the orientation (1=U/D=Horizontal=tangent, 2=L/R=Vertical=radial), the third the station number (1=ME+4, 12=ME-4), and the last number is the amount the DCOPS is displaced from the ideal position defined by a line through the two fixed DCOPS centers.
To compare with Link fits, look at these from the endfit.offsetlist
1 2 5 7.344545242730383642276078315 1 2 8 7.631446159885393184265493370 delta = .287 3 2 5 -7.083908653635520991129368413 3 2 8 -2.891952244522233901545265068 delta = 4.192 4 2 5 5.037019324244623962443188337 4 2 8 6.812762042076308562922673221 delta = 1.776 5 2 5 1.494906611654473942961233109 5 2 8 0.8389064064098574037913120765 delta = -.656 6 2 5 6.218742999887161939625860379 6 2 8 6.956086544064652122221780185 delta = .737The predicted MAB positions from 29-April are shown below, with the caveat that for Lines 1, 2, and 3 I don't have both measurements for MAB+3 and MAB-3, and one of them uses an ideal number.
7258.762051974592748153341631 -1 1 7242.479163586663911488892489 -1 2 7243.916948476407243907359227 -1 3 7252.538537139551466410003420 -1 4 7249.635319990481633774894694 -1 5 7258.539165215481947938027608 -1 6 7257.141696176121794247548393 1 1 delta = -1.620 (*) 7241.912812239789410971670551 1 2 delta = 0.566 (*) 7244.095779932023615151569026 1 3 delta = -0.179 (*) 7250.703931039655337716353890 1 4 delta = 1.835 7248.360641765007044842730220 1 5 delta = 1.275 7251.556342388121109129359786 1 6 delta = 6.983
| Line | Laser Fit Minus-Plus | Link Fit Minus-Plus | Disagreement |
| 4 | 1.776mm | 1.835mm | .059 |
| 5 | -.656 | 1.275 | 1.931 |
| 6 | 0.737 | 6.983 | 6.246 |
Unfortunately the agreement is nothing to write home about. The image from 29-Apr showing the Link positions is a little worrisome for Line 6. Agreement between Plus and Minus is much better for the other two measured lines, and agreement between measured and ideal is also good for the other three. So why such a large jump for Line 6?
The MAB calibration might be at fault. So I looked at the difference between the distances in the two calibrations for MAB+3/11 and MAB-3/11. Everything looks pretty close, except that the calibration distances are about 80 microns larger. It is, of course, impossible to directly compare the Plus (first 3) and Minus (second 3) distances, because the reference position is arbitrary.
| Point | Calibration Bench | Tilt Bench |
| 1918401 | 2698.14 | 2698.09 |
| 1918402 | 2800.96 | 2800.85 |
| 1918403 | 2707.50 | 2707.43 |
| 1948401 | 2717.46 | 2717.37 |
| 1948402 | 2811.26 | 2811.23 |
| 1948403 | 2708.46 | 2708.32 |
| Distances from Reference | X | Y | Z |
| Calib Plus Point 1 | -2660.81 | -22.66 | -446.68 |
| Calib Plus Point 2 | -2761.08 | -23.36 | -470.41 |
| Calib Plus Point 3 | -2661.46 | -23.85 | -496.60 |
| Tilt Plus Point 1 | -1554.51 | -2205.26 | -0.080 |
| Tilt Plus Point 2 | -1608.07 | -2293.22 | 0.35 |
| Tilt Plus Point 3 | -1519.41 | -2240.89 | -0.23 |
| Calib Minus Point 1 | -2673.69 | 15.74 | -485.51 |
| Calib Minus Point 2 | -2773.41 | 15.60 | -459.52 |
| Calib Minus Point 3 | -2673.16 | 15.54 | -435.59 |
| Tilt Minus Point 1 | -1547.46 | -2233.71 | 0. |
| Tilt Minus Point 2 | -1636.40 | -2285.87 | 0.66 |
| Tilt Minus Point 3 | -1582.42 | -2197.94 | 0.16 |
Modified 01-June-2011 at 23:09
http://hep.physics.wisc.edu/~jnb/cms_dpg/01Jun2011
| Previous notes | Next notes | Main slide directory |
