Pad geometry

The full pad detector photo is here, and another one.
The areas marked on the (smaller) photo below show where I took micrographs. These are in the table below.

In the table below, images correspond to rectangles outlined in the picture above. Click on them to get the full size.

  1   2   9   10   11   12  
21
20 		.  .  .  .  .  21
20
21
20
19 		  19
18
17
      17
16
15
14
      14
13
12
11
      10
9
8
7
    dumy spacer text 7
6
5
4
3
  3
2
1
  1   2   3   4   5   6 7   8   9   10   11   12


In addition, I found these pictures in a QM97 poster by GuanHua Xu:

with an explanation on this page





In this presentation from 1996 is a scan of the silicon strips, showing that there is virtually no dead area between strips. Shown are the readouts from two adjacent 200-micron wide strips (blue and red), and the sum of the signals.

Note that the signals fall off at <70 and >230, where the IR beam is intercepted by the Aluminum readout trace.

The area of the dip represents a 1% loss.



Also from GuangHua's poster is a postscript file of the pad detector. Here is a reduced gif:

with an explanation on this page. Here is the original postscript file

The ps file is 209095 lines long, and I edited it down to 1930 lines, so that only 21 pads along the center line are drawn. The boundaries of the pads in phi are arcs, and along r are straight lines. The corners are rounded. Thus each pad is made from 42-136 line segments. Here is a link to this single-row file.
Note the thick red line marking the detector boundary. Comparison with the microscope pictures shows that the silicon cut line corresponds to the inside of this line.





I edited a version of the pad detector ps file that prints out to about 75x85 cm (ftp to p25plotter2, then put big.ps). Measuring pads along the center line gives: 

     --------drawing--------   -----software----
pad   r1     r2   phi1  phi2      r3    r4  dphir
-------------------------------------------------
 1   26.2   46.2  20.4  21.0    5.25   5.45  2.4
 2   46.2   66.4  21.4  21.9    5.45   5.65  2.4
 3   66.4   86.4  22.3  22.9    5.65   5.85  2.4
 4   86.4  111.5  23.2  24.0    5.85   6.10  2.4
 5  111.5  136.5  24.2  25.0    6.10   6.35  2.4
 6  136.5  161.5  25.3  26.2    6.35   6.60  2.4 
 7  161.5  191.8  26.6  27.5    6.60   6.90  2.4
 8  191.8  221.9  27.9  28.7    6.90   7.20  2.4
 9  221.9  252.0  29.0  30.0    7.20   7.50  2.4
10  252.0  282.1  30.4  31.4    7.50   7.80  2.4
11  282.1  312.2  31.8  32.8    7.80   8.10  2.4
12  312.2  342.7  33.1  34.1    8.10   8.40  2.4
13  342.7  377.9  34.5  35.6    8.40   8.75  2.4
14  377.9  413.0  36.0  37.0    8.75   9.10  2.4
15  413.0  447.9  37.6  38.6    9.10   9.45  2.4
16  447.9  483.2  39.0  40.3    9.45   9.80  2.4
17  483.2  523.2  40.7  42.1    9.80  10.20  2.4
18  523.2  563.5  42.4  43.8   10.20  10.60  2.4
19  563.5  603.6  44.1  45.6   10.60  11.00  2.4
20  603.6  649.0  46.0  47.6   11.00  11.45  2.4
21  649.0  694.1  47.9  49.6   11.45  11.90  2.4
------------------------------------------------
Column 1 is the pad number from the inside out. Col 2 is the bottom of the (red) pad relative to the top of the red boundary line, in mm. Column 3 is copied from the previous one, shifted by 1. The next column is the width of the pad at the inner radius, and the 5th column is the width at the outer radius. This is the 'active width', from the left side of one pad to the left side of the next pad.
The last 3 columns are the inner and outer radius, and the delta-phi for each pad as it currently exists in the code.

Note that since the measured numbers are from a x10 drawing, and inner cut line is at 50.0mm from the z-axis, the following relation should hold:

r1/100 +5.0 = r3
This is good to 20 micron in the middle and 300 microns on the outside.

Using the numbers in the table I plot the ratio of areas. Note that there is a slope to the ratio. This is due to the fact that the assumed dphi3=2.4° implies a smaller gap between pads on the inside than on the outside.
The error bars are now correct, and the total effect is now about 7.5%.

If I take the measured values for r1 and r2, and convert them to the coordinates used by the software (cm, active area), I can calculate an effective angle phi_eff such that the area comes out to the measured area. These new values should replace the old values in the code (last 3 columns of the previous table). 

 
pad    r1       r2     phi_eff
-------------------------------
 1    5.252    5.452   2.216038 
 2    5.452    5.654   2.233846 
 3    5.654    5.854   2.250410 
 4    5.854    6.105   2.261362 
 5    6.105    6.355   2.262403 
 6    6.355    6.605   2.276802 
 7    6.605    6.908   2.293868 
 8    6.908    7.209   2.297191 
 9    7.209    7.510   2.296660 
10    7.510    7.811   2.311130 
11    7.811    8.112   2.324506 
12    8.112    8.417   2.329408 
13    8.417    8.769   2.337040 
14    8.769    9.120   2.338082 
15    9.120    9.469   2.348670 
16    9.469    9.822   2.355274 
17    9.822   10.222   2.366840 
18   10.222   10.625   2.369118 
19   10.625   11.026   2.373764 
20   11.026   11.480   2.382871 
21   11.480   11.931   2.386204
-------------------------------
Last update 29 Oct 2003 - HvH