April 25, 2001. Both LEDs installed. Timing Test

Did runs (see page 125 of my notebook) adjusting timing of logic signals out to LEDs. Plot "c" has the "yellow" (colour code of cables, not the light emitted) LED delayed 8 ns. Plot b has the yellow LED delayed 4 ns. Plot a has no delay and plot d has the 'red" LED delayed 4 ns.

ps - eps

April 16/17, 2001 Both LEDs at STRB

Did runs with LED A and B. Looking triggering LED B to affect output of LED B. Did a run with a scope probe across LED A. Ran an infinite loop of LED A only and averaged 512 signals. Then Did a run with LED A and B and looked at 512 signals averaged. Here are the two scope traces.

ps - eps

Prototype light source: first pass December 13. Work done December 7 on page 88 of my notebook.

Did a series of runs with the prototype light source using a 2 mm Nichea blue LED. First to calibrate the PMT, a spe run was performed. The of the run was guaranteed by doing a "gated-charge" run and ensuring that most counts were pedestal. Note also that the pedestal was a single channel wide and at ADC count 54. This 54 was subtracted off all plots

ps - eps

In order to get a reasonable solid angle a variable shutter was attached to the end of the PMT. The shutter was closed to its minimum size, 1/16 of an inch. A diffusing plate was put in front of the light source. The distance from the difusing plate to the shutter was 11.75 inches. The solid angle subtended by this hole is approximately 430 times smaller than the solid angle of an 8-inch PMT at 72 inches distance. An attenuating filter was needed to reach spe levels. When the filter was removed the follwing plots were obtained. Note that the energy scale has been approximately calibrated using the spe plot above.

ps - eps

This is unsatisfactory (read ugly) so I had a closer look at the spe spectrum to get a decent spe calibration. This is a very nice PMT. If one does a Guassian fit to part of the spe spectrum, the following is obtained:

ps - eps

The plot of the number of pe shifts slightly using 10.2 as the calibration.

ps - eps

The plot clearly looks better but there is something going on here beyond statistics because the chi-square is still way high.

Let's take six pe as the number of pe. To get 20000 pe from the light source we need a source 3300 times brighter. But we have used only 1/430 th of the solid angle so we are only short a factor of 7.

There are two big buts