On Nov. 13,2002 , the last hour of the last day of of CLAS G7 running, an attempt was made to produce plateau curves for the tagger E-counters. This was done with some limited success and the results are shown here.
E1 -E16 | 1.484V |
---|---|
E17-E32 | 1.016V |
E33-E48 | 0.820V |
E49-E64 | 0.859V |
E65-E80 | 0.820V |
E81-E96 | 1.016V |
There was a significant software problem that led to the scalers being recorded with HV values that were not neccessarily at the the actual value. This causes some distortion of the X-axis.
For those who are interested, here are the details: The EPICS scaler from which the E-counter scalers are read is an array of 384 values, only in reverse order from the e-counter numbers. Hence, value 1 corresponds to E-counter 384, etc.... Since all channels are set to the same HV values, this would not have been a problem except for the following: The plateau program can fork itself into several independant programs, each handling a block of E-counters. This is done to significantly speed up the procedure. These processes run in parallel and do get out of sync over the course of the program. Typically, they are never out of sync by more than 2 or 3 steps which could translate to, at most, about 100 V. In priciple, this effect can be observed (to some extent) by the two different passes that were done. If the points from the two passes line up, then the HV is probably pretty accurate given the passes used different step sizes.
The values from pass 2 were scaled down by a factor of 2/3. This is because the values in pass 1 were accumulated from 2 scaler reads while the values in pass 2 were from 3. The scaled values are shown in all of the plots below.
All of these plots are scaler counts normalized to beam current (in nA) vs. HV. The blue points are from pass 1 and the red points are from pass 2 as described above.
Here is a 4MB Postscript file with all 384 plateau curves.
Here you can view all of the curves via WWW. They have been sorted into groups of 16 for your convienience.