Fast Scan

  Some troubles arise with the beam stability, apparently a 60 Hz parasite was affecting the shape of the beam, making it 'breathing'. To investigate the problem more precisely, a temporal scan of the beam with quite little granularity (around 1 kHz) was required. Our standard acquisition system was limited to less that 100 Hz. Some other routines were written, that bypassed the EPICS SNL system completely, using very nasty empty loops for delays. Because of this system, every low priority tasks (as channel access) are almost stopped during the acquisition, but as a sampling lasts less that 2 seconds, we can afford it.

The average counting in PMTs being around 1 kHz, the total counting per sample is less than 10 (by summing every PMTs). So the direct view of the result is meaningless. A Fast Fourier Transform (FFT) routine was written to extract periodicals informations from the background.

The program is almost done, except for the GUI part, the Figure 3.2 has been done with it.