Dear Peter, Bernhard sent me a good suggestion last night: he proposed switching the CAENs to their internal 40 MHz clock, and seeing what effect that has upon the artificial time structures. If the structures disappeared, it would then be worth looking at the shape of the internal clock signal and seeing what we could learn from it. Unfortunately, the problem persists even when the CAENs are operating on their internal 40 MHz clock. I've included the familiar IU-OU histograms from my internal clock tests; the results are the same regardless of which CAEN the hodoscope cables are connected to. I would summarize my findings to this point as follows: the artificial CAEN time structures are sensitive to the duty cycle of the CAEN clock, with a minimized effect at 50% duty cycle. They never disappear entirely, however, even when the CAENs operate on their internal clock. There seems to be an inherent timing problem among the CAENs' 4 TDC chips. Tom
Attachment:
40MHz_int_timestruct.ps
Description: 40 MHz histograms