User name J. Volmer
Log entry time 04:15:48 on November 17,1999
Entry number 350
This entry is a followup to: 349
keyword=consistent HMS&SOS el. live timesFirst a correction to the previous entry: the plot does not show the dead times versus elreal, but the dead times, scaled to R=1MHz! That's why they are flat as a function of the elreal rate.
Following John's prescription from nucpilog entry #52, I calculated the widths of the EL60, EL90 and EL120 scalers of HMS and their differences. I arrive at the following numbers:
HMS SOS
60-90 41ns 36ns
90-120 14ns 34ns
60-120 55ns 70ns
Where available and meaningful, I also calculated the difference between 30-60, elreal-60 and pretrg-60, which all give the same results:
30-60 13ns 13ns
I take these last numbers to be the amount of time that the EL60 gates are wider (or longer) than the dead time of 50ns that we know are caused by the hodoscope discriminators. Then the gate widths (would) become
HMS SOS
EL60 63ns 63ns
EL90 104ns 99ns
EL120 118ns 133ns
I redid the electronic live time calculation with these gate widths. Fig. 1 shows the HMS (left) and SOS (right) electronic dead times (this time not scaled to elreal=1MHz) versus the elreal rate. The red dots use the EL60 and EL90 scalers, the blue ones the EL60 and EL120 scalers, and the green ones the EL90 and EL120 scalers. All of them agree rather well, and now, for both HMS and SOS, the rate of descent is about 5%/MHz (Rick Mohring quotes 3.5%/700kHz in his thesis, which is the same).
FIGURE 1
