Steps to be taken in making and testing a small scintillator counter. Kari van Brunt June 10, 2003 These are based upon Kari's experience in making a small scintillator counter in the past week. I. Calculations: 1. Estimate No. of Cosmic Rays (see PPB/PDG,pp. 180) If detector is 1/2 m**2, flat, how many soft, hard, total? [This is how many we expect!] 2. Delay time by adding cable? How far does light travel in 1 second? How many feet is that? How long does it take to travel 1 ft.? What if n = 1.6? 3. How much charge is delivered? If signal is 1 V, and it travels down a 50 ohm cable, what is the current? If the signal lasts 5 ns, how much charge? If gain is 10**7, how many photoelectrons? 4. What does photoelectron mean? 5. How many photoelectrons are delivered to the photocathode? How many photons does 1 cm thick scintillator yield? How many have appropriate angle of incidence for TIR? What is quantum efficiency of photocathode? Suppose we lose 1% each time they hit the edge of the scintillator. How many do we lose? [lost = 1-e**N_bounces] How many bounces are there? Liouville Theorem says you can't put the same number of photons in a smaller phase space (where phase space is the product of the available delta x * delta y * delta theta_x * delta theta_y). If our ligh guide goes from 20 cm -> 2 cm in diameter, by what factor do we keep photoelectrons? 6. We have a signal of 10,000 cts/s. If noise is due only to Poisson statistics, how much uncertainty is there? (sqrt(N)) 7. How many data points fall within your uncertainty? (1 sigma; 2 sigma; 3 sigma?) ( > 3 sigma). (For good data, expect to be no more than 3 sigma away..) 8. A signal shows up in bin No. 75. If each bin is 0.25 pc, how much charge was deposited by that signal? 9) A 1 pe signal is found in bin # 22? What is the gain? 10) Accidental occurrence rate: R1 * R2 * delta time. II. Procedures. 1. Build counter. a) Clean light guide and scintillator (we didn't do this) b) Prepare light guide for gluing by taping plastic flush with the edge of the surface to be glued, so you catch excess glue. c) stand lightguide on ring stand with part to be glued up. Tape it to ring stand. d) Mix epoxy on foil. e) line edge of light guide with epoxy. f) set scintillator on top of light guide and hold. DO NOT HANDLE SCINTILLATOR DIRECTLY! g) allow it to set. h) wrap foil around combo, leaving a layer of air. This way you don't change n for the scintillator. i) put cookied on PMT j) add PMT k) push tightly against block, secure with metal ring. 2. Test the counter I) a) know how to sue scope b) take anode signal and put in ch 1. c) triggering on 1, vary the trigger and look at different signals. d) find the 1PE signal. e) trigger more negative to find cosmic ray signal f) hook oscilloscope to counter. g) see how many cts/sec do you see? (Ea. time scope is triggered the scaler counts) II. At this point, can do 2 things: a) put small ctr in coincidence with large counter, and look at coincidences on oscilloscope, or hook small counter to logic. b) place counters in coincidence (want rays to only pass through scintillator) c) use scope to count rates in S1,L1, or both. How do coincidences change with threshold? Investigate 1 p.e. signal and cosmic ray signal. How can you see a coincidence on scope? How likely is it you'll see a 1pe signal coincidence? (Calc. 9) d) make a threshold plot. change the threshold and record cts. III. Test Counter Cont'd. Now that you have convinced yourself you can see signal, let's do some of the same investigations with logic instead of scope. a) send dynode signal from S1 to discriminator, and L1 to discriminator; output to coincidence. count rates in S1. (is it what you'd expect?) count rates in L1. (is it what you'd expect?) count rates in coincidence (is it what you'd expect?) b) test uniformity of counters. take a number of rates for each ctr. (10 or so) calc. mean find uncertainty find out how much each differs from mean (1 sigma, 2 sigma, 3 sigma, etc) (expect 32% within 1 sigma 5% within 2 sigma 0.27% within 3 sigma (PDG p. 242) create a scatter plot that show cts vs time taken. note any trends. does the number depend on time? . Add CAMAC cd c:/dtake dt 3 returns At menu, type sta to start choose n (# events) choose s at the end of run: save data with: DDUMP save histogramst with : HDUMP (choose a run not already in use, and use the same run number for the data and histograms.) hit return twice at menu, go to dos. del wat.dat load your data ("dlo") look at data ("dis") 1 for ctr 1 ADC 2 for ctr 2 ADC 3 for ctr 1 TDC 4 for ctr 2 TDC (note above could be changed by the dtake setup, but this is how they are setup now.) Here you can try to resolve the 1PE signal, pedestal, peak charge deposited, # cts within a bin range; How do light leaks affect your data? How well does the ctr perform wrt light leaks? (try measureing cts w. lights on,off, oreplastic blankets, lights on, off, etc.) III. Theory 1) How does the oscilloscope work? how does CRT work? discovery of e-? 2) how does the PMT work? 3) how does scintillator work? 4) how does light guide work? 5) what does discriminator do? 6) how does coincidence gate work? 7) why invert signal from dynode? 8) why factor 2/3 difference from dynode than anode?