Testing PMTs

R. J. Wilkes

3/28/05

This procedure tells you how to test the basic functionality of a PMT, before attaching it to a scintillator. You will measure the dark current (average output of the tube with no light input, a measure of noise levels), and the dark counting rates, as a function of HV and threshold settings.

Equipment needed:

• PMT
• HV supply
• Oscilloscope
• Coax connector with 50 ohm resistor
• Multimeter (with microampere scale for current measurements)
• DAQ card
• Dark chamber (lightproof box, with a lightproof access port for feeding HV and coax cables out). Note: it is not easy to make a box lightproof at the sensitivity level of a PMT!
• Small piece of scintillator (to fit in front of PMT in test chamber)
• The usual cables and connectors

Preliminaries: make sure PMT has clean face and check base for obvious loose connections. For preliminary tests, leave the plastic scintillator disk in the front (PMT face) end of the dark chamber to provide low-level light signals from cosmic rays.

Things to remember:

• NEVER expose a PMT to ordinary room light levels while it has HV applied; this may damage the PMT.
• After a PMT has been exposed to room light with HV off, it should be allowed to “cool down” in the dark for 1 to 10 minutes (depending on how long it was exposed to light) before attempting measurements; it will take a while for its photocathode to settle down.
• Similarly, after turning on power, you must let the HV supply, PMT, and its base circuit warm up for about 5 minutes before taking meaningful data. Efficiencies and HV levels are temperature sensitive and take a few minutes to settle. You can use this settling time to make preliminary measurements and adjustments.

Test Procedure:

1. Insert PMT into dark chamber and connect cables before closing the chamber. If necessary, use black electrical tape to light-seal the cable feedthrough opening.
2. Apply HV to the PMT: start at 1950V for the 12-stage (square base card) tubes, or 1350 V for the 10-stage (round base card) tubes.
3. Dead-or-alive scope test: connect the signal cable to an oscilloscope. Be sure to use a tee connector with a 50-ohm terminator on the scope input. If at first you see no evidence of signal pulses, try running the HV up as high as but not beyond 2000V (or 1800V for
4. 10-stage tubes).
1. If dead, put the tube into the “dead PMTs” drawer for later diagnosis.
2. If OK, turn off HV. Wait a few minutes, then open the chamber and remove the scintillator disk. Reseal the chamber and apply HV again.

Repeat the following steps for each HV voltage used. Start at the mid-range, typical voltage setting (1950 or 1350V), to get a feel for what you are doing. Then start at 1700V for 12-stage (1200V for 10-stage) tubes, and step up HV by 50V at each iteration. Stop at 2100V (1600V).

5. Dark current measurement: connect the signal cable to a coax connector that has a 50-ohm resistor on one lead. Connect an ammeter (initially set to 10 mA) between the resistor and the other lead of the connector, so the ammeter closes the circuit formed by resistor across the coax signal cable. Record the current (if too low to measure, just record “< 1 microA” or whatever the ammeter’s minimum is) for each HV voltage used.
6. Connect the signal cable to the DAQ card and set the comparator (discriminator) level for the channel used to 20 mV.
7. Use the LabView interface or the Linux DAQ program to set the card up to take a reasonable number of counts (at least 100 for each data point). Your goal is to measure the rate (counts/time).
8. Repeat, increasing the comparator threshold by 20 mV. Stop when you have reached 160 mV, or when the counting rate gets down to 0.1 count/sec.

When done, enter the data into an Excel spreadsheet, and make plots of dark current vs HV, and count rate (in Hz) vs threshold for each HV value.