# What happens if a PMT (photmultiplier tube) is run below its operating voltage?

I have this Hamamatsu PMT that says its operating voltage is 1500 V. However, it has serious electronic/background noise at LLD=0 at that operating voltage. If I lower it down to about 1.02 KV, noise is supressed by much. My question is, is it always necessary to run a PMT at its operating voltage? What happens when one runs it at a significantly lower bias, like 1000 V or 700 V? I feel like I swamp my detector a lot less using a lower voltage, but I'm not sure that it's correct to run a PMT at any voltage but its operating one.

• Are you interested in discrete counts, or in the photocurrent and is dependence on other variables? If the latter, I'd worry about possible nonlinearities. – Emilio Pisanty Jul 17 '17 at 20:50
• @EmilioPisanty I'm interested in the total number of counts given a certain LLD, but at the operating voltage of 1500 V, I'm seeing a huge pulse (larger than the noise pulses) on the oscilloscope. I'm trying to figure out the lowest I can go without corrupting, if that can be possible, any counts. – Betsy Jul 18 '17 at 5:42

If you can still see your signal, by all means use the lowest V that gives enough gain - but the gain is very sensitive to supply V

The photocurrent multiplication factor is a high power of the Volts applied to the tube

This also makes power supply noise an important consideration

https://www.hamamatsu.com/resources/pdf/etd/PMT_handbook_v3aE.pdf

tldr:

http://www.bo.infn.it/ams/Hamamatsu-PMT.pdf

Since photomultiplier tubes generally have 9 to 12 dynode stages, the anode output varies directly with the 6th to 10th power of the change in applied voltage. The output signal of the photomultiplier tube is extremely susceptible to fluctuations in the power supply voltage, thus the power supply must be very stable and provide minimum ripple, drift and temperature coefficient.