The problem is most probably false triggering caused by acoustic vibration. If the pulse frequency as shown on the meter jumps around, then acoustic vibration is almost certainly the problem. Pyroelectric sensors are sensitive to vibration, and they in fact detect acoustic pulses through the same physical mechanism with which they detect laser pulses. On the more sensitive scales of sensitive sensors such as the PE9 and PE10, they may be very sensitive to vibration. You can see this by setting such a sensor to a low energy scale (e.g. 2 mJ) and clapping your hand once, just above the sensor's surface; you will get a reading.
The Ophir PE-C series sensors have a trigger level that can be set to above the level causing false triggering but below the level you wish to measure. You may set the user adjustable threshold to above the noise level to eliminate the false triggering. An additional solution may be to put an acoustically absorbing material such as a thin piece of soft foam plastic under the base of the sensor to damp out any vibration; acoustic noise carries primarily through the base (rather than through the air).