Even just a few degrees off perpendicular can cause a loss of signal or result in the detection of an unwanted echo reflection. This becomes even more critical when trying to detect targets that are near the range limits of a given sensor.
Is there an adequate clear path from the face of the sensor to the target?
Keep in mind that we’re dealing with sound-wave signals, which spread in the shape of a cone similar to the beam of a flashlight. So what constitutes an “adequate” path? The answer to that question can get a bit complicated. It depends on a few things, such as the sensitivity & pulse settings, the furthest distance to the target, and the reflective characteristics of any perimeter targets. To be on the safe side, you can always use the sensor's published specification for beam spread as a guideline (APG uses the maximum beam spread as the specification).
Is there anything physically coming into contact with the transducer portion of the sensor?
The transducer face vibrates when transmitting sound waves. Anything that comes into contact with the transducer portion of the sensor can cause coupling, where the vibrations reverberate through the contact material. This can greatly prolong the normal ringing process and result in signal interference and false readings (often jumpy).
Is the sensor mounted on a standpipe?
If so, is the pipe mounted straight? Even just a few degrees off-perpendicular can cause significant signal loss.
Does the standpipe have smooth interior walls?
Seams, burrs, weld slag, etc can reflect enough signal to be detected on its 2nd or 3rd bounce (back and forth between the target and the sensor). The imperfection may be inside the blanking zone, but it can appear to be a target at a distance equal to a multiple of the actual distance.