I am not an antenna range expert, but have done some work in the field (no pun intended). You did not mention the frequency or bandwidth you were measuring or the power levels, which could be helpful to me or someone else trying to help you figure out the source of the variation.
Assuming that you are not making these measurements inside an anechoic chamber, then one source of variation can be due to multipath if there are many reflective objects in the area, especially vehicles that could be in different places for each measurement. Your dish probably has a fairly narrow beam with lots of sidelobes in the proximity of the dish, so changes in geometry within a few feet of the dish can change the main lobe shape, especially if the dish is being pointed close to the horizon where it might not be designed to work.
Also, having the receive horn too close to the dish can increase sensitivity to geometry changes. There is probably a rule of thumb like having at least 10 dish diameters between the dish and the receive horn. I will defer to experts for that rule.
Another source could be if you are using a power detector at the low end of its sensitivity.
Have you confirmed that the power from the source is stable? Are you measuring a single tone or a modulated signal?
An unlikely but possible cause could be if the dish just happened to be pointed toward the sun and the noise floor could vary enough over a short period of time if all the conditions (frequency, bandwidth, etc.) were worst case.
Anyone else have ideas?
- Kirt Blattenberger