Adding a radioactive check-source, like a thorium welding rod, may produce some interesting results. ;)
* Eating a banana: 0.1 µSv
* Background daily dose: 10 µSv
The calculation is probably more complicated, but as a back of the envelope estimation I expect that a banana increase the background radiation in a 1%: So it will be difficult to measure with a cheap equipment just counting the number of detections.
[If your device can measure the energy of the particles, perhaps you can see some peak?]
[If you log all the event for a day, with and without a banana, perhaps the 1% difference is statistically significant.]