My robotic lawnmower (Landroid M1000) stopped to work, complaining it is out of its work area. I knew the wire had some weak spots, but where? It is a few years old, several hundred meters long, hidden under grass... An interesting challenge.

Time for some low cost high tech. Frist attempts with portable oscilloscope (Hantek 2000) have failed. The induction loop generator works at some 10 kHz, the frequency was too low to receive with a random wire antenna I quickly built. A loop antenna from the AV receiver was no good - it picks up the magnetic component, but there is hardly any current and magnetic field when the induction loop is broken. As an afterthought, I should have tried one-sided connection.

I picked up my Made in China XR2006 generator. Actually it is Made in Germany, since I soldered it. I have 2 of those cheap generators and both have a problem in the sine output. Seems somebody is selling out faulty chips. I used the sqare wave output for max amplitude and harmonics, maximum frequency (about 1.2 MHz) for stronger capacitive coupling to the antenna, powered the generator from a 12V LiIon battery from an old laptop and grounded it. The FFT mode of the SmartScope disappointed me, so I switched to the RTL-SDR v3 Dongle. It is an inexpensive (< 30€) piece of Software Defined Radio hardware capable of HF direct sampling 500 kHz .. 28.8 MHz. This finally worked! The peak in SDR# was not very strong (so I was not disturbing my neighbours much), but I was able to find the fault quickly and within some 20..30 cm, trying from both sides. It was a bypass I made 2 years ago. The stick I've marked it with has mysteriously disappeared...