The key is that the DCF77 signal is weak, and the lawnmower loop is very close. It's a classic example of poor signal-to-noise ratio, made worse by the DCF77 receiver being within the "near-field" of the lawnmower loop.

You could improve the S/N ratio by:

- separating the wiring for the two systems as much as possible, and fitting a mains noise filter to keep the loop signal out of the mains wiring.

- putting a remote (ferrite rod) antenna on the DCF77 receiver and experimenting to find a quiet location (possibly the interference can be nulled out by rotating the antenna).

- attending to the shielding, earthing and balance of the lawnmower loop (especially shielding and twisting the feed wires).

You could document the problem and write to your local radio licensing people. Unfortunately, I expect that they would consider that neither the DCF77 clock or the Lawnmower Loop come under their jurisdiction.

You could also write to the manufacturer of the loop and ask for their assistance. They may be surprisingly helpful.

Finally (if you haven't already) you could research the methods that Radio Amateurs have developed for receiving VLF signals in noisy domestic environments.

Bottom line: In my experience, correctly earthing the loop and shielding the feed wires should do a lot to eliminate the interference.

But as before, you need to understand the modulation characteristics of the lawnmower loop in detail. What bandwidth does the system actually need to operate? Can the interference be reduced by fitting a better low-pass-filter to the loop transmitter? Can you put a notch filter at 77.5 KHz?