Diagnosing false alarm

[PlumbMark]

I would like to diagnose and fix a faulty septic alarm.

The problem started in January of this year in the freezing cold winter.
My septic alarm turned on.
I called a company to come over and diagnose the problem.
They came and determined that my pump needed replacement.

While they were inspecting the pump chamber they also said that the wiring for the septic alarm "wasn't done properly"
and should be attended to in the spring (they don't do electrical repair work). They mentioned housing it in a box.

When they left (after replacing the pump) the alarm remained on. I silenced the alarm at the box, but the red alarm light remained on.
It stayed that way for weeks or months.
One day I went to the alarm box and reset/unplugged it, one more time, and this time the alarm stayed off.

Fast forward to today (months later).
I had a company come and pump out the septic chambers.

I asked the man to open up the pump chamber so I could test the septic alarm to make sure it was still working.
He lifted the alarm float and the alarm was activated. When he dropped it back down, the alarm stopped.
The water level in the pump chamber was very low, nowhere close to the alarm float.

About 30 minutes after the man left with his truck, the alarm turned on spontaneously!!!

I tried the reset switch and unplugged the alarm box, but when I plug it in the alarm stays on.
I have silenced the alarm, but the red light remains on.

Any ideas how I can diagnose/fix this particular problem?

TIA

 

[Matt30]

Chances are you have a junction box buried under ground near the tank. It's now filled with moisture and its tripping off the alarm. If someone were to put an ohm meter on it, the resistance reading would likely by very sporadic. The same can also be said for the float switch itself if it has never been changed

 

[PlumbMark]

Chances are you have a junction box buried under ground near the tank. It's now filled with moisture and its tripping off the alarm. If someone were to put an ohm meter on it, the resistance reading would likely by very sporadic. The same can also be said for the float switch itself if it has never been changed

Thanks Matt for helping out!

Today I went to the alarm box inside my house to take a continuity reading.
The wire to the alarm float read 16K, which was close to zero resistance, showing the circuit is closed.

So either the float switch isn't working properly or wiring somewhere outside is screwed up.

How could I tell if there is a junction box buried under ground?

To me, the biggest potential clue is that the 'spontaneous' tripping of the alarm occurred just 45 minutes after we tested the alarm float.

For months the alarm was stable and now 45 minutes after the float is played with it triggers on its own!