Water Pump Control - Presscontrol failure

Update 19 October 2019: the cause of the problem has been found, details are in a new video (part2 - solution):    • Water Pump Control - Presscontrol fai...  
Not much is to be found on the internet about a 'water pump control failure and what the cause/remedy can be. Hence this 'temporary solution' video.
My 'Water Pump Control' unit (already a replacement for a faulty DAB Presscontrol unit) started acting strangely.
4 times out of 5, the unit would pump water and then all of sudden stop. This repeated itself with irregular cycles.
When looking at the Water Pump Control unit, it would start in a normal way, but near the end (e.g. when toilet tank is getting nearly full) the 'failure' led would come on. With the failure led on, if water was requested (toilet), the unit would start and water flowed (!). As said, 4 times out of 5 - and then for some reason the pump wouldn't start. If the 'Restart' button was pressed, the pump would come alive again (for another couple of cycles), always with the 'failure' led coming on, near the end of a run.
After I had opened the water pump control unit for inspection, I noticed by coincidence that when I moved my screwdriver close to certain sides of the unit, the failure led would disappear. Eh?
This video from Farabielec (   • Domestic water pump controller تفكيك ...  ) helped me on my way.
Apparently the water flow is controlled by means of a magnet system.
My magnetized screwdriver somehow influenced the working mechanism of the magnet operation. Aha.
When I attached my screwdriver to the side of the unit with powertape, the problem was gone.
What a lucky shot that was... I tried other magnetized screwdrivers, magnets, nothing worked... only my favorite screwdriver did the job. Eh?
I took the printed circuit board out, and resoldered all solder joints.
Problem gone, failure led did not come on again.
Until I closed the unit ! Failure led came on again.
This printed circuit board contains two magnet switches, it seems. Before closing the unit, the printed circuit board was 'loosely' in place, not all the way pushed to the back of the plastic casing. Aha.
I took a plastic 8mm nut, widened it with a 9mm drill so it could fit over the 3 plastic guiding posts (see video). Hmm, the 'Pump on' led came on (after 230V was applied), but did not go away (after some 15 seconds or so, this led is supposed to go out).
I then cut/sawed the plastic nut in 2 halves (so I had 2 pieces of approx. 2.5-3mm) and slid them over 2 of the 3 guiding posts. Hmm, the Pump On led went out, but the Failure led came back on when water was requested.
In the end, I used 1 piece of a 2.5mm thick nut, and slid it over the right hand side guiding post. This did the trick. The angle of the printed circuit board vis-a-vis the moving magnet is correct for proper operation (in this case).
Could one of the discrete components have changed of value ?
A friend suspects the magnet has weakened over time, which in turn causes the malfunction.
If someone knows the offender (component/cause) for this kind of failure, let me know and I'll post it here.

If you decide to try this temporary solution yourself, it is your responsibility. You should have someone qualified look at/service the anomaly.
Enjoy the short video which shows the fault and temporary solution.

The musical score once more is by Rex Wheeler with his take on When the Levee Breaks (   • When the Levee Breaks - cover by Rex ...  ).