That's a shame.
I've thought of another experiment. If you can stand the scales on only three feet, by putting spacers under them to keep the fourth foot off the flooe, and standing on it inside the triangle so that it doesn't tip, and it reads steady, then the isolated foot would be what's causing the problem.
Perhaps for loads under 2 stone, only three of the feet are making contact with the floor.
That's not going to be easy, but I'll give it a go. Thanks for the thought.
I tried that, thanks again for the suggestion. As anticipated, it wasn't easy standing on the scales without tipping them but I managed it.
With foot number one off the ground, the reading was immediate and steady. The same was true with foot number two not bearing any weight. And also with foot number three. And foot number four.
With all four feet back on the (solid, level) ground and me standing on the scales, we were back at square one: a reading which wouldn't settle and flashed continuously over a range of about a stone.
I'm beginning to think it might be me...
Now that IS interesting. If the four feet are truly planar, they will only touch a truly flat surface, but for only three feet they will all touch any surface. Maybe your scales are on a completely hard surface, so the slightest rocking will raise one of them off it, and the software is too slow to recalculate. The next thing to try is to put the scales on a slightly resilient surface, but not enough to let the mounting round each foot touch it. A slightly squashy surface would let each foot find some sort of contact even if you move your weight slightly, and the scales should sum the four forces.
I'm surprised you said the reading was immediate - it sounds as if there is no averaging over time.
Actually it would be even better to stick some rubber on the feet so they always make some sort of contact on the uneven surface.
The trouble with that is the same as using it on a soft surface; an indeterminate amount of the weight on each leg is taken by elastic deformation of the the rubber, and does not go to straining the strain gauge.
Update: I tried the scales on a variety of surfaces: a wooden floor (where as far as I could determine , all four feet were firmly down), a very close-pile carpet, a rubber mat, and a tiled floor. In every case the result was the same: wildly erratic readouts which never settled.
Since the scales used to work perfectly, I can't see any other reason than some sort of electronic fault somewhere along the line.
Incidentally, I was mistaken earlier when I said that only two of the feet engaged with strain gauges: all four do.
The trouble with that is entirely inside your head. They tell you not to use a soft surface to make sure only the sensitive feet touch something. Putting rubber on the feet can only compress the rubber, with all the force going straight through it to the foot - there's nowhere else for the force to go. All the indeterminate forces add up to your own weight.
After your tests, I have no explanation of why your readings were steady via any three feet, but not with four. When you step off, does the display freeze with the latest reading, as scales normally do? If not then I can't think of anything you can do.
In an early post you said the feet were connected to the pcb by one wire, but then you said twin wires, which is what I would have thought. I've just examined my scales and each sensor has three wires, one red and two blue. Are your two wires inside a common outer screen which also connects to the board?
All four feet must always have sensors, otherwise the reading would vary depending on which ones you put your weight on.
As your scales do not have a glass platform, is there any possibility that the platform can flex slightly? If so, when you were standing on one corner to test three feet, it was not getting flexed, but when standing normally, the case flexes and so does the printed board which might have a cracked track which then opens.
Also, Have you recently installed a new wall-wort or HomePlug that might be causing interference?
It does, yes, for some eight seconds. And then the display clears.
It is twin wires. They're so fine that I thought they were single.
There is definitely movement, yes. So your suggestion of a broken track on the PCB sounds very likely.
Thanks to you and everyone for the help.
Apologies for the delay in responding. No, I haven't. And see my previous reply about your broken PCB track theory. That does seem very likely to me.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.