Super Pinda mesh values not matching actual bed flatness  

RE:

Posted by: @baunegaard

So i have been struggeling really hard with the first layer ever since i assembled the machine. When setting the z height based on the first layer calibration, it needed to be much higher than when doing an actual print in the middle of the bed. After much frustration, i did the silicone mod in order to properly level the bed.

Now is seems really good, even though readings from G80 seems to fluctuate a little between each reading, i am able to keep the variation between 0.02mm

But if i take a metal ruler and put it on the x axis in the middle if the headbed, there is a clear gap in the middle big enough to slide a piece of paper under.

https://photos.app.goo.gl/7jkJZSa6QdMktnqy5

I dont get it, if i even out that gap manually, the readings will be way off on the probe. I seem to cannot win this.

Maybe the initial premise (the Heat bed is Warped) was incorrect , maybe the bed was flat and the sensor was wrong. (Assuming the steel ruler is not bent... Which is unlikely)

The Super Pinda, is a Sensor, not a measuring device. if you test the same surface repeatedly, you will get different readings, although the Supa Pinda is supposed to be more thermally stable than previous sensors, there is no guarantee that it is working as specified. 
the Super Pinda, is affected by magnetic influences, which is why the 7x7 Mesh Bed Levelling  has the option to ignore some of the test points because they are close to the bed retaining magnets. 
are you using 3x3 mesh bed levelling, or 7x7 mesh bed levelling. 
is it worth setting the bed level with the steel rule  and the silicone mod, then find other reasons why the Pinda is not seeing the bed as flat

In the days of the Mk2 Printer, which had the original Pinda, many people found that better results could be achieved by starting a print, with the Pinda over the middle of the heatbed and about 60mm above the heatbed. this tended to keep the Pinda temperature more regular when starting prints. if the pinda were located nearer the Heat Bed during the heating process, it tended to get hotter than if it were further from the Heat Bed. 
Similarly if the Pinda were located near the edge of the Heat Bed,  during heating, it will tend to be cooler. 

differences in temperature resulted in different sensitivities for the same Pinda.

Many of us found that by using the same starting point, we had fewer difficulties with first layer quality

so far, none of my 4 Prusa's have the variety of bed warp adjustments, that other people have invoked. 

your own cross comparison, between a Steel Rule and the results of the silicone method suggests that there is a disparity. 

regards Joan

P.s, I use 7x7 Mesh Bed Levelling

RE:

I think you are correct about the heatbed not being warped in the first place, i just tried unscrewing the heatbed and checking it with the metal ruler without it being fixed, and it is actually really flat just by itself, adjusting it based on pinda readings seems to make it much worse. Im really getting the feeling that the Super Pinda is defect in some way. I have tried using both 3x3 and 7x7 for measuring the bed, if im using 7x7 i get even weirder readings, it seems to read the points between the 9 regular points to be quite a bit higher than the rest, even though (based on the metal ruler) they are lower.