Bed leveling problem  

Values

Unless I am missing something, The second list of values on the picture are not bad.  

Bed variance

Hey thanks cwbullet.

I see the same, but where does the bad "Bed Variance" value come from?

Mesh leveling

Define bad bed variance.  My values are similar and the mesh leveling is able to adjust.  

Posted by: @joku

Hey thanks cwbullet.

I see the same, but where does the bad "Bed Variance" value come from?

Are you actually experiencing any first layer issues? I don't see anything here that MBL shouldn't be able to compensate for easily. 

First Layer

No, I have now managed it a little better and have now left it as it is. I have no problems with the first layer so far. But I am interested in how this value comes about.

RE: Bed leveling problem

I believe that Mesh bed levelling, on a Prusa Printer uses an inductive sensor. to estimate the proximity of the build plate 

when you use the sensor to aid your adjustments for either the Nylock method, or the Silicone Sleeve method, or the Spring Method, 
you are relying upon these estimates. 

Every time you run the codes to estimate the proximity of the build plate to the sensor, you are likely to get a slightly different result, especially if you start the process from different starting points. 

the Pinda / Spinda, is quite good at estimating proximity, but it is NOT Measuring the separation between the PINDA/SPINDA face and the build surface, it is only estimating the distance / Separation. Chasing perfection with an inductive proximity sensor is likely to keep you busy for an extended period. 

If you achieve satisfaction, turn the build plate over, and the chances are that you will achieve a different set of results, even if you are using a smooth build plate. different areas of the spring steel build plate will have different magnetic properties, and turning the plate over is likely to re position these areas and disturb the estimates. 

you report seeing 10 degree differences between consecutive tests.
the M3 screws, used to locate the heatbed, have a pitch of 0.5mm per revolution. 10degrees rotation equates to approximately 0.014mm it's not a massive variance 

your last reported low point was -0.04mm below the centre point of the bed 
your latest reported high point was 0.02mm above the centre point of the bed. 

that looks like a total deviation of 0.06mm or  6 onehundredths of a milimeter. or 60microns. 
a human hair is estimated to vary in thickness between 17 microns and 181 microns (Wickipedia) and typical accepted thickness id 75 microns, so your total deviation looks to be smaller than a human hair.

not bad for the build plate of a computer controlled glue gun...     

 I currently have a silicone mod in it, before that Nylock mod and before that I tried with springs. no change.

And

I have an MK3S + that is one month old.

Why did you try all these mods?  How far out was the print bed initially?  What were the first prints like?

It reads as if your build had some basic issue that put the print bed out of line and which continues to do so -

BUT, some of your figures look reasonable.  Please show us a first layer calibration.

Cheerio,

My gut is telling me that probably none of these mods were necessary... From what I can see there's nothing here MBL shouldn't be able to handle easily. But then again, maybe it's just the baked beans I had for dinner talking...