Bed leveling problem  

Re: Bed leveling problem

Sounds similar to my experience.
I now think that the printer bed is warped and I've been checking it with a straight edge. Hopefully I'll finish doing that today, but so far it looks as though there is a big (0.18mm) fall off in the left rear corner and smaller amounts elsewhere.
Note that you need a 'proper' straight edge to check to this level of accuracy. The side of a steel rule is not flat enough.
I'm using the body of a vernier caliper which is a ground surface and the I've checked as flat on a granite work bench using a 0.04mm feeler gauge.

Re: Bed leveling problem

Yeah, thought about that too. The problem is I only have a metal ruler to check. And when I put the edge back to front I do see more and more light coming through towards the back side and then tapers off again about 2 centimeters before the back end of the plate.

Re: Bed leveling problem

For what it's worth I made a video of my checking:
In summary, using the co-ordinates printed on the bed surface I found the following gaps:
Left/right
Cold: 0.18mm at 25,180; 0.15mm at 25,130; 0.04mm at 15,80, and at 225,180 and at 240,130.
Hot (55C) for half an hour: 0.18mm at 25,180 and at 25,130; 0.06mm at 15,80; 0.04 at 220,180 and at 240,130.
Hot (55C) for about 4 hours: 0.18mm at 75,180 and at 25,130; 0.13mm at 225,180 and at 25,130; 0.06mm at 15,80; 0.05mm at 125,25.
Front/back
Hot (55C) for about 4 hours: 0.04 at 75,180 and at 125,180 and at 175,180 and at 225,180.

Re: Bed leveling problem

Hey,

just saw your video, excellent idea to do it like that. But even using a regular metal ruler I notice some major gaps throughout the platform. I hope the help desk can see my predicament as well, because besides this flaw I'm very happy with the machine...

Re: Bed leveling problem

I'm getting exactly the same symptoms. Have Prusa responded to this yet?

My kit was ordered about 3 weeks ago just as they hit the production hump and it was delayed. Are we seeing a QC issue here?

Re: Bed leveling problem

Any chance of a a firmware update that allows a greater range in bed level correction? Also, a 9 point bed level correction would be fantastic. I would be happy to be a beta tester.

Re: Bed leveling problem

As far as I know you can use bigger values for bed level correction if you use g-code start script instead of LCD tuning. (G80 L R F B, for example G80 L-60 R60).

Re: Bed leveling problem

I never thought of that. I'll give it a try. Thank you!

Re: Bed leveling problem

This worked great! I finally got no gaps and no smears in ALL calibration squares!

I reset Bed Level Correct to all zeros.
Next I did the V2Calibration program to adjust live Z to the best I could get.
Then I altered the G80 line in my 30x30-5 square program, entering the numbers that I previously had entered on the LCD panel as a starting point.
I did 3 prints of the squares and ended up with this for my G80 line:

Original: G80 ;mesh bed leveling

Starting point: G80 L-50 R-50 F50 B-5; mesh bed leveling

To: G80 L-60 R-100 F100 B-10; mesh bed leveling

Note: As I was playing with this, it appears that "100 to -100" is the full range. Entering a number of say -120 starts to go back in the direction you did not intend to go. But the numbers above worked much better than anything I've produced before and in fact are ridiculously good.

I will alter slic3r's beginning gcode script accordingly so these numbers automatically go into all future programs.

Thank you again for the tip. Hopefully this will help some others.

Re: Bed leveling problem

What I don't understand is if there is any sort of level issue with the bed, why doesn't the Pinda probe and the pre-print startup level checks account for this?

I find it really frustrating that all the trusted youtubers were singing the praises of how consistent prints were from the MK2 and so far it is proving less reliable than my FlashForge dreamer.

I have no idea how to measure the offsets I need for Bed Levelling Correction. I have tried all sorts of values and I can get filament to stick, but then I print something more complex and it messes up big time.

Re: Bed leveling problem

With inductive probe you get quite good precision in detection distance, but accuracy *may* differ a bit over different probing points, that's why there is G80 command available for fine-tuning the result.

Re: Bed leveling problem

I'm not sure how that helps me, when all I want to do is get consistent prints from Simplify3D.

Also, is it right that the bed level is probed once it has been pre-heated along with the nozzle? This seems a bit odd to me. Also, because it does it just prior to a print, the nozzle is also hot and deposits small amounts of filament curling out of the nozzle onto the bed.

When you Calibrate Z (or all axis) it does it with he bed cold. If conductivity changes with heat (which I would assume it does), which does bed calibration now also heat the bed? If it doesn't, then why not do it prior to heating the nozzle?

Re: Bed leveling problem

Just a quick update on my original post. I've been in contact with the (very helpful) people of the helpdesk. And after a few things I had to check, they were kind enough to send me a new print bed.

So it just arrived today, and of course I immediately installed it. The strange thing is, I have the exact same problem in the exact same place. Aaargh!

My XYZ calibration still says everything is aligned perfectly. Tried to adjust the "Live Z adjust", and it still barely touches the back part of the bed, while the front part of the test print is completely squished onto the bed.

So yeah, I could start editing the G code for every print I make, but I'd lose a lot of time editing it every single time. I hope the Prusa ppl pick up on this and make the "Bed level correct" margins a LOT bigger on the next firmware update.

Re: Bed leveling problem

Sounds as though one of the legs of your aluminium Y chassis could be bent.

Certainly worth checking.

Peter

Re: Bed leveling problem

I have the same problem.
The only difference is that it was more or less OK at the beginning. The v2calibration code produced a line compressed a little more at the front of the bed, but it was 'acceptable'.
now the line doesn't stick at all at the back while the nozzle almost touches the table at the front.
The correction works, but:
1. I have to put a positive number at the front setting and negative at the back (this is opposite to what the manual says)
2. 50um is by far not enough.

I did not move the printer, I wonder what have changed...
I also have a little play on the bed - but it looks like it is on the linear bearings.

The correction value grater than 50um would probably solve the problem for now, but i am not sure if it isn't something that would progress.