Welcome to the SeeMeCNC Community

[img]https://lh6.googleusercontent.com/-zx2j ... o/onyx.jpg[/img]

red is low spot

yellow are high spots

i calibrate the bed with the same amount of grab (that i can tell by hand) by using scripts 1-4 (modified for 280mm beds)

the middle 200mm prints great!

Any thoughts?

Double check that all the mounting screws to the Onyx and base are nice and snug and as evenly tightened as best you can.

Unfortunately this will cause a complete recalibration to be needed.

But uneven torque will cause this issue.

Also they will relax over time as the wood compresses as will all the screws on the machine.


But you are not too far off at the moment.

Edward's post about tweaking tower positions sounded like it could resolve this, but I'm not sure anyone else was able to replicate his success.

Do the high and low spots move when you rotate the glass build plate?

I know that I'm the minority here, but I still have a hard time believing that such a small piece of glass can bend and warp just from its own gravity and the tension of some mini binder clips. BTW, you may try putting the clips closer to the high spots and further from the low spot as a test.

Do you have feeler gauges or are you just using a piece of paper when you calibrate endstops?

I just got a .051mm feeler gauge and used it to calibrate. I got a much more consistent first layer.

That's a tough print, but it may get better if you have more consistent endstop levels

Tonight i will rotate the plate, and i printed out a harbor freight holder so ill get exact measurements... more to come

ok so i rotated glass, re-torqued onyx screws. ripped the hot end off and put a 1" dial on it.

here are my target values give or take a +/- .001 for my fat fingers

scripts 1,2,3 (modified for 280mm bed) x,y,z : .042"
script 4 : .43" ish

so the steps for delta_radius stuff went well, but watch this video as i test the path to the edge.

I wrote this code to approach the screw holes

http://youtu.be/mpIuipEHogs

I dont know whats left to try i would like to use my entire bed not just the inner half

bubbasnow wrote:ok so i rotated glass, re-torqued onyx screws. ripped the hot end off and put a 1" dial on it.

here are my target values give or take a +/- .001 for my fat fingers

scripts 1,2,3 (modified for 280mm bed) x,y,z : .042"
script 4 : .43" ish

so the steps for delta_radius stuff went well, but watch this video as i test the path to the edge.

I wrote this code to approach the screw holes

http://youtu.be/mpIuipEHogs

I dont know whats left to try i would like to use my entire bed not just the inner half

Would you be willing to post that code so that people like me can use it to test their beds?

put this as script 5 or something

Ok, take this for what it's worth, ravings of a mad maker, too many brews... I'm going to throw it out there and let you all shoot it full of holes

What if the bed leveling were done mechanically to help compensate for software? Let me expand...

Rather than fasten the Onyx down, "float" it on a platform over a larger snowflake(That does sound mad ) This would allow it to expand and contract without restriction from the fasteners. Hopefully staying flatter. Over this is the glass(or whatever material you use) build plate. It would still rest on the Onyx, only it would be captured for registration by a frame circle around it that's fastened to the larger static snowflake with spring/lock mechanisms(yet to be invented?). Leaving enough space for the Onyx to move. A static probe is deployed from the effector to print height below the nozzle. This probe would run the calibration scripts, pushing the build plate and Onyx down at X,Y and Z. The "theoretical" spring/lock mechanisms would let the build plate/Onyx be pushed to print height and then lock in place. Eureka! as flat a surface as the materials involved Retract the probe and hit print

Problems I see:
-doesn't account for error in the software.
-Redundant? Is it worth the trouble? or is end stop leveling sufficient?
-spring/lock mechanisms would have to be precise.

I'm ready for the shooting now and remember to read the third quote in my signature below

After cutting a portion of the snowflake out to keep the power wires from affecting the onyx, I loosened the heatbed screws, brought the onyx up to 100° and tightened the screws down. Since then I haven't had any warp that affects the glass plate. I highly recommend this mod if you're having bed troubles.

lordbinky wrote:After cutting a portion of the snowflake out to keep the power wires from affecting the onyx, I loosened the heatbed screws, brought the onyx up to 100° and tightened the screws down. Since then I haven't had any warp that affects the glass plate. I highly recommend this mod if you're having bed troubles.

can you sketch something on a napkin for me so i know how far to cut?

When I did it initially I had just gotten pissed at the heatbed assembly and decided to punish it with a dremel tool. I later found that I wasn't the first to do this....

http://forum.seemecnc.com/viewtopic.php ... 1234#p6622

lordbinky wrote:When I did it initially I had just gotten pissed at the heatbed assembly and decided to punish it with a dremel tool. I later found that I wasn't the first to do this....

http://forum.seemecnc.com/viewtopic.php ... 1234#p6622

No donuts for you!

Batteau62 wrote: Ok, take this for what it's worth, ravings of a mad maker, too many brews... I'm going to throw it out there and let you all shoot it full of holes

What if the bed leveling were done mechanically to help compensate for software? Let me expand...

Rather than fasten the Onyx down, "float" it on a platform over a larger snowflake(That does sound mad ) This would allow it to expand and contract without restriction from the fasteners. Hopefully staying flatter. Over this is the glass(or whatever material you use) build plate. It would still rest on the Onyx, only it would be captured for registration by a frame circle around it that's fastened to the larger static snowflake with spring/lock mechanisms(yet to be invented?). Leaving enough space for the Onyx to move. A static probe is deployed from the effector to print height below the nozzle. This probe would run the calibration scripts, pushing the build plate and Onyx down at X,Y and Z. The "theoretical" spring/lock mechanisms would let the build plate/Onyx be pushed to print height and then lock in place. Eureka! as flat a surface as the materials involved Retract the probe and hit print

Problems I see:
-doesn't account for error in the software.
-Redundant? Is it worth the trouble? or is end stop leveling sufficient?
-spring/lock mechanisms would have to be precise.

I'm ready for the shooting now and remember to read the third quote in my signature below

Not a bad idea actually. The fact that the Onyx is held captive without the ability to move is what causes a lot of the doming effects.
As it heats it expands, it can't move laterally so it has no choice but to expand upwards. In my opinion your idea has a lot of merit.
In my case, I am going to use a 5/16" thick Mic-6 aluminum bed with a 1200 watt silicone heater controlled by a SSR. The heater is self
adhesive and sticks directly to the aluminum. I think that the Onyx would work if given some freedom to expand perhaps something as
simple as slots in the Onyx that are tightened only enough to hold it but still allow lateral movement.

Eaglezsoar wrote:

lordbinky wrote:When I did it initially I had just gotten pissed at the heatbed assembly and decided to punish it with a dremel tool. I later found that I wasn't the first to do this....

http://forum.seemecnc.com/viewtopic.php ... 1234#p6622

No donuts for you!

It looked at me wrong! Wait, no, I mean It came at me with a knife!

It looked at me wrong! Wait, no, I mean It came at me with a knife![/quote]

You are tooooooo funny!

Batteau62 wrote:... This would allow it to expand and contract without restriction from the fasteners. Hopefully staying flatter...

I think the BerryBot3d had a fixed print surface and the heater was held against it from the bottom with springs so it could expand downward. Similar idea to prevent the heater from pushing on the build surface.

I watched this video months ago, but here it is in case I'm remembering wrong.

[youtube]http://www.youtube.com/watch?v=gXUJQuNSLYw[/youtube]

I have always thought that a "floating" platform would seem to be much simpler, just never got around to fiddling with it. My UM uses four screws ran through springs to float the bed above the platform. It does require adjusting from time to time, but I can do that while it's printing the brim and get good results. And it's much more straightforward than adjust a screw, re-home and re-test.

MSURunner wrote:I have always thought that a "floating" platform would seem to be much simpler, just never got around to fiddling with it. My UM uses four screws ran through springs to float the bed above the platform. It does require adjusting from time to time, but I can do that while it's printing the brim and get good results. And it's much more straightforward than adjust a screw, re-home and re-test.

What is a UM?

Ultimaker

Nylocke wrote:Ultimaker

Thanks Nylocke.

My bad My students catch me throwing around acronyms sometimes, too, that they have to stop me and say, uh, what?

MSURunner wrote:My bad My students catch me throwing around acronyms sometimes, too, that they have to stop me and say, uh, what?

The Ultimaker is rated highly by various magazines. Did you buy yours as a kit or assembled? I was looking at them but when you throw in the monetary
conversion to dollars, they can be so very expensive. They appear to be open source but they seem to keep a tight reign on the files for the case and some
of the parts look to be proprietary.