Platform Leveling

[stanfwen]

Hi All,

I am a new M2 user and have been playing with my M2 for about two weeks now. After having time to experiment with several prints, I have come to realize how critical platform leveling, and Z height adjustment are. I saw and old post on this forum about ideas for bed leveling modifications. Most ideas I have read about all deal with physically moving the platform to each corner and moving the platform up until it contacts the extruder tip.

Has anyone considered using an optical solution? What I had in mind was a unit mounted on the support that the extruder linear slide is mounted on, that would project a circle on the build platform. The unit would consist of a motor with an angled front surface mirror spinning and projecting a laser beam down onto the build surface. As long as the laser and spinning mirror were perpendicular to the X,Y plane, anything but a level build platform will be seen from above as distortion in the circle. In other words the circle would become an ellipse in the direction the plaform was low.

This would provide feedback as to the levelness of the platform without having to move it from corner to corner. Some sort of circle would need to be on the platform to use as a reference. Screws could be adjusted while watching the circle and when they line up with reference table is level. The Z height of the platform could also be determined by the size of the circle.

Just wondered if anyone has tried anything similar of can think of a reason this would not work.

Thanks

[pyronaught]

It seems like a lot of work to do something that can more easily be done with more accuracy using a dial indicator. I don't know that the circle would distort enough to be noticeable with the tiny angles you are dealing with when leveling the table. A dial indicator takes very tiny changes and greatly magnifies them, whereas I don't think there would be much magnification with the circle distortion. If one corner of the table is down .2 mm relative to the opposite corner, how out of round is that circle going to be? Not to mention the cost and complexity of a laser, motor, mirror, mounting frame etc. compared with a $15 dial indicator that can easily be attached with two bolts when needed.

[jimc]

i gotta agree with pyronaught. the bed only needs to be out .001". you would never see that. a dial indicator or just a simple set of feeler gauges will get you there and can be had for 5 bucks at your local napa. the gap i tell everyone to start with is .005". if you get the level and gap close then you can print this file i made and fine tune it while its printing it.

http://www.thingiverse.com/thing:180970

[jsc]

There certainly are printers with easy bed leveling procedures. I was just watching a video of the Printrbot Metal on which you can attach a z probe, that will test your bed in different locations and automatically adjust the layer height for the first layer. There's even support built into our Marlin firmware for z-probe leveling. But as far as I can tell, nobody has done any such mods for the M2 yet.

The main problem with any such add ons is that it solves a problem that many of us have already gotten used to addressing, each in our own way, and that needs to be carried out very rarely. I think, too, the kind of people who choose an M2 over other printers are those willing to prioritize a sort of Soviet locomotive approach to build quality over ease of use. It falls between the printers being marketed to consumers, where ease of use is necessary, and the printers being built from scratch by Rep Rappers, to whom the printer features themselves are nearly as important as what they will do with it.

(My own preferred method for bed leveling is nicely demonstrated here in a video by MakerGear. It is very fast and easy if your glass is clean.)

[stanfwen]

I agree that dial indicator might be more accurate. I do some CNC work so I am familiar with their use. I think the advantage i was looking for was the ability to speed up the leveling process by getting feedback on what adjustments were doing without moving the table around. I designed a prototype and built one in a few hours with < $10 worth of parts. The unit can be permanately mounted on the M2 without interfering with anything else. As far as magnifying the error, the laser is projecting an 8" circle from 1" above the platform, so errors will be magnified. The laser module I used for the prototype was a very inexpensive one ($3) and would require a lens to make the circle very thin.

[Capt. John]

How do you set absolute level with the laser?

[pyronaught]

Well, it's good that you are actually trying this since that is the only way to ever know for sure. The rest of us are just theorizing.

Since the bed is leveled in relation to the X-Y plane that the extruder nozzle operates in, how are you aligning your laser circle to that same plane? I guess you'd have to first level the table using the traditional methods and then adjust your laser to align with the calibration circle on the bed. Of course now THAT will also have to be periodically checked to make sure it is also level with the extruder plane, otherwise your level checker could itself get out of level. If this method is accurate enough, the advantage would be not having to move the extruder head around. You should be able to get the table level faster with instant feedback from the circle distortion.

[stanfwen]

Capt. John,

The way you set the bed level would be to insure that the circle being projected on bed is perfectly round by comparing it to a reference circle printed on the bed. The circle will only be perfectly round is the bed is exactly perpendicular to the laser. If the bed is not perpendicular (not level) the circle will be an ellipse. Think about an object being illuminated by sunlight. The only time the object will not cast a shadow is if the sun is directly overhead. If not a shadow exists and the farther away from perpendicular the object is the longer the shadow.

Hope that explaination helps.

Pyrnaught,

I printed a model with bores for the motor and laser perpendicular to the mounting surfaces. I then mounted the bracket to the same surface the extruder head is mounted off of. If that is not accurate enough to get a good level platform, I would probably need some adjustments on the mounting as you suggest. It would be interesting to place jimc's leveling print on the platform and see how much distortion I get in the circle for the different height lines.

IMG_2751.JPG (80.77 KiB) Viewed 12137 times

IMG_2755.JPG (90.1 KiB) Viewed 12137 times

[mathisyourfriend]

"I think, too, the kind of people who choose an M2 over other printers are those willing to prioritize a sort of Soviet locomotive approach to build quality over ease of use."

We won the "Cold War" but still use their rockets to get up to the Space Station.

I don't know why but your comment made me laugh out loud. I can't believe Rick et al are commies. Their printer design is too cool.

[insta]

"I think, too, the kind of people who choose an M2 over other printers are those willing to prioritize a sort of Soviet locomotive approach to build quality over ease of use."

We won the "Cold War" but still use their rockets to get up to the Space Station.

I don't know why but your comment made me laugh out loud. I can't believe Rick et al are commies. Their printer design is too cool.

I think it was more how the Russians were like "truck with 28 wheels got stuck in mud, so we made truck with 30 wheels" -- although I also don't get how it applies to the M2.