New Build - Root Cause: Onyx bed is domed/convex

[BobH]

I'm at the stage of squaring the towers to the bed in my new Rostock build. The problem I'm having is that all the towers are angled inward by about .5 deg using a digital angle gauge. If you get one tower perfectly square then the others are forced out of alignment even more. Is this "close enough"? The easiest option that I see to correct the issue would be to take off the top tower mounts and elongate the t-nut holes enough to allow the towers to "lean" outward more. Any other ideas? TIA - Bob

[mhackney]

Since you've identified this, I would attempt to rectify it. My V1 is a little different than your V2 but if you can elongate the holes and shim to remove the lean, you'll be much better off. The aluminum from soda cans makes excellent shim material.

What I did on my build that I think will work for a v2 is to install the towers in the base and get them all perpendicular to the plate (no lean). Then loosely fit the top and see where you might need to shim or elongate mounting holes. The idea is to get the top to clamp into place without forcing the towers out of position.

[BobH]

Thanks for input. If i did my Trig. right a 1/2 degree lean equates to about .25" over the length of 30". Better to fix it now. Thanks again! - Bob

[BobH]

Here's what I'm up against so far: the center line of the extrusion & the center line of t-nut bolt holes is off by quite a bit. More than I'm willing to hog out. I'll see if I can shim the base supports to get a decent angle.

[img]http://www.wetconcepts.com/rostock/X_Tower.jpg[/img]

[img]http://www.wetconcepts.com/rostock/Y_Tower.jpg[/img]

[img]http://www.wetconcepts.com/rostock/Z_Tower.jpg[/img]

[Tonkabot]

I would try spreading the base of the towers out - possibly in addition to moving the tops out. Seems like you could do a little filing in the square tower holes in the melamine, and then shim it out a little.

I don't understand why - in the top and bottom plate the square tower holes should line up. maybe if you clamp them together you can 'port match' them (what you do with your intake manifold and head to get free horsepower).

[BobH]

Update: If I remove the Onyx bed assy. and just square off the melamine base surface everything is fine. I'm investigating why the Onyx bed surface is so far out when it's mounted...

The Onyx bed is convex or humped in the middle. On one edge there's at least a 3/16' gap from the center. Should I ask for a replacement, or forge ahead a wait until I apply heat to see if it flattens out?

[Tonkabot]

So if you remove the onyx and snowflake, you should be able to put in the towers square and tighten them up. Then when you put the snowflake and onyx in, do not over tighten. also You should have the glass on top of the onyx, and the glass is flat. I don't think anybody prints directly onto the onyx PCB. PCB's are not all that flat - especially when you start bolting them down to something, they are pretty flexible.

[BobH]

So if you remove the onyx and snowflake, you should be able to put in the towers square and tighten them up. Then when you put the snowflake and onyx in, do not over tighten. also You should have the glass on top of the onyx, and the glass is flat. I don't think anybody prints directly onto the onyx PCB. PCB's are not all that flat - especially when you start bolting them down to something, they are pretty flexible.

The issue of the convex Onyx surface is then translated to the glass if the clips are installed to hold it in place. The glass then bends to the pressure of the clips. So the end result is a convex build surface. I would predict that the glass would eventually break from the bending stress.

The gap in the picture is a little over 1/16" between the onyx & the glass with no clips...

[img]http://www.wetconcepts.com/rostock/Onyx_Glass_Gap.jpg[/img]

[BobH]

I could try and flatten the onyx in my oven using the glass plate as a reference surface. Any ideas on the glass transition temp for the onyx pcb without damaging it in the baking process? Shouldn't the onyx relatively flat already from the factory/SeeMeeCNC?

[Tonkabot]

Maybe I am wrong, But I am thinking that when you clip the glass to the onyx, the onyx will flex to the shape of the flat glass, and not the flat glass bending to the shape of the onyx fiberglass PCB.

Try bending the onyx with your hands, and then try bending the glass with your hands. I am pretty sure you will easily feel they onyx flexing, and you will not be able to detect the glass bending.

Take a straight edge on the glass, then put the clips on it. did the glass bend? You have the borosilacate glass and not some plain window glass, right?

[ I have not tried this for myself, I have always assumed the glass does not bend - at least not enough matter. I will accept being proved wrong.... ]

[Tonkabot]

Also, what I think your saying is that your towers are tipping inward at the tops when you have the onyx attached. This would imply that the melamine is bending to match the shape of your onyx.

I don't remember - is there a mounting screw in the middle of the snowflake or just around the perimeter? If there is one in the center, you could shim it (or the edges) to force the melamine into convex or concave shape at well. You could always add a block in the center under the snowflake to 'shape' it [at least convex - it would need a screw to make it concave]

[BobH]

I've asked SeeMeCNC to replace the warped/domed Onyx bed with a flat one.

Thanks for your help in troubleshooting Tonkabot. Here are some answers to your questions:

The glass bends to conform to the onyx curvature when clipped. Since the onyx is fastened to snowflake & base it doesn't have the ability to bend, so the glass is stressed to the shape of the bed. There are no fasteners or ways to shim the bed in the center. The bed & snowflake only has attachment/mount points around its perimeter.

[Tonkabot]

Interesting. I would have thought that the PCB would be [a lot] more flexible than the glass. Sorry to doubt you.

All the PCB's I have ever dealt with have been reasonably flat, although the standard thickness (.063 or 1/16") is certainly easy to flex. I have made some .125 boards that are much stiffer, and you see boards for cell phones and tiny stuff that run .032 or less (just checked an iphone3 board, and it is .025 thick). If you left one with weights on it in a bent shape long enough, I am sure it would retain the bent shape [heating would probably help]

[Jimustanguitar]

Interesting. I would have thought that the PCB would be [a lot] more flexible than the glass.

The PCB is more flexible than the glass, but the pcb is screwed down in 6 places, so if it's domed, those 6 screws kind of lock the dome in place.

The trick I've heard before is to loosen the screws, heat the bed, and press down on it with a known straight object like a level while you tighten the screws back up.

I also suspect that some folks are fighting a wad of wires under the center of the bed, but obviously that would vary by case.

[BobH]

The trick I've heard before is to loosen the screws, heat the bed, and press down on it with a known straight object like a level while you tighten the screws back up.

I also suspect that some folks are fighting a wad of wires under the center of the bed, but obviously that would vary by case.

No wad of wires underneath. It's domed just sitting on a table top. I could continue the build and hope for the best, but I would be pretty pissed if I couldn't fix it and had to rip it all out again just to replace it anyway.

It seems to be a very key component to have as flat as possible given all the gymnastics others are going through in the "Unsolved Mystery. Weird Z0 behavior around build perimeter" thread & Tips & Tricks forum for Z compensation via probing & software.

[mvansomeren]

Interesting. I would have thought that the PCB would be [a lot] more flexible than the glass.

The PCB is more flexible than the glass, but the pcb is screwed down in 6 places, so if it's domed, those 6 screws kind of lock the dome in place.

The trick I've heard before is to loosen the screws, heat the bed, and press down on it with a known straight object like a level while you tighten the screws back up.

I also suspect that some folks are fighting a wad of wires under the center of the bed, but obviously that would vary by case.

I've found this to be true. My Onyx was slightly domed as well but if you lossen up the mounting screws and then clip on the glass, the Onyx bed conforms to the glass, not the other way around. I imagine that after a few times of use, the Onyx will remain flat on its own.