I’m trying to tram my machine a little bit better. Many videos use a piece of ‘float glass’ which as I understand it is pretty much all glass that isn’t hardened or tempered. Just regular old glass.
Everybody wants glass that’s flat, but what’s truely flat? Essentially, what do I buy and how do I know it’ll be flat enough?
Float glass is glass made using a specific manufacturing method: the glass is literally floated on a bed of molten metal.
I’d suggest that the easiest way of finding the right glass is to look at glass 3D printer beds from stores that cater to 3D printing folk. It might not the cheapest way but it should be around $20.
As for “flat enough”, honestly I’m not sure what you’re using it for, so it’s hard to say. I guess you put it under the spindle when tramming so you have a smooth surface? I wouldn’t bother, just tram against the machine bed, the bed needs to be flat anyway.
Thank you for your reply. I’d be using it for tramming, like in this video:
Unfortunately, the glass on my 3d printer is tempered and tempered glass is ‘bumpy’ as it’s been through rollers. I admit I had assumed it was tempered glass (which it is) but some other printers seem to use borosilicate glass so maybe I could buy one of those.
Hmm, it looks like in that video, the machine isn’t properly square to the table, or the table isn’t square, so he shims up a glass plate to make a reference surface.
This will result in a trammed spindle but it also means the machine is cutting on an angle relative to your bed, so your workpieces are never going to be really accurate anyway.
I don’t think that makes a whole lot of sense.
Instead, get your machine square to the table first, then tram your spindle. If you do that your machine will be more accurate and you don’t need a glass plate.
As for the bed, sorry, I assumed all 3D print beds were the flat kind. TBH, a wobbly 3D print bed doesn’t make sense to me either…
I agree that bumpy 3d printer glass isn’t great and I’m sure the makers of my Ultimaker 2 know that. I’d guess it’s not that much of an issue for the first layer of a 3d print. At least on a hobby machine.
“Tempered glass units are heat-treated in a horizontal oven and contain waves created when the units come in contact with the rollers during the heating process”
I could post another 10 videos of people tramming CNCs with sheets of glass, including on Shapeokos.
Has anybody here used glass or do people only use glass on Youtube?
I used a piece of plate glass from Home Depot. It was 3 bucks and flat enough for my needs.
I think the most popular Shapeoko tramming process is probably the one described in the book.
I had a slightly different take on the “glass” recommendation.
First to answer your question I bought a plain/simple 8x10 glass panel, it was like $3.25 from Lowes.
I think the glass recommendation is because it’s simple and is usually flat enough. Another positive property of glass (and this is just my take on the recommendation) is that it is smooth so your measuring tool (when being rotated) glides easily and that makes the measuring easier (and more accurate?).
The reason they use tempered glass it’s being used as a heated bed, if it breaks you don’t want shards of glass flying around your 3d printer. And those waves from the rollers are NOT NORMAL. Good selling point thou if you can use it and don’t care about waves.
For our uses, something that is flat, regular old window glass fits the bill perfectly. Anyplace where you can get glass cheap is flat enough.
Yep,
As Lucas says, getting everything trammed is a multi-step process of checking all your V Wheel tensions etc., surfacing the spoilboard, tramming on the now hopefully flat to the machine X and Y axes and then finally re-surfacing to lose the ridges.
It appears that flatness is limited by wave formation in the float glass, but at ~0.1um should be almost flat enough for Lucas not to grumble too loudly whilst setting up his new not-a-Nomad 
Your point about tempered glass is incorrect. It was my assumption also, but not all 3D printer beds use temperered glass. Borosilicate glass has excellent thermal tolerance and doesn’t give a hoot about 3D printer bed temperatures and it’s used for a number of 3D printers.
Thank you for the scientific paper; I can’t say I was expecting that but it’s an excellent source.
We should all aim for 0.1um,
If what you’re after is easy tramming I’ve found that just having a larger radius between the spindle and the dial indicator was the most effective for me dealing with surface condition, bumps, T tracks etc.
I 3D printed a simple bar to clamp my dial indicator at one end and a 6mm rod at the other, it doesn’t need to be axially aligned with the spindle as you’re going to rotate it anyway so any up / down error in the indicator clamping is constant whichever direction you go to.
You may have more ‘fun’ when you come to tram front / back and try to rotate the X beam around the Y plates and discover there’s not enough wiggle, some shims behind the lower V wheels (assuming you don’t have a pro) are one good fix for that.
Since you mentioned my obsession totally reasonable habit…
The paper you linked says that float glass has a Rz of 0.1um, but Rz is a measure of surface roughness, not flatness. Your glass plate could be a cylinder and still have Rz of 0.1um.
However this page specifies a flatness of 5-10 fringes per 25mm at 633nm, which I think corresponds to 1.6 to 3.2 um per 25mm, which is 64 to 128 um/m.
To put that in perspective, a DIN 876/00 surface plate has a spec of 4 um/m.
So probably good enough a reference surface for a Shapeoko but not good enough for my insanity very normal practice of using a um dial indicator for everything.
You are adjusting a plane to be parallel to the gantry’s plane. yes it would be preference that your table bed is but with the typical bed materials (MDF) and slop in the frame mounting points, that isn’t going to be “perfect”. At least for the SO3, this might have improved with the hybrid table in the newer models.
Also I see doing the surface, tram, surface to be “imperfect”, you’re just getting invalid measurements inside the cuts tram error at two different points unless you ensure you are measuring the same spot with in the cut pass in the axis you are inspecting.
I use a B grade granite straight edge, shimming it to be parallel to the gantry’s constrained motion.
“[im]perfect” is all perspective to the operator and the requirements/tolerances they need 
I went with the 3D printer bed glass, suggested by Lucas. Specifically Borosilicate glass:
My dial indicator shows that the glass plate is, as hoped, not at all bumpy.
I surfaced the spoilboard and then put the glass on the spoilboard. I inserted the overpriced but apparently very useful tramming tool. Spinning it shows a 5 thou (0.127mm) tilt in one axis. I guess that as the glass is sitting on the peaks of mdf tram marks, there’s no need to setup a reference plane as people here suggested. Beats the hell out of trying to use a dial indicator on the mdf spoilboard.
Tramming tool:
https://www.machine-dro.co.uk/edge-technology-mini-pro-tram-system-with-1-4-shank
Thank you for your help, everyone.
This topic was automatically closed after 30 days. New replies are no longer allowed.