This thread claims GF has said 0.001".
3 Likes
Glowforge - the 3D laser printer says:
- Positioning precision to 0.001” (0.025mm)
- Engraving up to 1355 DPI
2 Likes
This. The motors have no concept of units.
Well, they sort of do. The smallest step might be (for example) engineered to 0.001"… or to 0.02mm. If the latter, then I will get the greatest accuracy by measuring my own patterns in mm rather than inches.
one step is the smallest increment of distance… what every the distance may be
1 Like
OK, it’s clear that I should have examined my own question more clearly before asking. I don’t think the answer to this is public. Sorry about that.
Maybe to articulate @Purplie 's question a different way… (or just invent my own question)…
How big is one step of the stepper motors? 
We know it’s around 0.001" or maybe 1/1355" (since that’s the maximum engraving DPI)… but knowing exactly what it is can be helpful if you really want maximum precision when designing, cutting, and engraving.
For example, if it really is exactly 1/1355", and I make something 0.5" wide… well… the laser would have to cut at either: 677/1355" or 678/1355" – neither of which is exactly 0.5".
Practically, I’m having a hard time thinking of when this would matter but I can imagine in some applications it might.
4 Likes
As an aside, I tend to think of this stuff a lot for digital UI design, where pixel alignment of text, icons, etc can matter for making things always look crisp on a screen. Aligning design to-the-pixel can make a big difference – and effectively what we’re looking to understand is exactly how big the “pixel” is that the GF can reason about.
1 Like
most steppers have 200 steps… 1.8 degrees
4 Likes
True but a lot of (most?) steppers are being set up with fractional steps these days.
4 Likes
Not really, it really is the motors. Inches or mm, to enough decimal points, aren’t really relevant.
A related discussion happened a few months ago.
3 Likes
At some point you are out stripping the practical limits of the machine. Thermal expansion of the belts, backlash and slop, variations in ketfwidth and thickness of material… Just go with the stated tolerances of about 0.001 inch and you’ll probably be fine, if you try to chase any tighter than that you’re probably going to end up disappointed.
Fundamentally it’s a digital machine but the last steps are very analog, you’re dealing with a laser burning through organic material. There are simply physical limits to how accurate and precise you can be when you’re doing that.
10 Likes
Yeah, you are making sense.
2 Likes
Oh, not to mention that your machine is probably very slightly out of square, and we know that the entire frame can flex if your table isn’t flat and there’s evidence that shows that kerf can be different if you’re running along the vertical or the horizontal axis. I think @geek2nurse exposed that one.
She even found some differences in the way the laser behaves depending on the color of your vector. I’m not sure that was ever resolved, maybe she can chime in.
Anyway all that is to say that there are many ways for sub-0.001” errors to creep into your finished piece. Best to design for a little tolerance than fight for something that is not reliably attained.
6 Likes
I think I eventually convinced myself the color thing had to be a coincidence. I can’t imagine how it could possibly change the laser’s behavior – they’d have had to purposely put it in the code, and I can’t imagine they’d have done that. 
Also, my Glowforge at the time had some weird issue going on with inconsistent cutting. It wasn’t long after that it set itself on fire cutting earsavers.
7 Likes
I have found that the easiest clue to what system they are using is what comes out in reasonable numbers. If the distance is 500mm, but the cutting range is called 19.68" chances are they are using millimeters. If it is 508 mm or 20" they are probably using inches. Just normal human tendencies.
5 Likes
To me the chances are more likely in line with physical realities that they backed into.
I suspect that they designed the machine to be as large as they could inside the existing case size (itself limited by realities of shipping sizes, anything much larger would require pallets etc)… and then they found out the realities of acceleration which pulled the borders in to a maximum feasible size.
My guess has always been that they experimented to figure all this out, calculating the acceleration and then knowing how quickly they could actually pull it off without losing steps on the motors… much easier to just back into it via experimentation to maximize cutting area while maintaining accurate positioning.
So ultimately I’m guessing that the available size is not a nice round number.
5 Likes
Yes you right! So many variables though that makes the accuracy. 
Well said! In the past I’ve tried to explain that every material has a max useable resolution, beyond which is wishful thinking, but I think this is a better way of saying it.
Kind of like Google Maps reporting lat/lon down to the subatomic scale. GPS simply isn’t that accurate. 
5 Likes
Tell that to Spock 
This topic was automatically closed 32 days after the last reply. New replies are no longer allowed.