XY home position

The questions of “is there an origin point” and “are there stops” shouldn’t be able to be turned into a 70-post thread.

We’re like Glowforge speculators here, trading theories about an object that already exists.

The answer to these questions could be as simple as "yes, and here’s a photo of them…

Or the answer could be “The origin point is generated each time the laser is powered up, so once you turn off the laser the origin point is lost. There are no physical stops and no provisions to add them.”.


I thought (although I could well be mistaken) that there was some discussion of the head-mounted camera being used to find 0,0 by looking for some logo or other (remember, you don’t have to find 0,0 in particular, just n,m). But a back of the envelope for the overhead camera suggests that position should be repeatable on the same machine to something like 0.01 inch without fancy subpixel analysis. Which is not great, but pretty good.


Pretty sure it’s the overhead camera looking at the Glowforge Logo on top of the flying head to determine the initial X, Y location.


Yeah, I remember that being mentioned somewhere as well.

Just waiting to see how it works.



My machine has hard limits, and when it homes it goes to the hard limit first, then moves the specified distance to the soft limit (35mm on the X). When I first got the machine I put a corner piece that runs along the edge of the soft limit and attached it so that I always know where my 0,0 is and can align pieces against it.

I havent had any issue in creating parts that fit flushly with others Ive cut on my CNC machine, and can easily queue up another to be cut without worry. I just shove the piece into the corner, then cut. On a number of those I do have to manually input the offset of the cut from 0,0 to get the cut where I want it to be, but I know that it can always hit it reliably.

Numeric positioning also helps me a lot when making sure I make the most of my materials. As different materials have different kerfs, I know what the minimum distance I can place and object away from another and still achieve a clean edge while maximizing the material usage. It may sound trivial, but there have definitely been times where an extra half a milllimeter has been the difference between being able to make the part now vs having to order another piece of material and waiting for it to arrive.


@dan, I think the question has still not been clearly answered. [quote=“dan, post:68, topic:3386”]
If I ever want accuracy better than an eyeball, I do it by cutting a piece-shaped hole in material and then dropping the piece in there.
Again, where do you ‘drop’ it? Do you just prop it against the enclosure walls?

Let’s try put it another way. The GF working area is predefined at 500x300. That should not be a variable. It should have pre-defined XY limits. How do we know where this envelope is on the bed? Trial and error?

Please clarify this for us.


We’ve gone around and around on this, so I must not be explaining it very well. I apologize.

Can you explain what you’d like to do, and how you’d do it on whatever system you’re used to? That might help me explain better.

@takitus: [quote=“takitus, post:78, topic:3386”]
When I first got the machine I put a corner piece that runs along the edge of the soft limit and attached it so that I always know where my 0,0 is and can align pieces against it.
You can do this with your Glowforge. Make that corner piece a little oversized and laser-cuttable, position it in the top left, then cut a square in the top leftmost corner. It will be aligned with the laser’s top-leftmost range of motion.


I think it’s as clear as it can be. (photo credit to @jacobturner)

Shove a piece of cheap wood in that corner. Cut an “L” or, as you say, a square in that corner by dropping a design in. Default position, from what I can gather, is the top left corner. Right? Could be tape or whatever if you wanted.


Thanks @scatterbrains, that image helps a lot.
There seems be a flat bar on the bed close to the work-area limit (I assume it is permanently fixed). So now I know where to ‘shove’ and ‘drop’ my guide.

Can’t wait to start ‘shoving’ and ‘dropping’…


Thanks for responding @Dan. What was missing all the time (or I missed it) is what to position the corner piece / jig against…

The image that Scatterbrains posted helps a lot.


Everyone has different approaches to understandiing. Over the past year I purposely didn’t bother asking questions about user interfaces, jigs, S/W design approaches, etc. Why? Because once the GF is in my hands I will realize that much of my planning has been for naught. Simpler solutions will suddenly become obvious or I will have assumed too much and have to completely rethink. Measure twice cut once.


I think @rpegg has the best conception here of what to do:
Chill out, don’t pre-think.

To be honest, the use cases where a hard stop at 0,0 that you can wedge material up against and you need to have done so are minimal.

The only thing which can align perfectly at a 0,0 corner is a rectangle. Most people who put a rectangle IN are not wanting to pull a rectangle OUT. ie - you are going to cut a shape out of the middle of this thing, and ditch the corners/edges as scrap.

In all of those cases… loss of a 0,0 endstop is a trivial loss.

In all of the cases where you are working with an initial “Not a rectangle” the loss of that corner is possibly a net gain, but at least not a loss.

So… the ONE use case where you want to have a 0,0 end stop, and may suffer for lack of it, is when you have something with a perfect 90 degree corner, and you absolutely do not want to cut away any of it (or you want to cut an insanely precise portion of it).

That simply is not a use case that will come up all that often for most people. Granted, for SOME people that is a “nearly every single time” use case.

But, it is still quite possible to create your own hard 0,0 stop. And there is a bit of a lip on the honeycomb which you could use (and keep a post-it with measurements for the offset from comb-lip to actual 0,0).

The manual drag and drop of your cut onto a picture of the bed will open up a lot of possibilities. But you will only make full use of this extra power if you actually use it.

Don’t judge the intelligence of the fish by asking it to climb a tree.


I’d like to put forth another use case where 0,0 might be particularly useful. (please remember, zero laser experience here)

What about the case where you have a piece that has at least one 90 degree corner and you want to precisely place an engraving on it that is a uniform distance from all of the edges? Lining that up by “eyeballing” it would be VERY difficult to be sure.

While the GF has workarounds for this kind of use case (scan the item in the middle of the bed, replace the item with a thin cardboard or wood to create a jig, test cut the image, put the item back in when satisfied about placement, let it engrave), I think it is a valid one to consider.


I think that having a 0,0 (or an M,N, don’t care about the number) is important for anyone who wants to do things more than once. Once I’ve made the jig for some job, I’d like a way to just plunk it down in the same place again, rather than doing the optical alignment each time. (And I think that cutting a corner piece and maybe some edge pieces should do fine for that.)


It is absolutely useful. And even though there has been much mention of “Make a jig” and counter of “I don’t like doing that” there hasn’t been much talk about what a jig is.

My latest repeat cut case on my own laser needed a lot of putting the material in precisely the right spot, because I was using the full width/height of the piece. I haven’t made myself a 0,0 stop yet (I swear, it is on the to-do list!) and wouldn’t have used it anyway (my laser doesn’t fire up immediately and so won’t cut the first 5mm or so of travel).

So, I “Made a jig”

This consisted slapping down a few sheets of paper, and running my job REALLY fast so it engraved onto the paper. I then put my work piece carefully over the engrave, laid more paper on top of that, and re-ran the fast run.

I now knew that my first piece was precisely where I wanted it, so I grabbed a bunch of thin magnets and put them all around the edges of my work piece. And lo-and-behold… I had a jig perfectly designed for my material and cut plan.

The cameras would have removed the need for me to do the paper engraving initial step. I could have slapped my material wherever, lined it with magnets so I can align the next one to the exact same spot, and then positioned the cut on screen (probably would still do the paper on top with fast engrave just to be REALLY sure I had the corners right though).


I was putting together a little jig creation tutorial last night when @dan answered the question…so I went to bed. Chuckle!

I can write something up once we actually have the machine and post it for folks who aren’t as familiar with it - it’s pretty easy to create some corners for the machine if you want to use them, or just make the jig part of your initial design the first time you run one.

And after that, it’s drop in your material and press a button.

(Would like to actually test it before i write it up though…just in case i guess wrong.)


And that will work if you have a bunch of repeat cuts to do NOW. But what about doing the same task next week or a month from now? How can I be sure that my next batch will be just as accurate as this initial batch is? I feel like I would need a 3 point assurance to know that my item and cut are perfectly aligned (with the 0,0 corner stop I’m assuming there are two “legs” in the x and y directions to give me the 2nd and 3rd point assurance).

I’m sure you’ve come across the situation where you have bought a card game of some sort, and you notice that the printing isn’t perfectly even across every single playing card. That is what I fear will happen with repeat cuts. Things may drift between cuts.

I don’t know that the 0,0 corner thing is the final answer. As far as I’m concerned if I had a Glowforge certified L block thing that I could toss onto the bed and the camera would automatically recognize it and it’s orientation and align the engraving based on that, that would be absolutely doable and acceptable. That way no specific coordinates are needed, as long as your item is a consistent size, and the software is capable, the engraving will be perfectly aligned every time despite the time and number of projects that happen in between each similar piece.


Depending on the material that you use for the jig, the sides of the machine can create your second and third points (ie: use wood to cut the jig, then align it in the same upper left corner each time and make sure that the sides are flush before you drop your material or items into the holes.). It should be right on the money.

I remember bringing up 3 Point Registration a while back, (can’t find it now though), but @dan and @Tony should be aware of it…They weren’t set up to code for it back then, and they had to continue as they had originally designed these things in order to get the machines done.

It’s a complete re-write of the code to do 3-Point Registration. The logic for how the head moves would use different reference points from what they are probably using now. (Can’t know for sure of course…haven’t seen it yet. But I trust that they have something in place that will work for what we want to do with it. We can create our own, physical, 3-Point Registration system using a jig.)

They might, or might not, opt to look into re-coding for it down the road, but they can look into whether it’s feasible or not after the machines ship, since it’s mainly a software thing.

In other words, it’s in


Good point. I rarely return to prior work, so just saving my file in the first place is hit or miss. I could see this being a nuisance for people who frequently return to prior work for one-off repeats.