Camera Alignment

Also opening and closing the lid between experiments.

Yes, I expect you have to do that to trigger the camera before you can print. That does bring hinge play into the mix though but I can’t see that making such big changes.

I remember reading somewhere, sometime, that it’s possible to require a reboot to re-home after some jobs have been run. I don’t know if that’s still needed or not with the current firmware/software. But it’s possible that TARDIS I ran that was way off was because it was the last thing I ran that night, and yesterday it was the first/only thing I ran.

So that theory lends itself to…

I think I will perform this experiment. (It might have to wait until the weekend though as I really won’t be home much otherwise.)

That’s also a good note… In last night’s TARDIS I did not have the room light on. Only whatever light was coming in the windows. My way off TARDIS, the lights were on. So this is also an experiment that needs a little proving out.

I think that’d be a separate experiment. But one I might also perform.

While it does trigger a bed image update, it’s not required for one. You can refresh the bed image from the GFUI.

3 Likes

Actually, in my mind, it’s the hinge and how consistently the lid closes that might introduce the single largest source of error. I’ve noticed huge registration errors with the slightest amount of material thickness mis-measurement. For example slightly warped material. I would expect even more severe errors if the lid camera were ever so slightly twisted because of any lid lean. Wouldn’t surprise me to see errors of this type introduced during shipping.

The unit is extremely well built but there is play possible in how the lid closes. I have to remember to press down on the lid to ensure it is fully seated on both front corners otherwise errors will be introduced. Consistency is important.

6 Likes

Yes I think how far the lid closes, the material thickness and warp and even how flat the table the machine is on will affect the corners but they shouldn’t make much difference in the centre. I think there are two separate problems.

The machine may be well built but for something weighing 70lb I don’t think it is very stiff. I doubt it will ever rock on a warped surface. I think it will just warp to match. How far can you lift one corner before the others lift?

Calibration drifting off in between homing is very worrying. That seems to imply the steppers lose steps, which would add errors in the prints if it happens while printing. Are the steppers permanently energised after homing or do they power off at the end of a print?

2 Likes

If the machine’s on, they’re on/engaged.

Just a thought: The head homing is done by the camera identifying the :glowforge: symbol on the head, right? That is much closer to the camera than the bed. I would suspect it is therefore a somewhat fuzzy image. I wonder how much slop there is in identifying the “home” origin at the center of the field?

1 Like

Or the camera switches to macro mode for homing. No clue.

Fisheye lenses with short focal lengths and small apertures will have infinite depth of focus starting very closely to the lens itself. No autofocus mechanics or macro modes required.

7 Likes

How 'bout that. I wish I knew more about lenses. And cameras, in general, really. I should take a class.

1 Like

It would be interesting to place a printed logo at the same height under the camera to see how it appears in the bed shot.

2 Likes

Unless I’m misunderstanding something, the centre will definitely have an error if the lid is not closed consistently every time:

  1. the machine homes and determines where the centre is (with the logo currently right under the camera)
  2. you open the lid and then close it slightly more tightly than when it homed
  3. the camera is now angled slightly more towards the back of the machine than when it was homed, and so the laser head will be operating on material slightly lower than where the design is displayed in the camera view

Depending on how you close the lid, you can introduces error in any direction - closing more tightly will result in cuts below the image displayed on screen, closing one corner more tightly will offset on an angle etc

1 Like

True but the lid is very large compared to the distance between the camera and the centre of the bed. So if you lift one corner of the lid by say 3mm it would only make the image move in the centre by about 1mm. @Tom_A is seeing errors over 6mm which would require the lid to be very obviously not closed I think.

The corners of the bed are another matter and will be very much affected by the lid to closing properly.

If the lid is not closed all the way the GFUI gives you a “Lid Open” message and you can’t run the project. There’s maybe a 2-3mm gap in the interlock “dead zone” though before it trips the sensor. However, the friction of the lid against the sides of the case will stop the lid above that point if you let gravity do the work, so it basically forces you to put pressure on the lid to ensure it is fully seated.

Here’s a little camera test. Looks like the lid camera will get a pretty clear/focused view of the GF logo during the homing/calibration routine. I may have stuck it in there nice and crooked but at least I’m not the one selling millions of these wax cans with spelling errors. :slight_smile:

10 Likes

Shinny is shiny skinny stuff :slight_smile: More efficient invented word :wink:

3 Likes

The only problem seems to be the complete white out of “following” due to it being close to the LEDs. Probably no coincidence the head is black.

1 Like

That was me shining a flashlight in there, note the shadow of the camera itself, which lends some credence to others that have had calibration issues (head banging) due to bright lighting in their workspaces directly overhead.

4 Likes

It’s killing me this week that I’m working so much. I’m going to try and run a test when I get home from work2 (which’ll be around 02:00). I’ll have to leave for work1 again around at 06:15, so I can either sleep, or 'forge. Hmmmm…

9 Likes

Oh, look… Another user with no alignment issue…

I’m getting pretty bitter now.

More like pissed off. I’m going to have to leave the forums for a while before I say something I might regret. I’ll do another test when I get home, but I’ve low expectations. Yeah… I’m gonna go now. I’m really starting to build up steam here…

2 Likes

Honestly, your experience in particular is making me very worried about my forge - I appreciate reading everything you’re posting and you have every right to be angry. I cannot imagine how frustrating this roller coaster has been for you. Is support on a line with you? Because … they should be.

1 Like