10W Laser; Is this why people say that camera positioning doesn't work?

Snapmaker 2.0 Product Feedback
, ,
1

I am beginning to think the 10W was rushed so they could move on to the 20/40W. I think any attempt at getting it right had been put to the bottom of the priority list.
As mentioned many times, the 10W camera resolution is awful, much worse than the 1.6W. It’s just the same camera, used higher up, and digitally zoomed.
And as also mentioned elsewhere, what is the point of the undocumented image quality setting??

But, I realise now also that the 10W (maybe Luban) has an error in that the positioning X,Y used by Luban to create the Gcode is based on Machine X,Y (i.e… the bottom left corner of the bed) but the machine attempts to process the Gcode with reference to the Work Origin, which is by default, in laser mode, the centre of the bed.
So, within Luban, you obtain a reasonable background photo of the bed with your piece of material. You position a simple shape from Luban’s tool-bar onto the image of the material, process a tool-path, and view it in the preview. You can see from there its position relative to the grid display that your design is in the upper right quarter, say X200,Y200 from the machine’s origin. this is confirmed when you go to ‘Workspace’. You then send the file to the machine.
On the machine you see that the work origin is 0,0 at roughly the centre of the bed. When you run the file or just the boundary you will see the machine tries to add another 200,200 and so goes to 400,400.
There are many work-rounds of course. One is to run the machine to bed 0,0 and then hit Set Work Origin, which means you’re basically back to manual positioning and the advantage of the camera is lost. It also will run directly from there to the work-piece hitting any clamps you have.
It’s no good glossing over these work-arounds though, it shouldn’t be necessary.
How difficult is it to fix!

2

I have a right angle guide at 0,0 and position my work material against that. Works perfectly, and 100% repeatable. I never use the camera.

3

Actually I have found that if you start the job directly from LUBAN, the positioning of the work item on the background is quite accurate. Its when you send the job to the machine and start it from there that I need to fiddle with setting origin - every single time.

Side note: i wish they would come up with a COMPREHENSIVE and EASY TO UNDERSTAND GUIDE to how to use work origin and how to solve these issues, as I have had my machine for 3 years + now and still cannot understand how it works.

4

Because the camera calibration isn’t 100% reproducible. You have to drag those corners to form the square, and everyone will drag it slightly differently. Hence, it’s not even possible to make a 100% accurate. I think most people have made their own workflow that works for them.

Personally, I’d rather the camera gave me a live view of the job in progress.

5

The issue is bigger than just the dragging of the corners of the square; it’s about the difference between bed origin and work origin.

6

As usual, you don’t get it (and just post for the sake of posting). Origin issues are just software and easily fixed (if they cared to). The human factor can’t be programmed out when doing the calibration procedure.

Honestly, I don’t even seen the need for the camera calibration procedure. There’s no reason it can’t be set at the factory and never need to be touched after that. A red dot pointer would go a long was as well (virtually every laser cutter has that as standard). Sorry, it already has a red dot, lest dual purpose it for work alignment as well.

7

Exactly, and that’s the point of this topic.