Angled cut edges

[paul3020]

Hi,

I have a LS3020 and am trying to compensate for the kerf of a cut. However the kerf was quite inconsistent some cuts had square(ish) sides and some had angled sides. I thought the kerf might vary with X and/or Y position, so I created a test shape which is a sawtooh shape with square teeth (see overview picture). I cut this test shape at each edge of the machine (Xmin, Xmax, Ymin , Ymax).

My test results are very strange, the teeth in the shape alternate between square and angled and this happens along for EVERY side of the machine. This indicates the issue effects the X travel and the Y travel and I think its todo with which direction the laser head is driving while its cutting (I'll watch the cutting head tomorrow to look for a pattern).

Has anyone seen this before or any ideas what might be out of alignment?


Overview: These are test calibration shapes, 2 were cut with X=min and X=max and the other with Y=min

Square side1: Engineers square against an edge to make the angle obvious

Square side2: Engineers square against an edge to make the angle visable


The images links dont seam to embed well (google drive) so here are the links
overview: https://drive.google.com/file/d/0B0utqY ... sp=sharing
square1: https://drive.google.com/file/d/0B0utqY ... sp=sharing
square2: https://drive.google.com/file/d/0B0utqY ... sp=sharing


Thanks
Paul

[Spooky]

Looking at the amount of edge rolling I'd say your focal point is adjusted a little too low, try lowering the workpiece and see if that helps.

The pics suggest the focal point is nearer the bottom of the material than it should be.

best wishes

Dave

[paul3020]

Hi Dave,

I tried lowering the bed this morning in 0.5mm steps from 0.5 to 3mm and found the behavior remained the same. I think the cutting power dropped a bit as the corners didnt quite cut (cuting a 20mm square) so I have probably gone past the optimal focal point.

I have found that the direction seams be related to this somehow. In the photo directions 1 and 4 give angled cuts with 2 and 3 giving square (ish) cuts. This was repeatable for 5 squares in a vertical line (spanning 150mm).
https://drive.google.com/file/d/0B0utqY ... sp=sharing

I guess it could be the caridge twisting or something but it doesnt seam very likely and I am not sure exactly how to test for this.

[paul3020]

UPDATE:

I have experimented with the drive direction and have found that the cuts are consistent with direction. (cutting a 20mm test square).

As you look at the honeycomb bed from above, cuts going left and down are square ish, cuts going right and up are angled. See pictures below
left : https://drive.google.com/open?id=0B0utq ... authuser=0
up : https://drive.google.com/open?id=0B0utq ... authuser=0
right : https://drive.google.com/open?id=0B0utq ... authuser=0
down : https://drive.google.com/open?id=0B0utq ... authuser=0

Now this is where it gets extremely strange. If I set NewlyDraw engrave settings to mirror the edges that cut badly flip!!!! I can watch the machine cutting in the same direction with mirror mode enabled and disabled and the cut will come out with a different angle!!!!!

How can NewlyDraw be effecting the angle of the cuts? This is crazy annoying to try and debug, has anyone seen this sort of thing before?

Also I am cutting 5mm acrylic at 10mm/s at 55% power, do these settings seam reasonable? too slow maybe?

Paul

[Daven]

Long time since I had the 3020 but 5mm did have a taper - try flipping over the lens and see how that affects the cut. Cut a circle as well and see if it tapers more on part of it

Cheers
Dave

[PhillyDee]

I would be tempted to see if there is play on the laser head / carriage.

[paul3020]

A quick update:

I have tested cutting out circles instead. This shows nicely that the issues is worst on the top edge, and bad on the left side edge of the circle. The right and bottom edges of the circle are ok showing minimal issues. This does show that the most extreme angle is right at the top of the circle, this makes sense as this is a 2 axis machine.

I have had a look around the machine for play, the laser head assembly feels fairly solid. But the the y axis belts on the left edge feel lose, when I press them I can see the main bar move. This is my best lead as to the cause.

At the moment I am having trouble getting to the belt. Trying to get the top off is tricky with its hidden screws which are tightly tucked away. I am trying to leave the machine intact as possible to save on recalibrating and introducing more issues with dissassembly and reassembly.


Is there a belt tensioner for the yaxis?

Does someone have some photos of the LS3020 bed open? it would be useful for knowing screw hole locations etc.

Cheers
Paul

[paul3020]

update:

I found a belt tensioner for the yaxis with aid of torch and webcam. This top left of the bed, so I'm hoping this will fix the lose belt and the issue. Unfortunately experimentation will need to wait till I get back later this evening.

Paul

[paul3020]

update:

Tightening the y-axis belt didnt solve the issues.

There dosent seam to be much play, but I can get a little play by twisting the X-axis bar a bit, sort or make it clunk between two positions. I couldnt trace where it was coming from.

I have also noticed that the left hand side of the X-axis bar, the rollers are not all in contact. So the left hand end of the bar can be pulled up and down about 1-2mm (estimate). I couldnt see a way to adjust the bearings and rollers on this end to clamp them down onto the rail.

I am running out of ideas a little bit here :-s.

[Daven]

Got me stumped.. If the taper was out the same amount but in an opposite direction on the bottom of the cut outs it would make more sense and could mean the table was slanted or something just as simple.

Only other thing I can think of (other than getting it to HPC) is to shim the mirror in the head to see if that helps - use a small dot of thin card or even paper at the six o clock position to start with.
Best
Dave