LaserUser.com Forum (formerly RDWorksLab)

I have been trying to diagnose circles sometimes not closing properly. It seems intermittent, but after performing some tests I can get it to repeat in some files.

I do not have issues with engraving, and squares look reasonably accurate when measure with calipers. I have been looking around the forums and found the ATK backlash test. When I run this I am getting consistent results. The same circles in the test have the same issues every time regardless of the output size or speed. Hoping someone might have some thoughts, thanks!

viewtopic.php?f=133&t=6131&p=31000&hili ... ion#p31000

Recommend performing the testing in the attached link.

Thanks Pete. I have performed those tests.

The engraving is performing well, and the cuts look good. I could probably fine tune the cuts further, but I think that would more likely solve a circle coming out as an oval as a opposed the periodic misalignment in the cuts.

First blush would say you are cutting fast enough to loose a step somewhere. But cutting is not the fast process engraving is to be in a position to loose steps.

When things don't line up in either axis, is it always the same direction it's off. When it is off, how consistent is the gap?

Dave thanks for trying to troubleshoot with me. My first thought was missed steps as well. The cuts aren't particularly fast, 15-20mm/s.

There is consistency in the lip when it does occur. In the first photo I posted the cuts in the first X we're moving left to right and starting at the 9 o'clock position. Every time I run the test it has errors on the third and fifth circle with almost identical results. This was repeatable even when I changed the speed, and when I reduced the size of the output by about 50%. Below is the file run three times, I never moved the paper and only moved the laser head between tests.
I have attempted to change the tension on the belts without any success, although I am not ruling out the possibility that tension is still incorrect. I have just been a little hesitant to change too much because I have not a good opportunity to have the machine offline for too long.

Sorry I re-ran that test again, and I realized I did not describe the path accurately.

On the first X row start/end positions and direction of travel are as follows:

Origin Top Right- CIrcles are order Left to right
1st Circle: Start 9 o'clock counter clockwise
2nd Circle: Start 3 o'clock Clockwise
3rd Circle: Start 9 o'clock Clockwise
4th Circle: Start 3 o'clock Counter clockwise
5th Circle: Start 9 o'clock clockwise

The lip is the most pronounced in these at the 3rd and 5th cuts.

Dumb question - have you adjusted the backlash settings in RDWorks? I don't mean the time delay in the power supply, I mean the actual backlash.

viewtopic.php?f=105&t=5733&p=29440

And check that the lens is tight in the tube.

When I try to compensate with the x backlash all of my circles start to become a bit misshapen with small ridges on both sides.

I also just double checked the tightness on the lens in the tube.

I am still getting the same lips.

Thanks again for troubleshooting with me.

Chris- I just noticed you have a 900x1300 table. I have had a laser that size for 6+ years and just replaced it with a 1800x1300 machine. I think we need to realize that these very large machines will have some drawbacks, heavy gantries and long cog belts for two. There is bound to be some slop somewhere. I don't know how large your test circles are but if they are small [which I suspect] you may be stepping up to the limitation of the machine. Large machine present new and funner problems I my experience. I have always had a 130W tube and have been able to do very fine work with it and have found that it will throttle back to very low output without a problem, but I don't think that large beds can be remedied. On my large machine I know that I should use the upper left quadrant for best cutting- the large bed is great for large sign blanks and we have enough power to do the cutting necessary but being able to finesse cuts over the whole bed is asking a lot.

Is your machine new[ish] and even if it is you would do well to check all the cogbelt pulleys and hardware.