Engraving and Cutting not lining up. Help Please!

[Tech_Marco]

Hi Marco,

I'm running Lasercad 5.95 and my os version is below.

Main Board version: 7.12.04.03
LCD Board version: 5.12.2.29

I am performing the upgrade to the current V7 OS update and lasercad 6.27

Thank you,

David

David:

I talked to the programmer, he said that you MUST upgrade the OS, not just the LaserCad.

Marco

[liaoh75]

Hi Marco,

I've upgraded the OS on the DSP and am now running Lasercad 6.27 as well so I'm fully up to date on the software and firmware side of the equation.

I've retested the same file and got the same results with my original settings. One thing to note though is if I bring my speed down to 50mm/sec instead of 100 or 120, things start to line up. X_swing still does not work correctly, but with x_swing unilatralism, my engraving and cutting line up correctly so at the moment there seems to be some relationship between engraving speed to how much it shifts. I hope this information can help you and the engineer track down the problem. My max accelleration for my x axis is currently set to 10,000. I have tried changing this figure to as low as 1000 but yielding the same results at the same 100mm/sec engraving speeds so the setting doesn't seem to make any difference.

I'm running lead screw with linear bearings powered by a Yaskawa 850 watt high inertia servo specifically matched for the load calculated by Yaskawa engineers to give me excellent positional accuracy up to 700mm/sec before having sync issues due to load.

Thank you,

David

[chakata]

Hi David,

I'm running a converted 40w Chinese laser (actually everything inside is now Lightobject except for original x-y stages and stepper motors and the blue box that contains everything) I have upgraded to latest firmware and LaserCad 6.27. Mainboard is 7.12.09.21 and LCD is 5.12.02.29 The quirkiness I refer to is I need to put a dummy cutbox (any speed, 0 power) around my work area otherwise my system will try to engrave everything at origin (0,0) at top left and prior to putting that box in I would get 'engrave over size' error practically anytime. In your example file , as the engrave layer was inside a cutbox which I altered (speed, power)to suit my 40w system I didnt need to apply another dummy box around the project, but the result was identical to your photo2 . What scaling did you try for photo1? If i get time I could also try that one as it appears the cut lines are at the opposite side of the diamond engrave than in the larger one.

Charlie

[liaoh75]

Hi Charlie,

Thank you for the information. I thought that maybe the delay was due to my laser tube bing 1650mm long and there were triggering and shut off delays. A 40watt glass tube shouldn't have much latency from what I've been reading. May I ask what you modified your engraving speed to? Can you try engraving at 50mm/sec and 20mm/sec and see if the shapes will line up for you as it has for me. If it does line up for you as it has for me at a lower speed, then logic will lead me to the conclusion that we have either identical errors in our configuration somewhere as our mechanical systems are completely different or the DSP is have difficulty timing the firing of the laser as speed increases.

If you could do that test for us, maybe we might start comparing our manufacture settings.

Thanks,

David

[chakata]

Hi David, In your file it imports into LaserCad as both layers engraving . I,ve altered the 'black' layer > cut. I,ve just tried your file just now setting engrave speed (blue layer) to 50mm and cut layer (black) to 100mm. Took a while for the job but looks same as before (photo2) . It appears that the engraved part (diamond) is shifted approx 2mm to the left

Charlie

[Tech_Marco]

Here is what I got. I compared the worked and the image file, I didn't see any difference. Please point me to the location where the 2mm "shift" movement happened.

Marco

[Tech_Marco]

By the way, I think there is an issue when using servo. Make sure that the pulse unit (um) to be 5 or above per programmer. If somehow you set it to 3.5 or less, there seems having a problem.

Also, sometimes the 'gap' was caused by the delay of On and Off of certain laser tube. He said that it's hard to prevent it from happening.

Marco

[liaoh75]

Hi Marco,

Thank you for testing that out for me. My pulse unit is set to 10 pulses per unit (um). I originally had it set for 1 pulse per unit (um) and when the Yaskawa engineer came to assist with tuning, he told us that the pulses were "not pretty" and that could cause problems after hooking up his computer to the drive as the dsp was driving the axis. We then set it for 10 and he old me it was acceptable. Could you ask the engineer what else I can do to make this work?

If you have been following my postings with Charlie, it seems that you guys have pretty much the same equipment yet your machine doesn't have any problems but his is exhibiting the same problem as mine and he isn't running servos. My question is, is it possible we both have the same machine setting problem and yours is set correctly.

Would you mind sharing your speed setting in terms of mm/sec and are you running x-swing or x_swing unilateralism? What is the next step for us to resolve this problem?

Thank you,

David

[liaoh75]

Hi Marco,

After magnifying the picture you posted, I noticed that your engraving also exhibits a slight shift to the left. If you look closely at the top to corners of the diamond and compare the two sides, you will note that the engraved part is not exactly centered as it is in the original file. I think mine is just more pronounced.

In problem solving mode, I think there must be some logical cause as both Charlie and I both have this problem yet our machines are completely different so I feel hopeful that we can solve this problem.

One correction on my previous post. My setting is one pulse = 10um oppose to my previous setting of 1 pulse = 1um of motion effectively reducing my resolution by a factor of 10 to get a pretty waveform from the DSP. I feel that this resolution is more than adequate so no complaints there. I'm even willing to bump it up to 1 pulse = 20 um if that solves this problem. Could you please ask the engineers if this would help.

Thank you so much for following up with me. I have included some pictures of my machine build for reference.

David

[chakata]

Hi David , I thought I'd run your file again and this time I added a scaled version (half x dimension) on the work area ,side by side. On my attached pic you'll see that both the large and smaller diamond shifted to the left , probably proportionally. {Interestingly that your small scale diamond appeared shifted to the right} I,ve taken some measurements and allowing of course for a few degrees of measurement error I,ve concluded that large diamond shifted 1.77 mm and the smaller shifted 1.15 mm. My points of reference are the apex cut box surrounding the diamond (right) and the apex of the first shape to its left , both having same y co-ords.

Charlie

[G-laserguy]

Hi David,

One correction on my previous post. My setting is one pulse = 10um oppose to my previous setting of 1 pulse = 1um of motion effectively reducing my resolution by a factor of 10 to get a pretty waveform from the DSP. I feel that this resolution is more than adequate so no complaints there. I'm even willing to bump it up to 1 pulse = 20 um if that solves this problem. Could you please ask the engineers if this would help.

David

Yes and no.
The controller is absolutely able to calculate high resolutions like 1µm/step, but:

A very high resolution will slow down the speed of your stepper or servo motors.
The AWC-608 is able to drive the "pulse-output" for X-,Y-,Z-,U-axis with 150kHz maximum Frequency.

If your resolution is 1µm/step, the pulse-frequency will be 150kHz at 150mm/s. By choosing a higher speed your machine will loose steps.

You can calculate this very easy by yourself for different resolutions:
max frequency (Hz) x resolution (m/step) = speed (m/s)
150.000Hz x 0,000001m/step = 0,15m/s = 150mm/s

After changing the resolution up to 10µm/step your max speed is now 1500mm/s. (from controllers side not your complete system!)
There can be more problems with stepper driver, motor mechanics ...for sure.


My tip to Charly: try to reduce the space_speed for your system. This could be a reason for your problem.

Good luck.



Greetings from Germany

Diemo

[liaoh75]

Hi Diemo,

Thank you for chiming in on our problem. Actually Both Charlie and I have the same problem. As a matter of fact, I was the one originally reporting problems with the shifting of the engraving. Then both Marco and Charlie were kind enough to try out my file on their machines.

The Yaskawa engineer that came to fine tune our servos evaluated the controller's pulse stream and informed me that the waveform wasn't very clean (he used the term "pretty") at 1 pulse = 1um so he advised me to change the value to 1 pulse = 10 um or higher before the pulse looked clean enough to the engineer that he felt confident there wouldn't be any missed steps. I wasn't aware that each axis had a max pulse rate of 150 kHz. Maybe that was why the pulses were not "pretty" to the engineer. Good bit of information to know.

Now on to the space_speed setting. What exactly does that setting do? I read the manual and it still isn't very clear. Can you give us an example of how that would alter the shifting of engraved images?

Thank you,

David

[G-laserguy]

Hi David,

Your Yaskawa engineer was right, the pulse-output-waveform is not perfect. The reason for that are the output optocouplers used in AWC-608. But this will become a problem only, if you use digital stepper drivers, which need a input-signal with short rising edge and defined pulse-length. Most stepper drivers "eat" any kind of signal as long, as it is a pulse.

The "space_speed" sets the speed, the laser head moves between cutting-end-position and engraving-start-position or when the head moves back to origin after finishing a job or while engraving/cutting the moving-speed from one object to the next.

If you cut out some parts from a peace of wood, the laser head will cut the first part, then "jump" to the next part ...and so on. If this "jump" is too fast, some microsteps can get lost.

Diemo

[chakata]

Hi Diemo, great to see another contributor to this particular post, I did as you suggested and reduced my space_speed from default of 500 down to 200 (this figure from a User Parameter posted elsewhere on this forum), and re-ran the job. If anything it may have shifted slightly more to the left by a few tenths of a mm (but difficult to accurately measure) using this slower speed.

Charlie

[twehr]

Hi Diemo, great to see another contributor to this particular post, I did as you suggested and reduced my space_speed from default of 500 down to 200 (this figure from a User Parameter posted elsewhere on this forum), and re-ran the job. If anything it may have shifted slightly more to the left by a few tenths of a mm (but difficult to accurately measure) using this slower speed.

Charlie

Sounds like excessive slippage and/or backlash in your x axis. Since it is more noticeable at higher speeds, that leads me more toward the slippage side.