SH 350 skipping steps or software related...?

[Arjen Abbestee]

I have a Shenhui 350 Laser. I use it mostly as a hobby machine, so even though it's three years old it has not got that many hours on it.
Today I cut some 3mm (1/8inch) birch plywood but that did not go well at all: the line did not end where it began. (See attached picture, the little bump on the recess top right is intentional)).
353012
It looks like it is skipping steps, but the rest of the shape appears to be smooth. I would expect some small bumps on the curves where the skips happen, but the whole shape is smooth. Except of course where the cut begins and ends, here is a difference of 0.1 to a whole millimeter. It is not consistent, sometimes it cuts a shape just fine, the next try on the same shape it's off. And this happens in both X and Y, which is another reason why I doubt it is simply skipping steps...
I checked the tension on the belts, seems to be ok. When I slowly move the carriages by hand I cannot feel any obstruction whatsoever. I also tried cutting at different speeds, no difference.
Also this appears to have started overnight. I cut a lot of cardboard last week without this problem.
Maybe related is that as long as I have this machine, the x-axis is almost silent, whereas the y-axis motor always makes some noise. Now when I cut a straight horizontal line I hear the sound that is related to a slow y-axis movement. Moving the head with the controls on the machine itself still sounds as it always has.
Does anybody have any tips how to determine where this fault is coming from? Can it be a software fault?

[Doug Fisher]

Have you checked that all of the set screws (some people call them grub screws) for all of the pulleys on the X and Y axis drive shafts are tight and have not loosened? On many machines, there are two set screws to check per pulley.

[Arjen Abbestee]

Hi Doug,

If you mean the two screws you use to set the belt tension (English is not my native language, so I have to double check if I understand you), then yes. I checked the belt tension and in the process re-tightened the screws. I also checked for play in the rails, but could not find any.
One more thing to clarify, the item in the lower right in the picture is a 200mm long, straight piece with rounded edges and a small kerf left and right. When it started cutting it went to the left, and the left part has the right dimensions. On the right it is still OK (where the kerf is), but the the line goes 'up' to end approx. 1mm above where it should end. It is still a straight line though, not curved or wobbly, only slightly at an angle. As if I rotated the piece whilst cutting. Of course the piece was perfectly still. I even taped it to the honeycomb to be sure.

[Jerome Stanek]

No the screws that hold the gear on

[Doug Fisher]

Look for the Y axis stepper motor. I should use a belt to drive/turn a long shaft that spans the width of your machine. For many machines, this shaft runs across the back of the machine, right in front of the glass CO2 tube. Each end of that shaft should have a pulley with gears that moves the belts for moving the Y axis. Look at each pulley. There should be two set screws on each pulley that hold the pulley tight to the shaft. The set screw most likely will require a six-sided "hex key"/"allen wrench" that is inserted into the set screw to tighten it. Look all over the system at both the X and Y axis to find many of these set screws and then make sure they are all very tight. It is very common for them to work loose which then decreases the accuracy of your machine.

[Arjen Abbestee]

Ah OK, I'll check that as soon as I'm back at the shop. I did check the belt-tension screws, but not the ones that hold the pulleys to the motor shafts. First thing tomorrow morning (it's 7 PM here now)
Just to know if I'm not mistaken, if you look at the picture, it seems clear to me that the deviation is in both the X and the Y axis (item on the left: mostly X-axis, top right: both axes, bottom right: Y-axis). Or could this be caused by just one axis being faulty?

[John Noell]

If you run that file again, does it make the same mistake in the same place? I have seen where a file did not load properly and gave similar mistakes. If you re-run it from the laser (without re-loading) it gives the exact same mistake(s). If you reload it, it usually runs okay. The cause in such a case may be a bad cable.

[Arjen Abbestee]

Hi John,

Good thinking. I'll try that too (after I checked each and every screw...).

[Arjen Abbestee]

All screws appear to be tight. So I ran the test John proposed. Ran a complicated file from the computer, none of the larger shapes closed, smaller shapes don't appear to be affected. Also it seems the fault becomes worse 'later' in the file. So the first shape has a smaller deviation than the shape that is cut last. I ran this same file from the machine itself twice. Deviations are almost identical, differences may be caused by the fact I'm cutting thick paper at plywood settings to not waste any more wood. I then ran the same file from the computer again. Still deviations, but different from the previous ones. I then changed the USB cable. I even plugged it straight into the controller to eliminate a faulty connector. Still deviations...
So I guess the next step is to re-install the software (latest version to be found at http://en.rd-acs.com/ ?).
And maybe a firmware update on my controller? Or would you consider that overkill or even dangerous? It now says 'RDLC-V5.00.30'. Don't even know if there is a newer version...

[Doug Fisher]

When you say you are using plywood settings, you do readjust for the difference in thickness, right?

If you cut a 300 x 300 mm perfect square out of paper (tape together multiple sheets of paper to make one sheet if necessary) and very carefully measure it, does your machine cut exactly 300 x 300 in both the X and Y direction? If not, you need to input the correction factor via the vendor settings in RDWorks where you type in what the cut was supposed to be and then what it actually cut, then the software will automatically set the correction factor. Then run the test again, remeasure and refine the correction factor again if necessary. Make sure to get your "nozzle to material" distance height correct for the test and use only enough power necessary to cut the paper.

[Arjen Abbestee]

Plywood settings mean that I did not adjust anything in speed and power. I did not want to change too many parameters, only to find out the machine works well at 25 mm/s and gives me trouble at 6 mm/s. So I kept a close watch with one hand on the power switch and the fire extinguisher in the other ;).
Anyway, I uninstalled the RDCam software (V8.01.10), and re-installed (V8.01.18). Even though the 'USB driver install' said 'installed successfully' there was no communication between computer and machine. So I went back to the V5 software that I got with the machine. This time it worked, so I decided to cut the same file with the old software. No deviation to speak of. Re-installed V8, this time the USB driver install took a lot longer and I also had communication with the machine. I did not have time to cut the same file so I just cut a few small circles (in scrap plywood this time). It's easy to see and/or feel where the start/stop point in each circle is. I'll try the complex shapes tomorrow, but it seems as if the latest software is causing me problems. Which is weird of course, because I did not have any of these problems until two days ago.
I think the next step is to re-program or update the firmware of the controller, but on the manufacturers site is no firmware to found (they do mention however that it is easy for the end-user to upgrade the firmware :confused:).
Concerning the 30x30cm square, that test is part of my clean/re-adjust/re-calibrate routine I do every once in a while. So far I only had to adjust the Y-axis once, by a very small margin. Nothing like the errors I see now.

[Scott Marquez]

Do you think you are running out of machine memory.
Try clearing all of your files from the machine memory and just load the one you want to cut.
It might be worth trying before making software changes.
Scott

[Arjen Abbestee]

Will try that tomorrow, but I had a look at my 'saved files' this morning and there were only 12 files, all very simple tests from when I just got the machine in 2013. But I might as wel throw them away... Thanks,

A

[Arjen Abbestee]

Cleared the memory, does not seem to make any difference. I'm now back to RD-Works V7. This seems to make the cleanest cuts. But I will probably be reminded soon why I switched to V8. Cross that bridge...
Anyway, it was clearly a software fault. I need to take the whole machine apart and clean/grease all moving parts, but I'll save that for later.

Thanks for helping,

A

[John Noell]

FWLIW, when I tried to use USB on V8.01.18 it simply didn't work, but the Ethernet on my new Longtai is fine with that version. (And Ethernet is MUCH faster.) I was seeing some strange intermittent mistakes on my Shenhui and have finally concluded that it probably is a bad controller.

[Arjen Abbestee]

If only my controller had an ethernet connection. But I have to make do with USB...

[Arjen Abbestee]

And the problem is back...
It seemed to be gone after my software downgrade, but today I notice a difference between two items that should be identical. So I did the test again running the file from the machine. Where the last time it seemed the mistakes were identical, now there are major differences with each run. So it is either something mechanical after all, or the controller and/or stepper drivers are on a rampage. But if the steppers are off it seems to me I would have to see hickups along the path where the steppers misses steps. But the paths are all smooth, just not in the place they should be.
I did some speed tests, it appears the faster I cut, the better the result. At 100 mm/sec I don't see any mistakes, at 4 mm/sec it's off by millimeters. Shouldn't a mechanical problem manifest itself more at higher speeds?
In the pictures the lower left corner is the same for both layers. In the close up, the shorter one is wrong. But in that one the total hight is the same. So it's more like it added steps than missing them.
The pictures are from different cuts. As you can see, one has a large mistake in the x-axes, the other in the y-axis.

I'm lost...

353552353553

[Kev Williams]

I wonder if you have a bad wire like I did-

The stepper motors on my Triumph come with a wired plug, and the wires are about 3' long, which is about 3' too short to run thru the drag chain and down into electric hell where the other ends connect to the stepper drivers. I'm assuming many machines are the same way, and the fix is to simply butt splice additional wiring on. The problem, at least on my machine, is the butt splice was in the middle of the drag chain, where it looped up at about 17" down the Y axis. Wire bends easy enough, but the soldered ends don't...

Here's where the splice is in my Y drag chain...
353573


-here's where the bend got really tight-
.353571

This is the broken ground wire--
353572

and after my repair..
353570

--my machine started out like yours, a bobble here and there, although mine was usually during engraving when moving from one spot to another. Most of the time the machine was fine, but during troubleshooting I finally realized the problem always happened around the 17"-down area. The wire would seem to connect okay until it hit that tight bend. Anyway, it's been perfect since I fixed that wire!

You might do some testing, you can just use paper, fill the table with paper and run at the lowest power you can and mark (or cut) the paper. That shape in your last picture, copy that into a matrix of 15 across and 10 down (or whatever fits), then run them all, first at the speed where it messes up the most, then faster, etc... if the errors occur, check if they're confined to one area of the table- my problems occurred whenever the Y axis got more than 17" down, but stopped messing up at about 20" down, and would happen ANYWHERE across the 51" of X axis. are That's when I thought 'bad wire'. I then found that if I grabbed the Y drag chain and shook it while the machine was working, the stepper would go nuts! :) But mine was real bad by then.

This may or may not be your problem, but if you have spliced wires in your chain, chances are good...

[Arjen Abbestee]

My machine is not that big (50x30 cm work area) so I don't expect spliced cables. But of course it cannot hurt to check. Your symptoms are different though. In my case the fault happens all over the bed, not only in a specific area. I'll try wiggling the wires during operation to see if this will invoke mistakes.

Thanks,

A

[Arjen Abbestee]

Checked the wires, no problems there. To rule everything out I took the whole machine apart (except for the XY stage itself). That was the only way to reach the gears on the Y-axis. The gears turned out to have no set screw at all, they are simply clamped to the motor axis. But no play anywhere.
While I had the XY-stage out of the machine, I took out the belts and put them back in reverse. So the part of the belt that used to run over the smooth wheel now runs over the toothed pulley.
Put everything back together again. Took some time to re-align all mirrors etc, and than it was time for the test. So far so good. No errors and as a bonus it cuts better than before. Fingers crossed, hopefully the belts were worn and caused the errors. New ones are ordered.
Does not explain why changing the software seemed to solve the problem at first, but there may be some logic as to why the error was more severe at lower speeds.
Anyway, if this caused the problem, its solved. If not, I'll be back...

Thanks to all,

A