If you are familiar with rotary tables, please take a look at the video - Looking for advice and guidance.
Thank you!
Looks good…
There are two areas that case issues… steps/rotation and the drive ratio…
What is your steps/rotation value and where/how did you derive that?
I ordered the table from the manufacturer of the fiber laser, he informed me 25600 steps.
Didn’t get any information about gear ratios, somebody mentioned that there was parameters for that within eZcad.
See if this thread helps… it’s just a rotary and the setup is the same.
With the setup in your video, you should be able to use the gui test button to rotate one complete turn and back. That must work for the rotary to perform correctly…
You can compute the ratio by counting teeth between the two pulleys.
Let us know…
My fiber laser came from Cloud Ray and they didn’t mention how the motor driver was set. I asked them on-line and they stated it was 12,800 steps/rotation. I’ve been using that and it works fine. I don’t have an ratio to apply, like for my PiBurn.
Worst case is to find what the motor driver is set, then enter that as steps/rotation. The motor itself should rotate one turn and back with the test button in the gui… That will tell you if you have the right setting for the motor driver…
If your driver is set to 12,800 and your ratio is 2:1 then the 25,600 should be the correct value for the steps/rotation in the Lightburn gui.
Generally you can fudge the ratio, the rest of the system is digital.
In the end, I think you did what you needed to do by increasing the steps/rotation using such a large table.
Try the test button in the gui… and ensure the rotary does one complete rotation and back.
Make sense?
As a diagnostic step, I think it would be wise to take the jig out of the equation temporarily. I know you double checked the distance between the slots, but if that jig isn’t quite circular or the circular array of slots isn’t quite centered, that could be the culprit.
I’ll give it a shot , mark some tap on the bare table.
Here is the LB file for the jig in use, i’ve checked it a few times - but maybe a fresh set of eyes wouldnt mind looking over it.
Thanks - if you’re seeing the same behavior on just the table, there goes my theory. The jig file looks good to my eye too. I’ve asked internally for additional input - I’ve never worked with one of these personally, so hopefully someone will come along and point something out that will make us both say ‘doh!’
Set it to the factories suggestion and use the test button in the rotary setup … some of the controls you are using I’m not sure of their exact operation.
The test button I am sure of.
are you sure that the centre of rotation af the jig is exactly aligned with the Y axis of the laser? if it is not you will observe that a full rotation is bringing you back to the exact point where you started, but a partial rotation will be misaligned
No, other than just trying to eyeball it close to the center axis
How would I go about making sure the center rotation is in line with the center of the laser?
i guess you should center the laser on the first pocket as you did, then rotate just 180° and check alignment. Or laser mark a line on your machine table along Y and use it to make sure the jig is in place
I suspect if you cut this on a machine, it’s probably centered.
The larger the diameter of the table the more off a little bit will be if not right on the correct steps/rotation value.
Of course my comment only applies if the correct steps per rotation (including the ratio of pinion and pulley) are used. that is the first things to check, but has been already covered. I may add that only certain ratios are compatible with the common stepper motors that has 200 steps/rotation, but I assume that if you bought the indexing table this has been calculated. Different story if you just made up the ratio by yourself without taking in account that the ratio must be compatible with both a 360° rotation and the number of steps you are using for a full rotation… they must be disivible by the same number. (I learned this the hard way while building my 4th axis for the CNC machine, and while designing the gear trains for a few clocks)
Following along, as I am rounding up parts to build one of these. I have another combo rotary belt drive 6:1, and it has zero backlash. So something fishy about having to add that many extra steps.
One possibility is the acceleration is set too fast, so skipping microsteps. Second possibility is the driver in the laser is not putting out enough amps for that stepper motor and mass of the table causing lost steps.
Suggestion: Turn your acceleration down until motor grumbles, then adjust up till it just smooths out. then run the 25600 setting. Turn the max speed down same procedure. Open up your case and look at the dip setting on the driver, and compare amps to the stepper motor.
If settings are 6400 which I assume, and amps are much lower then the motor size, you could change dips to 3200, and use 12800 steps as a band-aid. Or put a larger driver in, which is pretty easy.
Right now you have 2560 steps between ea hole in your template, (I think you said 36 degrees?) don’t think you have to worry about the math.
Also notice it almost catches up about the second to the last hole. Like to explain that but ?
Could be something else completely different, I only have 3 stepper projects under my belt, just what I would do to troubleshoot. Good luck.
Good thought, but I think if it had lost steps, it would not return the original point, like it does.
1 Like
So something I’m missing in the first video is does it return to zero after test 360 at 28600 or 28850? you cut away before completion both times.
Seems like one or the other can’t be true. I’m sure you checked this already but the simplest explanation is one of the pulleys are slipping back and fourth on the shaft causing similar repeatable backlash.
By the way, the video is a great way to demonstrate issue. you would have spent days trying to explain with the keyboard alone.
Just a wild thought, but what you are describing in the Non CNC world is backlash, were the table doesnt move until some slack in the system is taken up common in worn nuts or drive screws. One way to test is to do a second 360 after the first, in the same direction. By not changing direction the lash is still taken up. If it goes full 360 then you have lash in the system somewhere, possibly a loose set screw on a pully.
1 Like
Loose set screws on the gears causing excessive backlash.