This is all digital, so there is no ‘fudging’ around until after you leave the ‘digital’ arena, which ends at the stepper motor. If it’s supposed to be 5000 steps/rotation, it is or it’s broken is the thought… Once correctly set you should never have to change it.
This should separate the problem are.
Did you read the switches on the rotaries driver correctly? Are you sure it’s set to 5000 steps/rotation?
I only ask because it’s easy to get the on/off directions opposite… usually it’s marked on the switch, it’s always been towards the board for ‘on’ on all the ones I’ve looked at. Inverted would be 3200 steps/rotation… Just double check, you probably have it right.
You can check that steps/rotation are correct by using the test button with the steps/rotation set as the drivers are set. In other words set it for the 5000 steps/rotation and see if it rotates the motor like it’s supposed to. Don’t use a ratio, just steps/rotation.
It should rotate one complete turn and back. If it doesn’t, then the steps/rotation are incorrect or you have a hardware problem. Probably steps/rotation…
If it does the one rotation and back, then that leaves your computed ‘ratio’ between the motor and the ‘wheel’…
Don’t forget to ‘reapply’ the steps/rotation * ratio…
Where did you get the ‘other’ math, such as the ratio… did you physically measure the pulleys? The best origin of ratio is from whomever manufactured it.
Is the motor on the rotary the same as the machine…? Do you have a link to the rotary… always nice to ‘see’ the device…
I took a picture of the dip switches they are off, off,on, off for 5-8 on Y stepper. Per the sticker that should be 5000. I saw your earlier answer and that is why I multiplied by 1.5. This is 24.6/16.4 the two pulley diameters I measured with calipers.
I don’t have a link to the rotary but it is the same one OZ replied to in the Forum. Not sure if motor on machine is the same since I can’t see markings the way it is mounted. THe rotary is a Smooth FH160424 Type 23HD450Y-31A1. When I go to the SmoothMotor website, I can’t find that product. UGH… this is frustrating…
Do not understand what you mean by "don’t forget to reapply the steps /rotation/ratio; I thought you said to NOT use the ratio???
Thanks
Also, I ge a slight “shock” when I lean on cabinet with damp forearm and no shoes. I have tried all grounding steps and still doing it. Also, I lost the USB about a month ago, you think I have a hardware issue??
This was only to determine if the steps/rotation of the motor is correct. Remember the software has to step properly to rotate your drive wheel one rotation and that takes the ratio to be correct.
I was trying to take you down a diagnostic path to identify if the issue is pre or post stepper motor.
Setting the steps/rotation on the stepper motor driver to the steps/rotation in Lightburn, it should rotate the ‘motor’ shaft 1 rotation and back.
If this works, you have the steps/rotation correct. This would indicate the ratio/wheel diameter you’re applying is not correct.
You will have to apply the proper ratio to Lightburn for it to finally work. I’m just checking to see if the motor is doing 1 rotation.
Make sense?
You getting a ‘buzz’ is definitely not right and you should chase down the problem… Something isn’t grounded and you are it’s return path. This can cause issues, not only for the machine, but for you…
When I set the steps to 5000 it does not rotate 1 rev, it stops short. If I bump to 7500, the 5000 times 1.5 gear ratio I calculated by measuring both drive gears with caliper(24.6/16.4) it rotates almost to mark. If I change it to 8000, the test rotates to mark and then returns to Mark.
I was measuring to diameter of the top of the gears, if i try to measure diameter at belt it is more like 20.7 and 12.8. I do not want to disassemble. This gives a ratio of 1.642857143, which is 1.6 when rounded down and 5000 x 1.6 = 8000.
If I do a burn test with a height of 100 mm on a piece of tape, should I expect the rectangle burnt on the tape to be 100mm or something shorter?
I have tried to isolate buzz with help of Russ Sadler and still an issue.
The rotary attachment I have is the same one Oz responded to in February 2020 post. Pictures of it about in middle of the thread.
The idea of setting it up like a chuck rotary is to see if the 5000 steps/rotation is correct for the stepper motor to rotate one turn and back with the test button.
If that works then the problem would have to be the computations with the gear ratio. Once you add the pulleys and belts, you have to compute in the gear ratio… Lightburn really needs to know how many steps to rotate the driving ‘wheel’ one rotation. With that and the wheel diameter it can compute the ‘surface speed’, I’ll elaborate on that in a moment.
From your description, it sounds like you isolated it as the gear ratio value isn’t quite right. You really have to measure the channel which the belt lies and it usually requires disassembly…
I think you’ve figured it out.
It should be the same as if you are using your machine normally. If you cut a rectangle on the machine without a rotary, would you want it to be what you drew or something shorter…?
In machining there is ‘speeds/feeds’, it applies but is a little misleading since it was here long before lasers. Speed refers to how fast the tool is rotating and ‘feed’ applies to moving of material relative to the tool. With lasers, speed is ‘power’ and feed is how fast you are moving.
There is a concept call ‘surface speed’ (also ‘feed’ rate) … in all machining, additive or subtractive. It refers to how fast the ‘tool’ moves past the material. It can be the head moving across it or the material being moved under the head or a combination of the two.
The rotary is supposed to act just like a table, but it moves the material around instead of the gantry. It doesn’t know that you’re using a rotary for all practical purposes. So the ‘feed’ ends up being the same for either the surface of the rotary object or it laying on the table… If feeds are the same, when it moves over the surface normally 1mm, the rotaries object surface should also move 1mm…
So … the 100mm rectangle should be 100mm on the rotary…
I have no clue how to even find that post… drop the link on the reply or highlight some text and ‘embed’ the link with the link icon in the tool bar… I think you found the problem, so it’s probably no longer necessary.
All of the Chinese machines I’ve seen in the US have a grounded cord. It has ‘hot’, common and ground.
Generally the ground is wired to the frame of your machine and ANY potential for a path to ground should go via this ground line. That’s the whole idea of a grounded case…
I can’t see how could it be anything other than the ground return isn’t working properly. If it was, you could not get any ‘buzz’… There is still an issue somewhere and correctly fixing the ground may point out other issues…
When they make these, the holes are drilled before they are painted. They generally do nothing to ensure a good connection when they bolt a wire onto the cabinet for grounding purposes. I took mine apart and removed the paint where it grounds… might check those areas.
I’d try as ohm meter, might give you a clue measuring from ground on the chassis to the common ground at the end of the plug.
I missed your point of testing the 5000 revs as a chuck style rotary, was doing it always as a roller. I think I have the rotary behaving as it should. So if I get a 100 mm high box resulting in a Y height of something different, I should calibrate the Y-axis accordingly in Machine Settings to get it perfect?
As far as buzz is concerned, I cleaned-up through hole where the ground lug for power cable is attached, are there others I should look for? Also, I think I have isolated the issue to the HV power supply. Once I disconnected both Main Power feed wires and the ground to it, the buzz was gone. Now I just have to find the replacement HV power supply.
It sounds like the ratio of not right. You will have to come up with proper values in the steps/rotation for it to work properly…
You know the steps/rotation is correct, I think, so it has to be the ratio. You will have to fiddle with them to get the drive wheel to rotate one complete turn and back…
IMHO … The machines case has voltage on it, that’s why you are getting that ‘feeling’. The current is flowing through you to the ‘earth ground’. If the machine were properly grounded, then it would flow through the path of least inductance, which would be the ground wire, not you.
Jack,
Sorry for Thanking Russ in earlier reply, I mean Jack.
Are you using the Linux flavor of LB? I am still having difficulty getting my rotary cuts to match the Y length of the box I am cutting. I tried to enable Rotary Setup and then retrieve machine settings and then scale Y axis but it seemed to have no effect. So I disabled Rotary and then tried to cut the same box on the curved object and then scaled to 100. When I was sure that worked perfectly, I assumed my Y step size was correct for the Y stepper. I then re-enable Rotary and do another test and the 100 mm long box is now only 92.7 or so. I then disable Rotary and retrieve Machine Settings and try to Calibrate Y again. My assumption is Step size needs to be bigger to move 92.7 out to 100mm. But to my amazement it is smaller! See picture below:
As far as I know steps size is hardware and cannot be changed, except via a hardware change.
I think Lightburn is modifying what it sends to accomplish the task… so you may have the ‘step length’ ratio inverted…
I’m not sure why this is off and we may have to call in some ‘bigger guns’ with this as it appears to be an issue with the configuration of the device and not the soft/hardware itself.
Maybe we can drag someone from @LightBurn to help with the issue…
Jack’s spelling may invoke the Owner/lead Dev. Several staff read and provide support and report/discuss problems day and night. I’ll go grab the docs to see what’s going on with the Ruida. Maybe I can help.
@Rick and I had an involved conversation about ‘Better’.
I my mind if we get a handful of good walk-through threads on the forum and get them well documented that makes it better for everyone. It’s especially good for the folks that are on the code if we can unburden them here by solving it.
Yes, Oz probably knows exactly what’s going on but he’s probably explained it 6 times elsewhere.
I looked at my Rotary and it’s a three phase delta wound motor. 5000 steps per rev makes considerably more sense now.
Is the 100mm (tall) of tape wrapped around the circumference of an object on the roller?
Counting cog teeth or gaps on each pulley may be a more accurate approach. What you did should be close but the tooth count would be definitive.
A Machine Component Design instructor once told me, “Mark a tooth and say ‘Zero’, and start counting there. When you get back to the marked tooth count that number and that is the number of teeth or gaps on the gear.” It’s a quirky method but it seems to generate reasonably consistent gear tooth-counts for most folks.
My concern with this approach is that you may be able to adjust the y-axis calibration and then compound it with the software-switched rotary calibration setting. I would calibrate the X & Y axis on the laser engraver and then Enable rotary then calibrate for the rotary until the axis is correct.
There are other sources for error like acceleration, lost motion (slipping) and measurement challenges.
This is the secret-sauce of the calibration - Take a number like 100mm, (11 or 12 thousand-ish steps calculated by LightBurn) - Measure it - apply the measurement - correct the step length. DPI and line interval spacing (your laser dot size) are common engraver settings but the dot and the engraving settings are an order (or two) of magnitude larger than the 8-9 micro-meter (3-3.5 Ten-thousandths of an inch) step length.
Since the steps are fixed, telling LightBurn that the steps are smaller than previously expected will invite it to throw more steps at the rotary next time to get where it wants to go.
I have a couple of comments, first I started with the documentation you referenced and when I do so, the Test button works; i.e. does one rotation and then returns back to the location I have marked on the roller attachment. However, when I do a cut of a piece of tape on a can of diameter 89mm with rotary enabled and then remove the tape from the can, lay flat on desktop and measure with calipers it is only 92.6mm long. Hence my confusion on what was going on when I try to calibrate Y-axis.
To disassemble the rotary attachment to count the cogs on the gears is a time-consuming process and although maybe the “Ideal Solution”, it is NOT very practical, so suggesting it is NOT very helpful.
Your suggestion to first do the calibration on x-y on engraver and then calibrate with Rotary enabled is exactly what I have done no fewer than 15 times. Problem is this assumes the stepper motor on the engraver is identical to the one on the Rotary. I cannot verify this w/o totally removing Y-axis from my machine, so again a not very practical solution. In addition, when I do try to adjust the Y axis only with Rotary enabled, it changes nothing in the height of the cut;. I have 7 attempts this afternoon and all result in LB decreasing the step size each time but NOT altering the cut length.
I am aware of acceleration and slipping but have tried to mitigate using a speed of 10 and putting rubber bands on the can placed on the rollers.
Seems decreasing the step size without altering the number of steps per revolution in the rotary setup would lead to a shorter travel distance. I have never seen the setup alter the steps displayed to the user after doing the calibration, so where is this number saved?. What you say may be correct but is NOT intuitively obvious, especially if nothing the user sees let’s one know more steps are being thrown when the step length is decreased.
Final comment, I have over 50 years experience in IT, mostly as a programmer and system admin and manager; so I resent “implications” that I need to read documentation or disassemble hardware to do something that should be fairly straightforward. Also, if you guys “monitor” the Forum, why did it takes Jack’s prodding to get you guys involved and if OZ has explained this numerous times why is it not adequately documented and programmed to reflect what the calibration code is doing under the covers? Before entering my Post, in addition to following documentation, I searched and read all posts on Forum I could find by searching for “rotary setup”.
Sorry for being “preachy” but your comments ignited my fuse after spending numerous hours trying to resolve this myself and with Jack’s gracious help. I just wish you would have chimed in earlier rather than responding like I was an idiot.
When in rotary mode, you don’t calibrate the machine axis, you change the rotary steps.
The field in the docs shown as “mm per rotation” will show as “steps per rotation” on a DSP controller, and that is the field that you should adjust, not Y axis steps for flat movement. If you change that, it won’t affect the system when in rotary mode, and will mess up the scale of flat work.
When using the Test button, you want the rollers to make one full rotation, not the object. (From the docs: "Click the Test button to verify that the roller or chuck makes one full rotation, pauses, then returns)
So being NOT sure of current Machine Settings due to all my fiddling around. How should I “reset”?
Should I first do a test flat burn of a 20 by 100 rectangle, calibrate the axes and then switch on Rotary to set it up?
Thanks,
That would work, yes. Many machines use the same step size for X & Y, so you could start by just copying the X step size to see if that’s accurate.
LightBurn also automatically stores and keeps up to 50 backups when you change settings, so click ‘Load backup’ and look for one from around the time you started changing things.
I just spent some time first doing a flat burn of a 20 by 100 rectangle and got it perfect. I then went through Rotary setup, first testing rotation in Chuck mode and then switching to Roller. As has consistently happened my Y cut length is about 93mm. I thought I might have had slippage, so I slowed speed down from 100 to 10, did another test burn and rectangle was a 93mm.
Should i expect to cut 100 mm when cutting rotary?
Since you said Axis calibration is ignored in Rotary, what should my next steps be to get 100mm rotary cut?
I have a libre Office writer document with all steps done so far, how do I get that to you for review?
Also, I assume all code bases are the same,but just in case it matters, I am using latest Linux version Ubuntu 22.04 LTS…
Oz,
While waiting for your response (not complaining, I’m sure you are a very busy guy), I decided the only two parameters I could mess with were the steps/revolution or the Object diameter. Since Rotary doesn’t use object diameter, I fiddled with the steps.
I just scaled up the number of steps so 93 would be 100, 93/100=8000/X , so 93X=8000*1000=8602.15. Tried that number of steps/revolution and my cut was a little long, so I just truncated to 8602 steps/revolution and got exactly a 100mm cut.
It is not at all obvious to me why such a scaling is necessary, maybe rounding error or truncation in the calculator or just “slop” in the steppers? Any insights on why this is NOT exact but “black magic” may be helpful to all who experience issues with rotary setup.
Also, it just might be my machine being “flaky”, I noticed I have a high voltage power supply leak to ground that only manifests itself when I lean against machine when barefoot. I have ordered a new HV power supply and will retest after I install the new PS unit to rule out this as the cause of my issue.
It’s possible the rotary mechanism has a gear reduction in it, which could explain it - some do. It seems like a strange number (it’s probably 8600 or possibly even 8500), but still, if the results are accurate, leave it as is and go use it.
