can't figure this out.

[MikeUSNRet]

trying to do a litho in pvc pipe, but regardless of having CV On or Off, I get the same results.

As the cut progresses (starting bottom right corner), the right edge does not stay straight, but rather starts drifting off to the right with each pass.

I'm using a simple belt driven 40:1 reduction rotary, but at no time do i have any binding or slipping.

at a loss and looking for suggestions.

[NormanAlbert]

Hi: Have you calibrated the rotary axis ? Does it turn exactly 360 degrees when you tell it to? Norm

[MikeUSNRet]

I've calibrated the rotary axis till i'm every color in the spectrum and it is spot on 90 gets me 90, 182 is exactly at 182. I know this isn't a Vectric Product issue, but just trying to get some reasonable explanation of some sorts. I'm using Mach 3 and as I mentioned, there is no difference if I have CV on or off, if I cut at a moderate speed or set the system to something slower than the mating speed of a sea slug in antartica, I always end up with over-run of some sorts. as far as axis alignment to the cutting tool, that is dead on as well.

oh well..back to the drawing board and perhaps find something that has a geared reduction ratio instead of a belt driven one and the start of building an entirely new machine

[PaulRowntree]

Mike, can you show a preview of what you thought it should look like? Is the left edge of the cut deformed as well?
If you made the design twice as long along the axis of the PVC, does the error start at the same point?

[NormanAlbert]

Hi: Another question, is it a timing belt you are using? One with teeth and matching pulleys? That's what I have and it works very well. Norm

[NormanAlbert]

Hi Again: Another thought, are you sure the pvc pipe is not slipping? I've had that problem when chucking the inside of the pipe. I use a 4 jaw chuck with 2 rubber grips per jaw. Look for wood bowl turning lathe supplies. Norm

[MikeUSNRet]

Norman:

Yes, I am using a toothed ( i guess you would call it) timing belt that goes between the pulley on the stepper and the other one of the spindle. Definitely is not slipping.

Although I made the rotary so that I could put the chuck from my wood lathe on it, I was not using that chuck, but was instead using a very tight fitting plug on both the headstock and tailstock of my rotary that I had turned on my wood lathe. I hadn't thought about slippage. suppose it's probable even though I couldn't get any when I tested for this on my wood lathe.

Paul:

Not at home now and won't be for a few days so getting a picture of the model up won't happen for a bit.

using thick walled PVC pipe so :

Length was 5" OD is 4.25 and thickness is .25
setup up wrapped job and created component from bitmap.
image height was set to 0.15 and Gap Below Model set to 0.04

created a Rectangle border around the image in order to set a cutting boundary.

set the boundary to be .25 border on the top and bottom and the width of the image only covered 1/2 of the pipe diameter.

With all of this, it was a basic rectangular shape litho no different than one would see on a flat piece of corian.

Mike

[NormanAlbert]

Hi Again: Okay, one more thought. Have you gone to settings (in Mach3) and put in the radius of the piece for that job? Norm

[MikeUSNRet]

You might be on to something there....had I manually entered the A Rotation Diameter in the Settings tab before running the job? No. definitely something I had not thought about and overlooked..in fact, after reading more in depth thru the forums, there was a Post Processor written called Rays Wrap Y to A (Inch).pp and even a mm one along with a mach3 file that addressed this very issue which would automatically set that value in for us forgetful folks when you loaded the gcode into Mach3

[rdean33422]

You are certainly welcome to use the posted files but I don't think the radius offset function is your problem. Is it possible to send me the file so I can check it out here?

Email if too big.
rdean33422@aol.com

Ray

[MikeUSNRet]

Sending you email Ray.

Mike

[rdean33422]

Got the code file you sent and it runs fine in simulation i don't see any problems there. I gave you some things to try in my last email to you but just ignore what I suggested as that is not the issue.

I have attached a modified picture of the project so I can explain what I believe is happening.

The project 0, 0, is in the center of the work piece with X axis extents of -2.25 to +2.25. The A axis extents are about -102 to +102 degrees. The starting point is in the right lower corner and the tool path direction is shown as the blue line in the picture. Everything starts out just fine until we get to the top of the green line in the picture and then it starts to skew over.

Looking at the red line in the picture this cut stays where it should be through out the program so that rules out any lost steps in the X Axis.

I can not see any lost steps in the A axis either but if there were it would cause a pass to be cut over top of the previous one only a little longer.

If the gantry was not aligned properly with the X Axis then this issue would have started at the very beginning of the cut and have a linear curve but it doesn't start until the top of the green line. So what is happening?

I believe that as the gantry is moving from the -2.25 position toward the positive end of the table the V rollers on the far side of the X axis are climbing the rail and allowing the gantry to move in a Y minus direction. The farther the X axis moves in a positive direction the greater the skew becomes but when it moves back to the -2.25 position everything drops back into place again. It could be the rails themselves are loose or not parallel to each other. One of the X axis drive chains could be out of adjustment of something has slipped causing the gantry to only be driven from one side. The lower X axis V rollers could be out of adjustment and allowing the gantry to tip up on that side but what ever it is I believe you will find a mechanical issue with the machine on the X axis.

I believe it was in 2011 when you had an issue with a lithophane that was cut like a parallelogram and skewed sideways. This is exactly what is happening now but to a greater extent and the rotary axis curvature make it worse.

Trouble shooting an issue from afar is at best haphazard but I gave it my best shot hope it helps.

Ray

[MikeUSNRet]

I understand what you are saying, but I don't think that can be happening because I am wrapping Y to A, therefore the gantry stays stationary over the center of the A axis spindle. Only the X axis moves as A rotates. We can see that by the Red line being parallel to the bottom of the piece at an equal distance as it progresses, the X axis is returning to -2.25 consistently each time as it should so this would mean that something is happening with the A axis causing it to over rotate and progressively get worse with each pass.

Yes, I have had problems similar in the past and I have re-built this machine 2 or 3 times and continue to have issues so now with your thoughts on this, I'm starting to think more along the lines of electrical vice mechanical. Coming to mind are:
1. the Stepper motor has an issue which could be related to Thermal Breakdown as it heats up with use, although, it doesn't feel any hotter than the other axes steppers after longer use.
2. The driver is bad or going bad.
3. Power supply not stout enough to keep up with the demand of 5 steppers etc (2 on Y, 1 X, 1 Z , 1 A)... Maybe adding a dedicated power supply to the rotary driver would work.

I think I just need to scrap this chain drive build and start over with a Joe's R&P or something as funds become available. It has been nothing but problems for me and my neighbors becomes my best friend in the winter when he wants to light his wood stove. He hates when I only use my wood lathe..less food for the stove...lol

I am using shielded cable between the drivers and motors.

[Greolt]

What is the job supposed to look like? Can you post the Aspire file?

That will tell me what you are expecting and what machining strategy you are using.

Till then I am blind.

First thing to look at when something does not come out as expected is, does the machine return accurately to origin at the end.

Greolt

[MikeUSNRet]

unfortunately I am away from home and will be for a while longer so can't post any files / pics, but to sum it up, it is a simple rectangular picture that is being placed on a straight piece of PVC pipe that has no contours, so the left and right sides should be 90 degrees from the top and bottom, no curvature at all whatsoever