Spend the day putting on a new pressure roller support and putting in the new rollers. I must admit that it was a long day, to say the least.

Finished up at 6pm and this is what I have to show. Another calibration file, but this time cut into foam. I am still not happy with the accuracy of the machine. My circles are skewed and they are not equidistant from each other. Maybe this is all I can expect from the machine. I was hoping to have the accuracy of the MKII demo cutting the black and white puzzle pieces out of Depron foam.

We still seem to be having issues with X. I was running the code at the feedrate of F=75. When I calibrate the printer with software, everything just about measures up. I did notice some bumps in the grip tape (I am upset that I should have installed it in a spiral). There was a little bit of overlap when one of the students put it together. I am planning on putting grip tape on the pressure rollers.

Maybe I am missing something in all of this. If I could only get into cutting, the students would be enjoying this.

Rob

-----------

DSCF3150-sm by aechacc, on Flickr
First Cut


DSCF3148-sm by aechacc, on Flickr
Second Cut

 

Hi Rob,

It looks like you have some flat spots on the small circles, is that right? If the flats spots are oriented side to side (when the foam is in the machine) it could be backlash on the x axis. I had some and used the backlash compensation in CNCusb to get rid of it and now my circles are fine.

http://www.phlatforum.com/viewtopic.php?f=264&t=2525

- Jeff

 

The flat spots on the tops and bottoms of your circles indicate backlash on the X axis is still present.

This could be due to slippage of the X axis stepper motor pulley or belt tension too loose.

Try taking a sharpie marker and place a line across the pulley and end of the shaft on the stepper. Then run the X axis back and forth while watching the mark you just made. If there is any slippage on the shaft, it will be easily observed visually.

If the tension is good and you are still getting flat spots, you will need to setup the backlash correction in the software to compensate for it. If you don't have a way to measure it precisely, you can just plug in a starting value and work up till you eliminate it visually by cutting small circles. Some typical values I have found to be in the range of 10 to 20 thousands of an inch backlash. Mine is .014 but each machine will be different.

As to putting grip tape on the pressure rollers, you might want to reconsider doing this until you have exhausted all other options. I really don't think it is needed or desirable.

 

Flash, have you had any backlash correction needed on your Y axis? I'm tweaking mine now, really just doing the guess-and-check method for X-axis backlash correction. Circles aren't far out of round but they are just a bit. But I'm not considering Y-axis backlash at all... just wondering if I should spend any time looking at that.

RJ

 

If you have flat spots on the left and right of the circle rather than the tops and bottoms, then you would need correction along the Y axis.

The Y does not cause the problems that the X presents because it is a single pulley to pulley drive.

The X is another story altogether. The dual rollers offer the possibility of the slack occuring between the two drive pulleys when switching directions. It is nearly impossible to set the tension tight enough to reduce all the slack and still not put too much pressure on the drive train.

The flat spots are a result of the stepper pulley being under motion for a brief moment in time in which the drive roller pulleys are taking out the slack that results when changing from the driving pulley to the undriven pulley.

I have yet to see any significant backlash in the single roller systems that have been used on the MK1 or the Y axis of the PP3.

I think the dual roller design will always have a small amount of backlash present. Fortunately we can correct for it in the software.

 

I should have said... I don't really see flat spots, so much as the circle appears "squished" on a diagonal axis. There are definitely no distinct straight lines along the circle.

It almost sounds to me like I applied too much backlash correction. I've only done two sets of test circles, and I didn't cut deep enough on the first so I don't remember exactly what they looked like. Just figured I would ask the question now in case there were any tips for when I get back to playing with it.

 

Squished along the diagonal can mean that your X and Y axis is not calibrated accurately.

From the photos, robhsmit's problem looked to me like flat spots along the X, but I could be mistaken. This too will cause the circle to appear to be oblong on the diagonal.

Cut a pattern with a circle inside a square to be sure. Measure the sides of the square to be sure both sides are exactly the same.

I usually use a 2" square and circle to check for this. Once I get that setup accurately, I cut a 4" and an 8" and a 10" version and make sure that the accuracy continues to maintain over the different distances.

 

Thanks for the input. I haven't had any time to try what was discussed, but I did continue to test other materials. I'll do another test cut and look at which direction is a problem. I believe it is X. However, whenever I do original calibration all axes measure up.

I've noticed that after sending 3 DXF test files of a site model to the printer, it is now slipping. Is anyone having any problems keeping a tension on the X-Stepper Motor?

Again I didn't continue to trouble shoot because I ran out of time. When I press on the belt of the two X rollers it seems extremely loose. I need to see if the motor may have moved or the pulley came loose off of the stepper motor.

I am under a time limit because we have a site planning project and the printer was lined up to print contours.