I can't seem to cut a round circle on my new phlatprinter.



The X axis seems to be slipping slightly on the pulley and is causing a lot of backlash.

Attached files

 

I am not really clear on how to properly measure the backlash. I tried moving a specified distance and then moving back measuring the difference but that value didn't seem to do anything.

It almost looks like the belt teeth are not the same size as the pulley teeth. I don't get how so many people seem to be able to run without any backlash compensation when there is this much movement in a properly tightened belt.

 

I was just using cardboard to try out the sharpie attachment. All of my calibration had been done with plywood.

Here it is printing the test pattern. I'll try measuring the backlash again and see if I can get a round circle.

http://www.youtube.com/watch?v=hzHPYvWJ9LQ

 

I have the belt tensioned correctly so I am not sure why the belt can move like that on the pulley.

Are the belts and pulleys for the Z and X axis the same? perhaps I have them mixed up.

 

It looks like you might have a set screw loose on the stepper motor pulley.
With the pp powered up, loosen the tooth belt.
If you can move the pulley even a little with you fingers check the set screw.

The video posted does not accurately show backlash because of visual parallax.
Don't trust your eyes. Trust your fingers.

Make sure the set screw engages the flat of the stepper motor shaft.
Make sure the set screw is in all the way, and tight.

This is a gotcha that can have you climbing the walls, and it's simple fix.

Hope this helps.
GHB

 

The pulley was slightly loose when I firsst assembled the machine but I have since removed it and retightened it several times to be sure it was not causing me trouble. The pulley is rock solid on the shaft.

In the video dont look at the lines I drew on the pulley. Look closely at the teeth on the pulley and at the lettering on the belt I can assure you that the pulley starts moving before the belt. It is more apparent if you view it fullscreen.

I calculated the backlash which is a whopping 0.021" and entered it into the planet CNC software. The circles are closer to being round but everything else is completely messed up.

Backlash was calculated by moving a sheet of plywood 2 inches and then back 2 inches and measured the difference with digital calipers.

Attached files

 

Which software version are you using?
This bug might be fixed in beta version.

 

Helno, I looked at the video on my big screen and saw what you mean about the pulley slipping. It looks to me like your belt and pulley don't match, like if the tooth of the belt are smaller than the pulley grooves.

 

I'll give the beta software a try but I think my problem is more mechanical than anything.

Are other people using this aluminum pulley? what size are the teeth on yours?

 

The beta software made the print much better but the circles are still flattened on the X axis.

The error in the circles seems to depend on which way they are drawn so perhaps I have Yaxis trouble as well.

 

So if you use beta with backlash compensation set and it was not messed up?

There are 3 issues here:
- backlash on X axis
- backlash on Y axis
- software messing things up if backlash configured in settings

I'm trying to fix third issue,

 

Well I am still scratching my head at this.

I think I'll have to give Mark a call to figure out why I have so much backlash on teh x axis.

The software compensation seems to only do so much. The machine can cut boxes just fine but with so much backlash it cant cut a round circle.

 

Hmm, I wonder if the X stepper is stalling. When it's cutting a circle it's running through a wide RPM range as it goes around. I suppose it could be hitting some resonant frequency where it stalls as it does, that way it wouldn't effect square cuts, just round ones. I seem to remember something about problems like that if your belt is too tight. I sure someone else will chime in with some better information. (or even correct information if I'm completely wrong!)

 

Gefahren, where did you read this? I'm having the same problem and some other posted on another thread that they have the same problem too. Circles are not round. The tension on the x belt is way higher than the tension on the y belt.

I wish my stepper had the shaft extended on both side so that I could have made the resonant mod that was posted here somewhere.

 

Tigerpilot when you say you are having the same problem does it seem like you have alot of backlash on the X axis?

Do you have the newer aluminum pulley or the older plastic one on the x axis?

 

I have bought the new pulley myself form that horrible company. You didn't read my runt about the high shipping charges?

For the life of me I can't solve the unround circles. I bought on Sunday a dial indicator and measured all my backlashes. I had nil, nada, none on all axis' yet I still have an abortion for circles. I've been cutting mostly wood and I bought today some 1/8 bits on line for just in case that some of the problem is that the bit is a bit too thin for wood cutting. I tried it with a 1/16. I'll do some dollar tree foam cutting tomorrow to see how it cuts lite stuff.

 

Hi Yoram,

To measure the backlash with a dial indicator;

Jog the axis just in the one direction only then place and set the dial indicator and the Mach axis DRO to zero.
Jog the axis in the same direction as before away from the dial indicator some 50mm or so.
Use the Mach key “Goto Zero”.

Does the dial indicator read zero ?? – if so then you have no backlash on that axis.

Hope this helps,

Tweakie.

( Ivan... )

 

I did only 1" or 25.4mm. I'll try it with 2" or 50mm. I also didn't do the jog in one direction only. What I did was go forward and back and then I set all to zero. This way all axis' showed me nil, zero, nada backlash. You may be right though, about the 50mm, on the x axis since it uses more than 15", or about 38cm, to make one rev of the belt. That may explain the odd reading I have coming out of the jog. Did you read it on the Mach forum? I'll try it the way you said.


EDIT: On second thought, I jogged back first and then forward and then I put everything to zero. That means that I did do it like you said, about the direction, that is. I didn't do it for 50mm and will try it.

 

Its perhaps not the amount of the jog so much as maintaining the direction. The DTI can only be positioned after the first jog has been made, perhaps the following diagram explains it better.

Tweakie. Attached files

 

I did that the other day when I recalibrated mine and the measured backlash was much higher than I had measured before.

 

I'm out of town for the next few days but I'll be sure to contact you next week.