How to fix non round shapes?

Thanks everyone for the tips. I will try the circle test and post another cube result.

Having problems finding a 15mm. So I will use a 20mm? Should I be using thingy verse or make my own?




Sent from my iPhone using Tapatalk

 

This is really not the case and not the way stepper motors work. The stepper error and XY calibration is a red herring in this situation and just causes confusion. I think we can all agree that Peter should focus on the non-round large circle test case. From what we have seen there is no big problem with the X axis. The Y axis motion is stopping a a point symmetrical around the circle. The fact that this is repeatable is hard to explain by any slippage or skipped steps. If it is not exactly repeatable then a loose set screw with the pulley moving back and forth at some symmetrical point of extra force would explain the problem. I assume that Peter has checked this and ruled it out so I'm not sure where to go next.



For my XY calibration as a red herring claim I will put down a few comments for Peter and others to look into and leave it at that. For stepper calibration calculations I refer to Leon's discussion and calculation of the 80 steps/unit value:
http://forums.robo3dprinter.com/ind...t-x-and-y-steps-mm-are-wrong.1536/#post-10719
And Hunters calibration guide: http://reprap.org/wiki/Triffid_Hunter's_Calibration_Guide

A properly driven, nonstalling stepper motor will have not much more than 5% of a step error over one step and the error is noncumulative no matter how bad the error or stepper is. The stepper can do an entire rotation and it will no more than that % of a step of error. So for example using the 5% error value or the Robo3D the linear error due to the stepper would be (1/200)*0.05*(40mm circumference). The Robo's pulley circumference is 40mm so this gives around 0.01mm of error due to the stepper, noncumulative. This is tiny. The other error that could be larger and usefully calibrated for is belt stretch.

 

Here is my 15 mm cube I created in Cinema 4D.

My measurements are:
14.80 Y
15.05 X
15.01 Z



Using Matter Control, sometimes the print starts really heigh up not touching the bed. I set my z level, manually and it prints a good print. Then when I go to start a new one, the z height starts really high up. I adjust the z height as it's printing and once it's dialled in, I restart the print. Again, it starts height up. Same settings as Repetier. When using Repetier, doesn't happen.



Sent from my iPhone using Tapatalk

 

Peter, Did you try gently resisting the Y axis by hand moving it with the manual controls?

 

Y axis manual control? How would you do that? I tightens the belt underneath the bed to a nice snug. Is this what you mean?

I checked the visible screw underneath, but without taking off the bed, not sure what else could be tightened there. What kind of screws are those holding the bed down? They are so weak, as if made of butter. I attempted to remove the bed but gave up after the screw started to give. Until I can replace them with real steal ones, I decided to postpone that part.


Sent from my iPhone using Tapatalk

 

In Repetier Host or the LCD controller or Mattercontrol you can manually tell the printer to move an axis. It is under the manual control tab in Repetier. As it is moving try to resist the movement gently at first with your hand. I am looking for a slipping set screw on the Y pulley that is hidden under the bed.

 

Wow - I don't know what screws you mean that are like they are made of butter? The bed is held with steel screws inside nylon standoffs. If the nut over the rail or the nylock nut underneath the rail are loose then I guess it will move like butter. If this is the case you really need to tighten them or get help from Robo3D.

 

No no no, to turn the screws, they feel as if they want to strip. I am talking about the screws on top of the bed holding them down.

To manually move the y axis is working. Are you suggesting hitting move by increments and see if it skips?


Sent from my iPhone using Tapatalk

 

Yup!

 

Yes but you can't just turn the screws with a screwdriver from the top. If the rail to bed contact is loose (like butter?) you need to tighten the nylock nut under the rail. The screw might not be able to turn because of the nut in the middle under the nylon standoff. It is complicated for me to explain how to tighten if if needs it. Please don't strip them.

 

Peter, Was there anything wrong with this print from your earlier posts? If so what exactly.

 

Forget the butter comment. Nothing is loose. I rechecked.

What about my measurements? What can you see by the results?

The print you asked about, it failed due to the bed being hit and move the print out of alignment. Did not re-attempt.


Sent from my iPhone using Tapatalk

 

Ok Peter, I would focus on the circle and the Y axis to start.

Your cube looks great but I would defer to Tesseract on the tolerances.

My 15mm cube has some X and Y variance due to some overextrusion with my Colorfab filament I think. I get
15.05 to 15.28 range (Don't know if X or Y)
15.16 to 15.30 range (Don't know if X or Y)
15.00 on Z

The X and Y get better if I reduce the Extrusion multiplier setting.

 

Well actually while you are hitting the move selection I want you to try to resist the movement with your other hand, gently to see if it skips. OK? Try setting the move to 10mm for example. Try it in both directions.

 

I'm turning in. Good luck and I will be around tomorrow. I'm sure you will get lots of good suggestions in the morning from others that will conflict what I have suggested .

 

Here's another test. Looks round!



I am noticing that there is a shift at the bottom of my prints.

Could it be a broken tooth on a gear?

Sent from my iPhone using Tapatalk