Y axis going over by 1mm

i am working on calibrating my 1k x 2 machine and am having the issue of overshooting by 1mm -2mm. i have seen many who are under, but not over in measurement. i just finished securing the belts from slipping with success by using gorilla super glue gel which is working great. i also found the high and lows of the pots, then set them a little past the mid point. the belt tension is around 3.5lbs at 1in lift from midpoint on all belts. to note, the i have only testedthe y axis so far and have just ordered a metric machinist rule. Also to note, it is consistent when going back and forth. when i tell it to move 600mm it goes past the mark, but when i return to the other direction, it is spot on the start position.

my next plan of action is to loosen the belt a little and double check the alignment of the pulleys.

thanks in advance for any insight and advice!

My first thought is you need to calibrate the steps per inch of the steppers for your particular machine.
Each one is slightly different.

@RobertA_Rieke has posted a video on how to do this and it can be found here:

https://www.youtube.com/watch?v=vSxFutlfngk

how would i do that using rhinocam + ugcs? or does easel save it to the hardware.

went through his videos last night. great stuff robert! on behalf of all noobs thank you!

based on looking around, it looks like the settings are saved on the hardware, please correct me if i’m wrong. i will mostly be using rhino/rhinocam/ugcs to create and communicate with the machine.

You can adjust the steps per inch using Universal G-Code Sender. Under the command tab, you enter the values.

Thanks for the responses both of you! I was just not sure if the settings were saved to the machine or just for easel. i found a post on squaring the machine (although i know that isn’t going to effect this, but will hopefully save me some trouble down the road), so i squared it and set the steps through easel. now it’s spot on. just have to calibrate the rest. next ill be testing some cuts.

when checking for square on the z axis, it is very slightly off. how detrimental has this been? the x axis rail is perfect but the z itself seems to have very slight flex with the dewalt. roughly 1/32 inch.

after some test cuts with some “stair stepping” by around .5mm after the first pass, I checked over the pulleys for belt alignment as well as the screws and screw alignment with the flat part on the shaft. After checking belt tensions again, I ran the same test with the same stair stepping results. I figured i would try a pocketing test and see if it shows more of what is going on via the walls and floor of the cut. The results were almost perfect which is throwing me off on figuring out where to tune.

the parameters for the profiling test with the stair stepping were a 2" square and 2" circle inside cutting to a depth of “.2” . ; Everything was off by a bit in this test, however I do not have the results in front of me.

the parameters for the pocketing test are:
2" circle
.19 depth
.03 rough depth
.19 rough depth
.03 rough depth cut
.03 finish depth cut
cut speed 50 in/min
tool .125 2 flute flat
25% stepover

the results:
measure in the x = 2.071 with a .071 difference
measure in the y = 1.991 with a -.009 difference
measure in the z = .202 with a .012 difference

software used is rhino/rhinocam and ugcs

also want to throw this out there for whoever might find use.
it has the speed calculator that was posted on these forums along with a stepper calibration tool i threw together to make things go a bit quicker when calibrating the steppers

http://www37.zippyshare.com/v/22VefyhW/file.html

did the x axis bolt mod (inserting screws or bolts in between the rails in order to improve rigidity). Will have to try a different store to see if they have 3/16 steel or aluminum. Then I will give that method a shot.

Ran a 40mm test square and am having decent results, but not there yet.

top of the square in x axis is 40.01
bottom of the square in x axis is 40.15

right side of square in the y axis is 39.62
left side is 39.78

Is this normal to see? top to bottom in the x axis is a .14 variance and .16 in the y left and right sides. in other words, not square.

tomorrow or Friday night i will run another test going from 80mm down to 20 mm squares. It should be a bit easier to measure. I will also be rechecking everything for square.

did 4 test tonight. 80mm 60mm, 40mm, and 20mm profile cuts to see consistency. I did only one modification to the $11 by setting it to .010

the results are as follows
80mm test
top x =79.58 (-0.42)
bottom x=79.53 (-0.47)
right y =79.8 (-0.2)
left y=79.8 (-0.2)

60mm test
top x=59.64 (-0.36)
bottom x=59.58 (-0.42)
right y=59.92 (-0.08)
left y=59.92 (-0.08)

40mm
top x=39.59 (-0.41)
bottom x=39.54 (-0.46)
right y=39.8 (-0.2)
left y=39.96 (-0.04)

20mm test
top x=19.52 (-0.48)
bottom x=19.44 (-0.56)
right y=19.88 (-0.2)
left y=19.98 (-0.02)

things seem to be fairly constant. Based on reading around, it looks like i may need to tweak my pots at this point.

Adjusted my pots by adding more power to my x, and a tiny amount to the y. I rechecked the movement in both directions with a 600mm machinist ruler (same as in the video by RobertA_Rieke) the x was only off by .01mm roughly.

Results
80mm test
x top = 80.14 (+0.14)
x bottom = 80.14 (+0.14)
y right = 80.02 (+0.02)
y left = 79.94 (-0.06)

60mm test
x top = 60.34 (+0.34)
x bottom = 60.22 (+0.22)
y right = 60.01 (+0.01)
y left = 59.96 (-0.04)

40mm test
x top = 40.32 (+0.32)
x bottom = 40.26 (+0.26)
y right = 39.98 (-0.02)
y left = 39.92 (-0.08)

20mm test
x top = 20.4 (+0.4)
x bottom = 20.3 (+0.3)
y right = 20
y left = 20

Will be adjusting the pots again and re-calibrating the $100 and $101

I also noticed slight play in the z axis rail when pushing in the y direction. The x direction has no noticeable play. Also, the bottom V wheel with the eccentric nut will not tighten against the rail like the others will. It rubs against it slightly, however it is still really easy to spin.
I also measured diagonally (left rear to front right etc.) to check for square. it is around 1/16 off from the other, which is the difference before i squared it. It seems to have shifted back to where it was.

This post has more or the less become my logbook for tuning my machine. I hope that some of the reports may help someone else who is trying to adjust their machine. Any insight is more than welcome also.

Did my first project with considerable success tonight. It was a puzzle for my kids. It was successful in that, the cut went as intended as far as the machine was concerned. It was unsuccessful due to me grabbing the wrong endmill. I was supposed to grab a yonico 1/8 upcut, which actually measures to a .1175" diameter and instead grabbed the milecraft 1/8 which measures to .13".

I programmed the cut with the yonico .1175 in mind with a .01 offset to the joining parts so there would be a small amount of play. This resulted in .0225 amount of play which is a bit too much. Oh well, lesson learned.

Be sure that you measure your endmills and check that you are grabbing the correct one before starting a cut.

Redid the puzzle tonight with the correct endmill and it came out almost perfect. one thing I noticed while cleaning up the parts, is some chatter. So I will be continuing to crawl the forums looking at solutions.

here are the results.