Each Pass the position changes

Discussion about the arduino based g-code interpreter, grbl
johnc
Posts: 9
Joined: Wed Dec 24, 2014 1:46 am

Each Pass the position changes

Post by johnc » Sat Apr 18, 2015 7:45 am

Does anyone know what setting I have wrong to be causing this issue.
Basically each pass, it moves slightly to the left and top.

Here is a photo of the issue.

Image

It was working fine when I was using step stick drivers, but they were suffering from underpower and heat issues.
They were not rated high enough for the steppers I am running.

So I replaced them with these ones that are the right rating, the steppers are no longer skipping steps, but its got this offset issue

Image

WillAdams
Posts: 8514
Joined: Mon Apr 09, 2012 6:11 pm
Location: Pennsylvania --- south of the Turnpike, East of US-15
Contact:

Re: Each Pass the position changes

Post by WillAdams » Sat Apr 18, 2015 12:58 pm

Does the same offset happen if pushing a pen?

I’d check pulley set screws first.

Wiki helpful? http://www.shapeoko.com/wiki/index.php/ ... rmittently
Shapeoko 3XL #0006 w/Makita RT0701 Router w/0.125″ and ¼″ Elaire precision collets
Nomad 883 Pro #596 (bamboo)

cvoinescu
Posts: 4442
Joined: Thu Jul 19, 2012 6:50 pm
Location: Camberley, UK
Contact:

Re: Each Pass the position changes

Post by cvoinescu » Sat Apr 18, 2015 2:08 pm

Those drivers can skip steps in a different way. Try disabling the auto-power-down feature -- in many of those modules, it is badly designed. There can also be a problem with the DIR optocoupler being too slow, which can cause steps to go in the wrong direction occasionally. There's not much you can do about that in GRBL, except maybe try to invert the STEP signal (in LinuxCNC and Mach3, you can control the timing of the signals to a greater extent).
Proud owner of ShapeOko #709, eShapeOko #0, and of store.amberspyglass.co.uk

johnc
Posts: 9
Joined: Wed Dec 24, 2014 1:46 am

Re: Each Pass the position changes

Post by johnc » Sun Apr 19, 2015 8:36 am

cvoinescu wrote:Those drivers can skip steps in a different way. Try disabling the auto-power-down feature -- in many of those modules, it is badly designed. There can also be a problem with the DIR optocoupler being too slow, which can cause steps to go in the wrong direction occasionally. There's not much you can do about that in GRBL, except maybe try to invert the STEP signal (in LinuxCNC and Mach3, you can control the timing of the signals to a greater extent).
Thanks for the ideas. I changed a number of things and it seems ok now.

- Turned off auto power down in the grbl settings
- Changed the dip switches on the drivers to 16 microstep and recalculated the steps/mm setting for x & y in grbl

cvoinescu
Posts: 4442
Joined: Thu Jul 19, 2012 6:50 pm
Location: Camberley, UK
Contact:

Re: Each Pass the position changes

Post by cvoinescu » Sun Apr 19, 2015 3:39 pm

I was thinking more about the auto-power-off feature of the driver (the one in GRBL is not buggy, and unlikely to be involved here anyway). But glad it works now.
Proud owner of ShapeOko #709, eShapeOko #0, and of store.amberspyglass.co.uk

johnc
Posts: 9
Joined: Wed Dec 24, 2014 1:46 am

Re: Each Pass the position changes

Post by johnc » Sat May 09, 2015 6:20 am

I have done numerous test cuts and it is lining up properly all the time now.
I also did simple cuts moving the bit say 100mm then measuring each direction to ensure it is moving the correct amount.

However when I run a program from Easel the sizes are slightly out.

I cut a simple square out of some thin MDF.

Programmed Size - 90x156mm
Actual Cut Size - 85x145mm
Difference - 5x11mm

Programmed Position - X: 105 - Y: 122
Actual Cut Position - X: 93 - Y: 110
Difference - X:12 - Y: 12

I notice there seems to be a pattern of 11-12mm in the difference apart from the X cut size.

Can anyone give me any advice on what to look at to see where this is going wrong?

WillAdams
Posts: 8514
Joined: Mon Apr 09, 2012 6:11 pm
Location: Pennsylvania --- south of the Turnpike, East of US-15
Contact:

Re: Each Pass the position changes

Post by WillAdams » Sat May 09, 2015 11:30 am

How are you measuring cut position?

Are you taking into account endmill diameter and runout?

Typically one would drill holes and measure either from the hole centers or the left/right/front/back-most position of a pair in a given axis. (I guess it would work on diagonals if one did the math).

The traditional way to do this is to cut a diamond-circle square test and measure that.
Shapeoko 3XL #0006 w/Makita RT0701 Router w/0.125″ and ¼″ Elaire precision collets
Nomad 883 Pro #596 (bamboo)

johnc
Posts: 9
Joined: Wed Dec 24, 2014 1:46 am

Re: Each Pass the position changes

Post by johnc » Mon May 18, 2015 10:52 am

WillAdams wrote:How are you measuring cut position?

Are you taking into account endmill diameter and runout?

Typically one would drill holes and measure either from the hole centers or the left/right/front/back-most position of a pair in a given axis. (I guess it would work on diagonals if one did the math).

The traditional way to do this is to cut a diamond-circle square test and measure that.
Thanks for the suggestions.
After further testing I have managed to get the cutting right. The automated calculators were not working for me. I assume I was putting in values wrong somewhere.
I ended up making repeated lines with a marker, measure and adjust. Eventually I managed to get the X and Y producing shapes the right position.

Next was onto the Z. I placed a ruler next to the rail, then moved it up and down comparing the positions.
Moving up and down various distances matched the ruler.

However when running a cutting job it seems like each pass is going the wrong direction.
I place the bit on the home position
I start the job. The bit moves up and over to where it will start cutting.
The bit moves down and correctly cuts the first layer.
The bit then moves up again, then across to the start of the next layer.
The bit moves down, but doesn't touch the material.

This cycle goes on with it never touching the material again. If anything, it seems like each pass is slightly higher, like it is running the wrong direction.

It knows which is up and down though because other functions work fine.

Any suggestions of what to look for to fix this?

WillAdams
Posts: 8514
Joined: Mon Apr 09, 2012 6:11 pm
Location: Pennsylvania --- south of the Turnpike, East of US-15
Contact:

Re: Each Pass the position changes

Post by WillAdams » Mon May 18, 2015 11:04 am

Did you try the wiki link I posted earlier?

Looks like lost steps caused by overheating.
Shapeoko 3XL #0006 w/Makita RT0701 Router w/0.125″ and ¼″ Elaire precision collets
Nomad 883 Pro #596 (bamboo)

johnc
Posts: 9
Joined: Wed Dec 24, 2014 1:46 am

Re: Each Pass the position changes

Post by johnc » Mon May 18, 2015 11:59 am

WillAdams wrote:Did you try the wiki link I posted earlier?

Looks like lost steps caused by overheating.
Sorry I totally missed that link. Thanks for pointing it out.

I am not sure about being overheating issue as the other drivers are running fine. One driver is running two steppers on the Y axis with no issues.
Also it runs fine when I manually move the Z axis even after I have been running the machine for some time.
I will add some extra cooling though just to make sure though.

The link does show an interesting point about plunge rate for the Z axis.
My GRBL config doesn't have a feed/plunge rate for the Z axis, but I am using Easel and set the depth per pass for this. I am not sure if it is the same or not.

The current GRBL config is

$0=53.100 (x, step/mm)
$1=53.100 (y, step/mm)
$2=24.930 (z, step/mm)
$3=5 (step pulse, usec)
$4=100.000 (default feed, mm/min)
$5=250.000 (default seek, mm/min)
$6=220 (step port invert mask, int:11011100)
$7=255 (step idle delay, msec)
$8=25.000 (acceleration, mm/sec^2)
$9=0.050 (junction deviation, mm)
$10=0.100 (arc, mm/segment)
$11=25 (n-arc correction, int)
$12=3 (n-decimals, int)
$13=0 (report inches, bool)
$14=1 (auto start, bool)
$15=1 (invert step enable, bool)
$16=0 (hard limits, bool)
$17=0 (homing cycle, bool)
$18=0 (homing dir invert mask, int:00000000)
$19=25.000 (homing feed, mm/min)
$20=250.000 (homing seek, mm/min)
$21=100 (homing debounce, msec)
$22=1.000 (homing pull-off, mm)


I will do some further testing.
Edit: It looks like I have an older version of GRBL also. I will upgrade to see if that helps also.

Post Reply