Z axis calibration out by a scale factor!

[mACca]

Hello People, first post here.

I'm having an issue with the Cnc Shark hd3 at my local wood workers club. The x and Y axis's seem to be in calibration fine but the z axis seems to be out by a scale of 2.4. for example when i told it to move up an inch in the step menu, it raised itself 2.4inches.

The issue is also present when running a g code file.

Apologies if its an easy fix as I'm just beginning, but usually computers are my thing and this has got me stumped.

Cheers

[Rando]

I'd check the control panel settings for "steps per millimeter".

Open the control panel (v2) and type Ctrl-G, then enter the password "787". This is not the same as the Preferences dialog.
Check the values for X, Y and Z; they should be all the same, but one might be negative.

All the stepper motors are 220 steps per revolution, and the lead screws are 1/4" per turn. Yes, if you measure an inch of length on along the lead screws you'll find it looks like 8 TPI. But, there are two threads, so it's really 1/4" per turn.

On mine, the numbers are horrendously bizarre fractional numbers: +/-157.48031496063. I'm sure they calculate out somewhere, but they look like irrational numbers to me!

If you also have the 4th axis, mine is set as 33.334 for the steps per degree. If an axis value is wrong, that might be also.

Here's a screen shot of mine. IMPORTANT: ignore the +/- on mine, because I swapped the X and Y axes!

Control Panel Settings

And who knows...maybe just clicking "Reset to Default" will get you there!

Hopefully that will help. The alternatives have more to do with broken wires going to the motors, so hopefully it's not that.

Regards,

Thom

P.S., don't bother with the PWM and Spindle speed stuff at the bottom, if yours has it: she no work .

[mACca]

Cheers mate,

I was hoping for something like that! I'm heading back tonight so I'll let you know how I go.

[rungemach]

Hi Thom,

In looking at the number on your control panel, I tried to do the math to tie everything together.

starting with the 4th axis, I come up with 12000 steps per chuck rev (33.334 steps per rev x 360 deg /rev)
With a factory stated ratio of 6 to 1 on the pulley drive, it would be 2000 motor steps per chuck rev (12000/6, which is a standard stepper at 10 times micro stepping (200 steps per rev x 10).

If you have .5 inch per rev screws on the x and y, that would be equal to about 12.7 mm travel per rev of the screw.
at 200 steps per screw rev and 10 times micro stepping (2000 micro-steps per rev) you would get 2000/12.7 which comes out to be 157.480315 etc.. steps/mm as shown on your control panel.

My question is, does your machine have .5 inch per rev screws on all 3 axes. At one time the sharks were all were .25 in/rev pitch and then later with the hd machines the x and y were .5 and the z was kept at .25 inch /rev. Did they change the Z to .5 as well ?

I also noticed that, to me, the axis acceleration values seem a bit high, (compared to what I use on my mach3 machine). With acceleration values too high, you can lose steps from time to time if the machine can't physically keep up. Especially on the z when it has to lift the router weight on retract. FWIW, The max 3d acceleration of 250mm/sec/sec seems more in line than the 500 for the individual axes.

Bob

[Rando]

Bob:

I honestly couldn't tell you what the X and Y lead screws are, but I'll take a look....

My Y (your X, along the length of the gantry) shows 4 threads, 8 TPI. For obvious reasons my X (your Y, in line with the T-track slots in the bed) is a bit out of the way, but it looks the same. The Z is larger diameter. At one point I measured the Z-axis and it was 1/4" lift per full turn. I've never changed the values in the dialog, just the signs, since I swapped X and Y axes. Mine is an HD2+ model.

IIRC, the value was confirmed by Joe, the programmer doing the control panel, when he increased the field calculation resolution to 10 digits. It was one of those numbers where I said "um, okay, I'll just use that, and not worry about where it came from" . Sure wish it was an integer, though, would seem so much more neat-and-tidy. It's those metric-people's fault! hahahaha.....

Regards,

Thom