Z lag (?) eats through my tabs and part - help!

[chairguy]

I have a 2014 PRS Alpha 96" x 48" that I'm going to use for two-sided 3D milling of hardwoods. I've run the leaf bowl sample file included with the machine and had no problem a few weeks ago. But now when I go to mill some chair legs, the Z seems to occasionally stick in low positions and ends up eating through my tabs half the time, only to pull up and back into correct position, then do it again somewhere else. (It may even dive down prematurely - I can't recall). It also eats through the part (chair legs) on occasion. I've tried recreating the cut files and even tried running from a different computer, but the mistakes happen in the same places, so this isn't random. Attached are photos showing my paths (side view) in Partworks 3D (looking down the side of the file, you can see all the way through the tabs, so the file seems correct.) But in the photo you can see how the 1/2" round nose roughing pass eats right through the occasional tab or chair leg, leaving the whole thing looking like swiss cheese. Anyone ever seen this before? Any ideas?

I'd be incredibly grateful for any help.

Thanks,

Kevin

[chairguy]

For some reason my attachments got reduced to something too small to see. Does anyone know how to load larger pictures on here?

[gc3]

if preview look correct then is a machine issue or how you are setting up material on table and setting up machine for toolpaths

[steve_g]

Kevin…
What are you using for a spoilboard? If it’s been a few weeks since you surfaced it, good chances are, if it’s MDF its absorbed moisture from the sir and swollen. The swelling happens in a non-uniform manner, It may be fine where you zeroed but not 5 inches away…
SG

[scottp55]

Kevin, Try converting to a .jpg about 1900X1200 or slightly smaller, that's about the limit for attaching.
scott

[Brady Watson]

Wood moves when it is being cut - especially solid wood. Some have a tendency to bow up as internal stresses are relieved (eating tabs) - in those situations tabs alone 'ain't gonna cut it' - so you need something else too.

Excessive speeds will cause the tool to run rough. What speeds are you running when you witnessed the tool 'self-correcting' ? This would be your MS speeds. (EG - MS,2,1) in your SBP 3D finishing toolpath.

-B

[chairguy]

Thanks everyone, although I apologize; I may not have described my problem so well.

Brady: After rechecking my Move Speeds and Ramp Speeds, I noticed that the settings that came programmed into the software weren't at their defaults (I was a bit surprised by this, as I would think when I load the software it would start with default settings). For each value below, the first number listed is where the setting was; the second number is their default, to which I reset the values:

[VS]
[S]PEED VALUES
XY Move Speed - 3 > 3
Z Move Speed - 3 > 1
A Move Speed - 2 > 1
B Move Speed - 50 > 50
XY Jog Speed - 14 > 12
Z Jog Speed - 6 > 3
A Jog Speed - 4 > 3
B Jog Speed - 100 > 100

[VR]
[R]AMP VALUES
XY Move Ramp Speed - 2 > .4
Z Move Ramp Speed - 2 > .4
A Move Ramp Speed - 10.16 > .4
B Move Ramp Speed - 10 > 10
XY Jog Ramp Speed - 2 > .4
Z Jog Ramp Speed - 2 > .4
A Jog Ramp Speed - 10.16 > .4
B Jog Ramp Speed - 10 > 10
Move Ramp Rate - 5.08 > .2
Jog Ramp Rate - 5.08 > .2
3D Ramp Threshold - 100 > 100
Minimum Dist to Check - .15 > .15
Slow Corner Speed - 65 > 65
KeyPad Ramp Rate - .2 > .2
Small Circle Def (OBS) - .25 > .25

I also ran a ground wire to ground the machine and dust hose as mentioned here (http://shopbotwiki.com/index.php?tit...g_your_ShopBot), to rule out static. (To ShopBot: Given the apparent importance of this step to Bot functionality, not to mention avoidance of sawdust explosions, it might be worth including this step and supplies with the initial installation process).

To my relief, resetting the Move and Ramp Speeds to the above defaults definitely took all the heavy jarring and jerking motion out of the cutting that I had experienced on my first two cut file trials (the 2D shopbot logo sample file, and the 3D leaf bowl sample file - both of which milled perfectly, even if the motion was jarring)....

...But resetting the Move and Ramp Speeds didn't solve the problem of the bit diving into the chair leg tabs or chair legs themselves. After reading more in the forum, I don't think I was losing Z, since most people speak of this term in the sense of the Z gradually slipping down. For me, the bit was occasionally ignoring movement commands in the Z and driving straight along the X axis (my tooling paths are oriented along the X axis), meaning that it would occasionally eat straight through my tabs or my part, which ever was in the path of the bit at the time. After eating it's way straight through (these gouges ran anywhere from a half-inch to 6 inches long), the bit would then pull back into correct position and continue as if nothing happened. Does this sound like a case of "missing steps"? When I re-ran the file on top of the botched part, the bit would make the same mistakes, gouging through the same places (So I guess that rules out static).

As a test, I reran the Leaf Bowl sample file, reducing the feed rates even further just to be extra safe, and although my first attempt at the Leaf Bowl worked fine, even if it was jarring, this second attempt at the Leaf Bowl retained the chair-leg problem of occasional gouging along the X. I've attached screen grabs of my settings, and will follow up with images of the the cut preview, actual cut results, and the Cut3D .v3D file itself (if you load the .v3d just make sure the bit info matches mine). Hopefully since this Leaf is a file everyone is familiar with, it will be easier to troubleshoot.

All help is tremendously appreciated, because I'm at a loss at this point.

[chairguy]

Here are those images of the cut preview and actual cut results.

[chairguy]

Hmmm... anyone know a good way to share a 10.4 Mb Cut3D .v3D file?

[ron_moorehead]

Drop Box works for sharing large files