Yet another weird Z height problem

[beacon14]

This is one I've never seen or heard of. I spent yesterday cutting sample files of an inlay to be put in the center of a dining table. Everything ran fine except on this one file. The file is made of two toolpaths in VCP, combined into one .sbp file, the first one steps down from 0 to .97" in 4 passes, the second starts at .97 and steps down to 2.01" in 5 passes. The first passes ran fine, but the last five passes machine at a depth less than the depth in the file. I can watch the lines in the file and the control screen shows the tool at a higher depth. Oddest part is that the log file shows that the file ran to it's full programmed depth but the hole only measures the depth shown on the screen for the last pass.

I re-ran the file several times in air-cut mode and it always runs the same way. Restarting the computer and control box didn't change anything. Went back to VCP and output just the errant toolpath and it does the same thing. That is the only toolpath that didn't run properly all day. Each time it goes to the same incorrect depth for each pass.

programmed depth ------- actual depth and Z position on control screen

-.2425 ------- -.243
-.485 ------- -.485
-.7275 ------- -.727
-.970000 ------- -.970
-1.1780000 ------- -1.074
-1.386 ------- -1.232
-1.594 ------- -1.447
-1.807 ------- -1.663
-2.010000 ------- -1.863

Not sure I'm explaining this effectively, but the bottom line is that for half the file everything is fine but the second half machines at incorrect depths and the problem is repeatable so not grounding or electrical noise. The actual hole measures 1.86" deep even though the log file shows a minimum Z value of -2.01. The stepped area that machines .97" deep measures correctly.

I'm out of ideas of what to look for - except that I replaced my control computer a week ago after the old one died. The new one is a refurbished Pentium 4, 2.0Ghz with 512M RAM running XP Pro with all options disabled and only SB3.5.3 loaded. Speed test returns 83.4. I also just swapped the router for a spindle.

What's going on here?

[richards]

David,
The first depth on the second part of the code is off by 0.104, the second 0.154, and the rest are off by about 0.145.

Does the Z-axis raise to a safe height between the two sections of code? Does the Z-axis plunge to -0.97 at the start of the second part?

What I'm looking for is the actual instruction that is being used for each pass. When I hand-code I sometimes get caught by the Z-axis height just before a CP or CC instruction. It might be possible that the code generator for VCP uses the wrong logic.

If it's not code, it has to be mechanical or electrical. Let's hope it is the code. That's easier to fix.

[beacon14]

Mike,

I can't see anything in the code that is unusual. It does Z-up between passes and it uses a ramped entry. But the code is nothing more than M3's and J3's. The last section of code includes hundreds of lines all ending in ,-2.01 (Z depth), and the file runs through completion, the log file shows that it went down to -2.01, but watching the control panel the bit never goes below -1.86 and the hole is only 1.86" deep. I'm going to try changing computers tomorrow and see if the new computer has anything to do with it.

[bleeth]

David: Just a wild guess here but check your table limits and file limit checking. This sounds more like SB3 is causing it than anything else.

[wayne_walker]

I am having an issue with my "Z" height. I have a PR upgraded to a PRT with a PC router.

Yesterday I cut some 1/4" ABS completly thru and scored the spoil board.

Today I clamped a fixture onto the table that is about 1 1/4" thick. I clamped my bonded up ABS, 5/8" thick, to the fixture. I "Z" zeroed to the top of the material and attemped to do a cut thru for a handle hole.
It left about a .060" in the bottom of the cut. The "Z" was bouncing at the bottom of the cut.

I have cut this part many, many times without this happening. My machine is well grounded.

Any help is appreciated.

Wayne

[pwbell]

Had a similar problem. While cutting a 3D image, the border just offset itself a distance from where it should have been. After checking everything I discovered that the router had slipped in its holder just so slightly. After re-fitting and tightening, all is good and the 3D graphic cuts properly. As tight as I put the router in its holder, it is hard to believe that it could have moved... Good Luck!

[wayne_walker]

Phillip,

I will check to see that the router is tight in the mount.

Thanks

Wayne

[seana]

Wayne,

Also check the bearing under the stepper motor where it connects to the ball screw.
When i would watch mine i could almost see it stagger then i oiled up the bearing and got it all cleaned up and now it doesn't stagger step. I give it a blast of WD every morning and it seems happy.

O.T. did you ever get the 4g upgrade installed on your machine?

[wayne_walker]

Sean,

My "Z" has a rack & pinion! I checked the gear on the motor and it seems tight!

I just got my new PC set up for the bot, I hope to get it upgraded in the next couple of weeks.

Are you still cutting a lot of bamboo ply?? I never heard from him again.

Thanks

Wayne

[seana]

Wayne,
Blast, i forgot you had the rack and pinion, mine is the old ball screw.

Yea, still cutting bamboo and have set up one of there 2 machines and will set up there 2nd one when i get a chance. I'm doing all the coding for the all the machines keeps me a touch busy.

I haven't switched to Mach yet but i think that will take place in a couple of weeks.
The machine that i have set up for them is running the 4g upgrade, what a difference you will be really happy.

Sean

[wayne_walker]

Sean,

Thanks for the thought anyway.

I hope you don't loose all of the biz when they get their machines running. Maybe you can still do the coding!

Thanks,

Wayne

[wayne_walker]

Phillip and Sean,

Thanks for your help!

I found the problem. It was a small chip of plastic lodged in one of the rollers for the "Z". The "Z" would go all the way up but not all the way down.

I have not been running my DC since I need to make a new shoe. Time to get it done.

Thanks,

Wayne

[seana]

Wayne,
Glad you located the problem, nothing more frustrating as having a problem and not figuring out what the cause is.


in regards to the biz with bamboo, i will be doing all the code work for all the machines. The anticipation and projections are that there will be enough work for all 3 machines cutting. We will see what happens.

Sean