Please give me some Probe advice

[COBOB]

I recently got the ShopBot probe. Things are not as good as they could be. The probe seems to be working correctly. Number 1 Input is glowing green, and goes out when you move the tip. The scans seem to be OK except for the lines running thru it. I am assuming these are stepover related. I am using the 1/8" tip and left the stepover values for X and Y as they came from the factory, 0.01"
Is this my problem. I am running another scan right now, using 0.0625" stepover.

If anyone can look at the attached picture and give me some hints / answers I sure would appreciate it.

Thanks,

Bob

[myxpykalix]

im guessing since i haven't had a probe for years now. Do you have a tip that comes to a point like a needle? I don't recall if one came with the probe or not.

I would try that since the areas that the lines run thru are consistant my therory is that the probe is too big to get into that area and creating that void?:confused:

I would use a smaller tip and try a sample area in the same place it currently isn't scanning and see if that changes....just a guess:)

[COBOB]

Hi Jack, and thanks for the input. I have 2mm, 1/8 and 1/4 inch tips. I tried the 2mm and it was just as bad. I can't figure out why it is happening in the X axis and not the Y. When I changed the stepover to 0.0625 the lines were no longer there. Of course the resolution suffered. Still a mystery here.
Bob

[paul_z]

Bob,

If the lines are in the same directions as the probe is moving then I suspect backlash or worn pinion gear in the Z axis may be your problem.

Paul Z

[COBOB]

Paul, thanks for that. Machine is so new I didn't consider that. I will look into it though.

[myxpykalix]

paul the reason i didn't think of worn gears was because the lines are so straight and uniform where i would think if it was worn gears that it would not show up so straight and uniform of lines but be straggily lines(?)

[paul_z]

I still think worn gears are a possibility and should be checked occasionally anyway. Dirt packed into the gears or rack is also likely to cause problems.

What actually happens during the probing is quite complicated. The controller commands a plunge at a fixed speed. The probe touches the surface and causes an open circuit which the controller interprets as a touch. The controller then issues a stop and reads the current position. Because the controller is not infinitely fast, the current position is actually below where probe touched the piece that is being digitized. This actually is not an issue if the machines downward travel is at a perfectly constant velocity. Even though the position is off a few thousands, all positions would be off by the same amount. Due to vibration, varying drag along the z travel, (and God only knows what else), the reported position varies some. If you digitize the same object twice and subtract the point clouds from each other, you can see this variance. Slowing down the vertical speed would help but the digitizing process is already painfully slow.

Hmmmm … I wonder if the software could be modified to wait for the contact to happen, stop, move up at a very slow speed waiting for the contact to clear, then order a stop and read the position. This should not add much time to the overall process since the travel distance would be extremely short. It may result in less error in the point cloud.

I’m a bit too busy to try this right now. Any takers????

[COBOB]

Those lines appeared at a 0.01 stepover. Once I scanned as 0.0625 they disappeared. When I scanned at 0.03125 they were also gone. But at the higher stepover numbers the resolution was rather poor. If it was the pinion or slop, wouldn't you think they would still show? I don't know. I am asking, and surely appreciate anything you guys throw out there.
Bob

Tonight when I get home I will go back to the 2mm tip and decrease the stepover. If it works the same way as cutting tools one would think that would fix it, if it is indeed a stepover issue. I am surprised nobody has seen this before. Why me? GRIN!

[Brady Watson]

I've done my fair share of probing & I've found that the lowest usable step resolution is about 0.025" with a 1/16" stylus. Keep in mind that ramping is turned off during probing, and this can sometimes cause jittery results if the points are too close.

I'm not saying that .025" is the smallest that the machine can resolve, just that, with the factory probe speed settings, it is the smallest reliable step res that I would go. If you reduced probing move speed lower, you may get better results.

If you get totally frustrated & strung out with the probe, send me a PM. I can laser scan it for you for a reasonable fee or barter...

-B

[COBOB]

Thanks for the info Brady. I will keep all of that in mind.
I appreciate the response.
Bob