Need help on a shuttering issue.

[knight_toolworks]

I have had a bit of shuttering on a part of a curve all over the table. if you have a circle in front of you it is always on the bottom left and some on the bottom right.
well I had cut a few of these parts they are all 1" oak one cut per board same location on the table. well the shuttering got far worse today right off the bat. Only on the finish pass with only .01 cut depth and whats left from the climb cuts. I am cutting at 3ips 8k with a 1/4" downcut bit it is pretty fresh with maybe 30 minutes cutting on it. communication speed is 82 it was shuttering even when ramping in in that area and pretty much not cutting anything.
http://i154.photobucket.com/albums/s266/knighttoolworks/posting/8d91d220.jpg

[cnc_works]

Sounds like it might have something to do with x axis and y axis handing off to one another.

[knight_toolworks]

how would you deal with that? called shopbot and they think it is mechanical since it is so repeatable.
I checked the pinions and they are tight. They may need to be replaced though.
I slowed the final pass down to 1ips and it is much better. I will replace the pinions as soon as I can get them.

[richards]

Steve, This is an interesting problem that has a lot more to do with way a CNC router cuts a circle than the mechanics of the CNC router.

Make a spreadsheet showing how many steps the X and the Y axis has to make when cutting a circle. (Remember that a circle consists of straight-line segments.)

A stepper motor cannot "split" steps. It can either make a step(s) or it cannot. If the math tells an axis to move 1/3rd step, you can only hope that the CAM software compensates.

So, dig out those old "trig" books and do some computations. An Alpha normally moves 1,000 steps per revolution, so an Alpha machine with 1.5" diameter pinion gears and 7.2:1 gearboxes would move an axis 0.00065" per step. Cutting a circle requires that the minor axis move a multiple of 0.00065" per "segment". If your spreadsheet is similar to mine, you'll probably find that the CAM software is dividing the circle into way too many segments.

Even if you tell the software to make "coarser" cuts, which would seem to indicate that the "studdering" or "chatter" would increase, there is another problem that as even a higher impact on quality. That problem is resonance. At low speeds, stepper motors tend to "resonate" or "vibrate". That vibration tends to force the micro-stepping of the Alpha drives (or the Gecko drives) to be nullified and the stepper seeks the "full-step" detent (or position). The net effect is that the cut is 10X more coarse than expected.

I know of no way to eliminate the problem. I literally spent thousands of dollars chasing "chatter" when I had my PRT-Alpha. Changing to belt-drives (and 7.2:1 gear boxes) helped, but the "chatter" was still there. Using servo motors adds considerably to the cost and to the complexity of the machine (and they have their own "quirks").

The best advice that I can give is to know your machine and your software. Know when you have reached the limits and then get out the sandpaper ad smooth those "chatter" marks.

[bruce_taylor]

what are you using for hold down?

[knight_toolworks]

the wood was nailed down. I was thinking it was ratting around but I watched the machine shuttering.
I know there will be some shuddering but this is past excessive and just started.

[gc3]

vectors???

[knight_toolworks]

vectors???
no just a smooth arc. I cut several they were fine Tuesday started today and there was the shuttering.

[knight_toolworks]

well replaced all the pinions why my new spoilboard was vacuuming down. Isjut did a test cut in some bamboo plywood using a 1/2" bit at 3ips and .7 depth of cut and only minimum chatter.

[Brady Watson]

Make sure you lube the racks or the pinions will be trashed in about 2 weeks...

-B

[knight_toolworks]

I lube them once a month. the left side of the machine can go months if I let it compared to the right side. Still got to battle the z when I have the spare time.