Problems with small circles

[jbworden]

I'm having a problem cutting a series of 1.75" circles out of 1/4" clear acrylic. The problem is that they are not round. It looks like the center point shifts slightly about halfway thru the cut so the outer perimeter is not smooth.

I'm cutting 13 of the silly little things all of them seem to have similar problems. I generated the toolpath file in Part Wizard by drawing a small circle and making 12 copies of it. I then make a shallow cut (1/8") on each circle with a 90 degree V-bit on the line before making a thru cut on the outside of the circle with a 1/4" end mill. That should give me a disk with a slight chamfered edge.

Once these are done I need to cut 8 larger 4.25" circles using the same techniques.

Anybody have any ideas about what's going wrong?

[Brady Watson]

You are either running the machine too fast (non alpha) or have a loose Y pinion. My money is on the pinion.

-B

[terryd]

I agree with Brady, a close second would be only one of the X pinions is loose. Been there, done that, a few less hairs for my efforts.

TerryD

[jbworden]

So how do I fix this? This is a PRT96.

I slowed the cutting speed down to 10 inches per minute. No joy. I actually screwed the acrylic down to the spoilboard to make sure it wasn't being scrubbed around by the dust skirt. No joy. I checked each of the pinion gears (on the motor shaft?) and all look okay (that is to say none of them had big neon signs pointing to them saying "I'm all screwed up".) They appear to be seated properly in the racks and all appear to be held tightly in place by the spring/turnbuckle.

In every case, no matter if I'm cutting one or many circles, it appears that a line has been drawn from the 1 o'clock to the 7 o'clock position (with the x axis running from 9 to 3 and Y running from 12 to 6). The half circle on the left side of the line appears to be shifted in the negative Y direction along the line by about 1/32 inch. It appears to be consistent even thru the bit change. The 1/4" end mill gives me the exact same shift in exactly the same positions. It would seem to me that if this were a loose gizmo somewhere the error would not be so consistent. The path in Part Wizard looks smooth and running the toolpath in preview mode look smooth.

How do I check the pinions to see if they are loose?

[Brady Watson]

Jim,
The best way is to *gently* push on the Y car with the control box on to feel if there is any play in the shaft & pinion. Don't ram it....just a deliberate push & pull to --> Feel <-- for any play. The main culprits here are the hex grub/set screws that act as sort of a keyway when you tighten them down to the motor shaft. You REALLY NEED TO CRANK THEM ON TIGHT to keep them from walking loose. I JUST had this happen to me yesterday...I was replacing pinions on my PRT and felt that it was cranked on pretty tight. I ran a 3D file and it was really squirrely looking at the end of each pass. I left the power on, pushed the Y car & sure enough, it moved a good 1/2-3/4". The set screw had walked it's way out. So to make a long story short, I went to my local Ace Hardware and bought some new 1/4-20 set screws and a brand new hex wrench and some loc-tite. I put a drop of blue loc-tite on it & I cranked the grub screws on so hard that they bent the wrench - which is exactly the right torque.

If you are in doubt, pull the Y motor down and really have a look at what's going on. You'll need a 9/16" wrench and a 9/16" deep socket to pull off the motor/mount. Then set it down (leave the power on), grab a rag and try to turn the pinion. When the power is on you will not be able to move the pinion gear any measureable distance.

Inspect the pinion for wear. If you have not replaced the pinion since you have bought the tool, order up a set & replace them...your cut quality will improve greatly. AND be sure to grease them...otherwise they'll be worn out in less than a month of normal use.

-B

[jdgrahamwaldorf]

Have you checked that your router is attached tightly? It may be a loose clamp since the error is allong the X axis.

[jbworden]

You guys are very good. I checked the Y car and there did not appear to be any movement at all. But when I actually put a hex wrench into the set screw I managed to give it almost half a turn. The problem has gone away.

Thanks to everyone.

Jim