Help with egg shaped circles

[Trucks]

Hey guys, well got my machine running, and i cut a few parts that were not circles and everything seemed ok. I am cutting some circles out of pink foam, and on one side its 3.01 and then if you measure around it goes down to 2.95 in one spot or side.

I am running antibacklash nuts on 1/2" acme rods, with .500" per rev travel, using 387oz motors, with a Gecko 540 drive, rapids are 120 ipm, and cutting speed is 50ipm in Mach 3. I am using a .25" spiral upcut bit, at about 18k rpm, which is the lowest my router will do.

Anyone have any suggestions? Could i just be seeing a little racking on the X axis? My y axis is super solid, and my x has a slight wiggle. Or am i just asking too much for the resolution of my machine? I should be able to get a decent circle shouldn't I??

Thanks, Andy

[ger21]

Which axis is the 3.01, and which is the 2.95? Can you post a pic?

It could be that the screws are slightly different, and you can adjust it by changing the steps/inch.

[Trucks]

Ok, here is a picture.

The X is the front of the machine, and the X with the dots is the back, or the home position, and the Y is the home position in that axis. As you can see the X and Y directions are pretty close, but the egg shape happens basically in the corner of the cut. In the program the bit is plunging where the Y is and going clockwise around the circle, i tried several different feed rates and seemed like it did not do anything.



Thanks for the help guys, its greatly appreciated.

[SlowRiderr]

Originally Posted by Trucks Ok, here is a picture. The X is the front of the machine, and the X with the dots is the back, or the home position, and the Y is the home position in that axis. As you can see the X and Y directions are pretty close, but the egg shape happens basically in the corner of the cut. In the program the bit is plunging where the Y is and going clockwise around the circle, i tried several different feed rates and seemed like it did not do anything. Thanks for the help guys, its greatly appreciated.

AFter you tune the motors, adjust for backlash compensati in mach3 as hoss does on youtube.com

[arizonavideo]

I would check the antibacklash nuts for well... backlash.

I bought 4 -1/2" 2 start 5 TPI from Dumpster and they all had backlash. I replaced them with some from Roton and now all is better and I can hold about 3 thou.

[ger21]

Are you using a single G2/G3 for the hole circle, or is it broken into smaller segments? How slow have you tried to cut it? Are you sure the X and Y axis are perpendicular to each other? If you cut a square, is it square?

[Trucks]

Well to start off with the sad news, i tried to cut a square, and it definately is NOT SQUARE!

So i took a measuring tape, and sure enough when i move the Y-axis across the table it measures out differently between the bit and the end of the table, by about 3/8" so i moved all that and cut another circle, and now it is really close to being square.

I will try and cut another circle and double check everything, as well i am going to check to make sure that when mach 3 is moving 1 inch, the router is moving exactly 1 inch.

Thanks for the help guys

Andy

[Korellibopper]

check out the smooth stepper If your already sure its not a mechanical problem. I think its smooth stepper .com

[CarveOne]

Originally Posted by Trucks Well to start off with the sad news, i tried to cut a square, and it definately is NOT SQUARE! So i took a measuring tape, and sure enough when i move the Y-axis across the table it measures out differently between the bit and the end of the table, by about 3/8" so i moved all that and cut another circle, and now it is really close to being square. I will try and cut another circle and double check everything, as well i am going to check to make sure that when mach 3 is moving 1 inch, the router is moving exactly 1 inch. Thanks for the help guys Andy

That sounds to me like a steps per inch (or mm) setting in Mach3 is not correct. It is easier to measure the error on a 6" or 12" square than a 1" square (or circle). You can get a finer resolution in your measurements. All other things being equal, the "steps per" should be the same number for each axis. The only time they should be different is when compensating for different ratios in lead or ball screws.

CarveOne

[ger21]

I think what he is saying, is that the Y axis is out of square by 3/8" from one end to the other.

It is possible to have Mach3 compensate as the Y moves across the table. If you can't get it square, I can tell you how to do it.

[Trucks]

Gerry,

how would you change that in mach 3? I am still off by about 1/32" or maybe a little more between the 28 inches of travel i have right now.

I am also thinking of purchasing some linear rail off ebay, and redoing my gantry.

Thanks, Andy

[ger21]

You use the formula functions. Are you saying the X changes as you move along the Y?