[CNC-Joe]

Hi Group,
- I've been wanting to implement a rotary (A) axis on my milling machine for a while.
- I actually have already done this with a 5C Spin fixture as a workholding device. ( CNC-Joe's Workshop - Poor Man's CNC Rotary 4th Axis Project Page: )
- Basically - I figured out the gear reduction, and the number of steps for rotation (its been a while).
- I checked its accuracy (sorta) by mounting a protractor and a piece of paper with a line through it, and would have it rotate once, then 10x, 100x, 1000x for checking accumulated error. (It was fairly dead on after 1000 rotations).

- Thinking back on doing this again (as I have, of course, torn the set up down in between using it years ago.....) - IS THERE A BETTER WAY of calibrating the rotational axis?
- I'm thinking of mounting a bar and attaching a digital angular gauge (similar to the Wixey) - mount it via magnetic mounting - zero the gauge, and set it through its paces.

Wondering if anyone else has experience with calibrating rotational axis - and how they did it.

[CitizenOfDreams]

You know how many teeth your gears and pulleys have. You know how many microsteps per turn your motor has. Just calculate the "step per" value, there is no need to "calibrate" anything. A full circle is always 360.000000000 degrees, never 360.1 or 359.9.

[cncuser1]

CNC-Joe: How much error are you getting after 1000 revolutions? For the type of work you are doing, is it acceptable? My guess is you're good.

But after looking at the picture I have another comment. You don't have a big gear reduction. So that means the holding torque of the motor will be used to its maximum. You face the possibility that if you push too hard on the work piece ( with the cutter) the motor will be back driven. If you are doing small work close to the centre of rotation you should be OK. If your work is bigger and further away from the centre of rotation you might get into trouble.

If you haven't already done so, do some test cuts to find the limits of the 4th axis holding torque. Better to do it on a test then on real work.

[CNC-Joe]

Thanks for the input, group. I made the "Poor Man's rotary 4th axis" out of parts that I already had laying around (ie a "no cost" weekend budget project).
- It is limited in holding torque - but I mainly intended to use it for engraving purposes and not really 4th axis machining.
- It could definitely use a good 60:1 or 90:1 rotary table for more holding torque and to reduce the chance of it being back driven.
- After 1000 rotations it was off a small amount...
- I'm going to get it hooked up again and put my digital angle gauge (a wixey knock off clone) and see how it holds up using that as "reference".
- I think that a better way of doing this would be to use a dial indicator over the top of a ground 1-2-3 block... with a ground parallel on top (or something similar).
- zero it across the block.. move the block out of the way... rotate.. move the block back in and measure...

I'm just trying to find a good way to calibrate it to "zero".

[cncuser1]

Joe: Doing 1000revolution is different than getting a zero level across a flat.

Doing 1000 revolutions is to calibrate the steps per revolution in the control software.

Running an indicator across a flat to get a zero degree position is for a setup or measure a part. I think they are for two very different purposes, but I'm making some assumptions about your goals and procedures.

And when you mount the unit to the machine take great care to get the axis of rotation of the 4th axis parrallel and perpendicular to the relevant axiis.

a couple of degrees off parallel and perpendicular and the engraving tip (or mill) will be WAY off.

[CNC-Joe]

-Doing 1000's of revolutions - shows you the accumulated error. Eliminate the accumulated error - you eliminate that portion of the problem.
- Running an indicator across will tell me quantitatively when that is zeroed out, as well.
- Agree that any amount of parallel and perpendicularity that is out - will only increase the errors in the finished product.