First successful indexed 4axis project

Not much to say other than it took me a really long time do dial in the rotary and write the CAD and CAM.

The outcome is pretty stunning though.

This is a four sided cut with automatic rotation. There is only one manual tool change.

Software: Deskproto

Roughing 1/4 end mill, doc 2mm step over 2mm, speed 10000mm/min.
Yes, the 1F is screaming, so roughing takes about 30 minutes.

Finish with a R 0.5mm tapered ball end step over 0.1mm feed rate 10000mm/min which might be a little on the high side.

Takes about 4 hours to finish.



5 Likes

Nice! DeskProto is great for rotary stuff.

Which method are you using to set the zero point for the rotary axis? That’s the tricky part, since being off by a fraction of a mm really stands out on indexed rotary jobs.

1 Like

I had a 1/4 inch rod in the spindle and used the touch probe. The chuck is precisely 100mm in diameter.

I use the G59 coordinate system for the rotary.

I modified the Deskproto post processor so that it will use the g59 command at the beginning of each cut which was quite easy.

I rarely measured X, Y and Z0. The Masso just remembers the work home. Of course this only works with the tool setter.

I remember that was a nightmare with the buildbotics controller.

My rotary is permanently installed just off the waste board, only the chuck is inside.

2 Likes

Cracking job.

I’ve had one go at 4th-axis indexed milling. It’s tricky to find the centre, and I was still off by a tiny fraction in my attempt.

I had two 1/4 inch rods, one in the spindle and the other in the chuck, but I still had to “manually adjust” the calculated values to cut a cube from four sides. I roughed with a 1/8th bit.

What was your method for "dialling in " the centre?

FYI: I used f360, which wasn’t too tricky. Using previous set-up stock and duplicating setups for the other sides was relatively easy, contrary to my expectations, and meant that I limited the re-cutting of previously cut areas.

I had a long rod in the chuck when I installed the rotary on the waste board to have it point in the right direction parallel to the y axis.

I think there best way to center is to use a rod in the spindle and touch of with the masso probe cycle on the chuck. Yes, you can also put a rod in the chuck.

The rotary is attached and has to be only probed once, and the coordinates should be in the masso forever.

The live end is far more difficult to center. I have to readjust and check it every time I move it. I use the rod in the spindle and micro adjust the tailstock.

1 Like

Yes, Fusion should be great for this too.
I am still curious about their pricing, do you have the free version, and is that limited or the subscription and how much is that per month?

I have the free version of F360.

It is restricted. The biggest restriction, for me, is the lack of rapids at rapid speed. They are throttled back to, I think, the last feed rate. It is a problem with large jobs that have fine detail spread over the job (e.g. rest machining).

For rotary milling i use vcarve.

2 Likes

@AndyP have you seen Tim Paterson’s GitHub project which restores rapids?

Fixes tool change restriction in Fusion for Personal Use. Also restores rapid moves the free version of Fusion limits to feed-rate moves. See section below.

2 Likes

Interesting. I remember researching this a while back and concluded from others attempts that it was too tricky. The f360 add-on you have identified above looks more promising, but it still claims to be experimental. I will give it a go. Thank you for the post

1 Like

Finding true center of 4th axis is really simple and easy, I do mine with an indicator, sweep centerline with indicator to get close on your 1/4 rod, then to fine tune it, position 4th at A0. And indicate one side of centerline then raise up in z and rotate indicator to opposing side and then rotate A axis 180 deg and sweep that side. Split the diff in axis. Should be in within .0001 or .0002” from centerline. Thats how I have done it for over 30 years. But that is true center of rotation.

3 Likes

I for got to mention, that when you rotate 180 deg, you need to lower z back to same z you had before, that is really important.

3 Likes

Wow. I think that is a spectacular result. Well done on the programming.

1 Like