Are you getting similar steps in your edges?

[Mayo]

Most of the work I do is of a temporary nature, so the edge quality doesn't become an issue. I recently had another job to make some Corbels out of AZEK and I was less than happy with the edge quality.
The corbels were each made in four pieces like a sandwich, with scrolls on the outer two surfaces.

I cut some of the job with a 1/4" down cut and some with an up cut spiral.
Cuts were done in 4 passes on 3/4 inch material and 5 passes on 1 inch material. I don't remember the exact cutting speed but I got shavings - not dust. I'm thinking it was probably around 2 inches per second at 13,000 rpm. Porter Cable router.
I can check the log files when I get back to that computer.

Hold down was screws in the corners of the sheets and I had two or three tabs holding each part in place. More tabs would have been better but I had 92 pieces to cut from 8 1/2 sheets so...

I replaced the pinions on my 2001 PRT96 last year and 99% of the time I cut EPS foam, so there is little stress on it to wear things out.
But the interesting thing about the steps I get on curves is they are in the same place on the same parts.

So I thought I would post photos of the edge of the parts to see what others think is the cause. Mechanical, part file, hold down, cutting strategy, bit selection, all of the above?

[gc3]

take a look at the nodes on the vectors

[steve_g]

Looks like my "before" photos with my 1:1 alpha drive... the belt reduction boxes made a big improvement.

SG

[chiloquinruss]

I agree with Gene, looks like the design for the corbels is not curves but segments. Where did the file come from? Russ

[gerryv]

The easy stuff:
Are the stepper tension springs that pull the pinions into the racks tight enough?
Are the eccentric V-groove wheels always in full contact with the tracks?
Are the set screws in the pinions really snug on the stepper shafts - I "think" that there are two, the lower one being the actual set screw and the one piggy-backed onto it being to lock the first one.

[Chuck Keysor]

Hello Mayo. If you want to cut one part on my machine to see how it turns out, you would be more than welcome. I haven't had the issues you are showing, so that would provide a simple (but not very brilliant) way to see if there is a problem in your file. Chuck

[cnc_works]

Mayo, I have a 2001 PRT as well. Other than the suggestions for machine tuning that have been made, I can tell you my experience.

I machined a great deal of acrylic disks and always ended up with annoying mill marks as seen in your photos. They were always the worst when the x and y steppers were working together. Mill marks along the x or the y were fairly minimal. Nothing I did would eliminate or even minimize them much.

Finally, I upgraded to a new control box (Ascension in my case) with Gecko drivers. Mill marks improved considerably, but were still there. Now it is just part of my world to know that I either have to deal with them or live with them. Like a lot of relationships I have.

[richards]

If you search for "chatter" on some old posts (about 2007), you'll probably find a photo showing similar chatter on my original 1:1 PRT-Alpha and a comparison photo with a 3:1 belt-drive. Later, I upgraded to the 7.2:1 alpha drives that were offered by Shopbot. Both the motors with a belt-drive and the 7.2:1 alpha motors had minimal "chatter". The "chatter" on the 1:1 drives was unacceptable.

Before building the original belt-drives, I tried every possible software setting, but nothing seemed to do much good. With a belt-drive or with the 7.2:1 alpha motors, the cuts were much smoother and almost identical.

[Mayo]

Firstly, thank you to everyone that has responded.
It seems like the issue was present in both the non Alpha PRT machines but also in the early Alphas as well.

My PRT came with Vexta motors PK296A1A 7.2 2 phase, .25/step with the gearboxes.
The tension springs are at the recommended number of turns for tightness but I could try making them tighter to see if it helps. Based on the responses though, I think it's just a combination of not enough tabs, hold down, cutting strategy, and mechanical limits of the machine.

When I replaced the pinions, there was only one set screw and I tightened them to the point of almost bending the allen wrench.

In the design file the curves look like curves - not steps. I don't think the post processor is converting the curves to straight line segments.

So Chuck, I may take you up on the offer to cut it on your machine for comparison purposes. I'll give you a call Thurs. or Fri. if either are good for you.

[Brady Watson]

Mayo,
A 4G upgrade and 30 tooth pinions would do wonders for your cut quality. I wouldn't crank those turnbuckles more than 3.5-4 turns. More than this and you'll wear out your rails prematurely.

Be sure to give your entire machine a good shake down with the motors energized. Pay attention for any slop or backlash felt or heard during the process. Keep shaking until you pinpoint its location and then make adjustments.

However, your pics appear to be rhythmic and this would be attributed to step resolution more than anything else. If your box is 1/2 stepping with those 7.2s you have the same resolution as a standard 3.6:1 PRT, which may be the case. If this is true, then a 4G upgrade will give you 5 times finer resolution and a nice speed increase.

-B