Cuts have steps instead of being smooth

6128 Views 25 Replies 9 Participants Last post by difalkner, Dec 1, 2017

difalkner

Discussion Starter · Nov 29, 2017

8

This is something I have noticed in some jobs but not others; the side wall of the cut should be straight but it has steps in some areas, smooth in others. It doesn't happen on every job I cut and on some it doesn't matter because there are subsequent operations that will remove the steps but it shouldn't be there.

I'll tell you what I know -
1) drive belts are tight
2) the machine is lubed sufficiently
3) there is no visible change in motion or jerky movement
4) the simulation in Fusion 360 doesn't show this stepping effect
5) nothing on the machine appears to be loose
6) the DC hose is grounded at the DC
7) there is no audible change when this happens
8) has happened with both downcut and upcut spiral 1/4" bits
9) using Mach4 although not the latest release, it's probably 6 months old and I need to update it, and using ESS
10) the jobs start and finish as designed in Fusion 360, no interruption or appearance of missing steps and being off from that point forward. This one is a 2D Contour profile.
11) spindle speed is 18,000 and feed is from 100 ipm to 150 ipm with depth of cut 0.125" to 0.150" in hardwood
12) the bit is tight in the collet and the work piece is well locked down, no movement at all

This is what I was cutting today, B-52 tails for a trophy shop, 2 up per board of Walnut, cutting time is about 6 minutes, 4 boards and 8 tails total, and the stepped area of cuts is in the same place on each one although more so on the first than the last -

Wood Wood stain Hardwood Table Tool

The red area cuts are very smooth and the green area cuts are stepped -

Wood Line Hardwood Plywood

Wood Arrow Electrical wiring Wire Bow and arrow

Right after that I cut the antennas, 10-up, and two of the bases; cuts are smooth except for some fiber tear but that's ok because these get completely rounded over -

Floor Wood Auto part Hardwood

Wood Hardwood Table Furniture Floor

Wood Hardwood Table Lumber Furniture

Here's the finished piece so you'll get an idea what I'm building out -

Hardwood Wood Building Furniture Inventory

Here's another example from a couple of week ago, the other side of the letters is very smooth - I believe this is 2D Contour, as well. I don't recall if this has happened on another profile such as Pocket or Adaptive clearing -

Ideas?

Thanks!
David

See less See more

1 - 20 of 26 Posts

DaninVan

· #2 · Nov 29, 2017

Phew! You had me worried, David; I thought you were making the whole damn plane!!
Sorry, I can't help you with the problem, although from your sample it looks like it's (always?) a gentle transition into the curve and then it returns...not a sudden right angle departure. But you already noticed that

difalkner

Discussion Starter · #3 · Nov 29, 2017

DaninVan said:
Phew! You had me worried, David; I thought you were making the whole damn plane!!
Sorry, I can't help you with the problem, although from your sample it looks like it's (always?) a gentle transition into the curve and then it returns...not a sudden right angle departure. But you already noticed that

Haha! That would definitely be an undertaking but if they want to pay me to make a Walnut B-52 then I'm game. :wink:

David

Reactions: 2

4DThinker

· #4 · Nov 29, 2017

2 reason for steps... Your spindle isn't mounted perfectly vertical, or your steppers are slipping a little during a cut. Slipping can be a loose couple between drive shaft and stepper shaft. A gear can be slipping on the axle it is mounted on.

Sometimes a bit will have a slight deflection of the cutting tip, leaving a line at each depth it does a pass at.

4D

difalkner

Discussion Starter · #5 · Nov 29, 2017

4DThinker said:
2 reason for steps... Your spindle isn't mounted perfectly vertical, or your steppers are slipping a little during a cut. Slipping can be a loose couple between drive shaft and stepper shaft. A gear can be slipping on the axle it is mounted on.

Sometimes a bit will have a slight deflection of the cutting tip, leaving a line at each depth it does a pass at.

4D

I checked vertical a few days ago and it seems to be good. If it was off, wouldn't this 'step cut' show everywhere? Also, if the bit had a slight deflection why would it cut the steps in the same place on all 4 boards? If it was leaving a line I probably wouldn't have mentioned this but what I'm seeing is a definite step, well past a line. When I do a final 2D Contour profile to cut a piece out the walls of the cut are very smooth.

It doesn't matter if I use an upcut or downcut bit or whether the feed is 75 ipm or 150 ipm, most of my cuts are very smooth but a small percentage have this step.

I checked all the drive gears and pulleys - nothing is loose. I used Loctite when I assembled everything and they're all still good and locked down.

With power off and moving the gantry by hand I can't discern any backlash in the R&P on the Y axis and none on the X axis. Z is the only one with a flexible drive coupling but it isn't a factor in what we're seeing.

David

MEBCWD

· #6 · Nov 29, 2017

looking at the last picture I would say it was a software glitch. Toolpath generator or post processor. The steps do not follow the profile, which would indicate an alignment problem, they are each a different path which to me says software problem.

Does this happen every time you run the same gcode file? If it happens every time it is most likely a software problem. If it is different each time then I would think it is a mechanical problem.

Material choices might also cause some differences.

difalkner

Discussion Starter · #7 · Nov 29, 2017

4

MEBCWD said:
looking at the last picture I would say it was a software glitch. Toolpath generator or post processor. The steps do not follow the profile, which would indicate an alignment problem, they are each a different path which to me says software problem.

Does this happen every time you run the same gcode file? If it happens every time it is most likely a software problem. If it is different each time then I would think it is a mechanical problem.

Material choices might also cause some differences.

I just generated the code yesterday, Mike, though I have several earlier versions of the same file (I keep making minor changes each time I make tails). Pretty sure it has done this before, though. I have seen it in Walnut, Cherry, and Maple and cutting on different places on the machine.

Here are some more photos after taking the tails out of the board -

Pretty obvious when placed side by side -

Tire Automotive tire Automotive wheel system Wood

You can see how it's smooth coming down from the top but starts right after the curve of the leading edge of the tail -

Tire Wood Automotive tire Hardwood Automotive wheel system

Waste side of the good cuts -

Wood Plywood Wood stain Hardwood Roof

Waste side of the stepped cuts -

Wood Wood stain Hardwood Wall Plywood

David

See less See more

MEBCWD

· #8 · Nov 29, 2017

2

Since you are generating new toolpaths each time it would be hard to pinpoint it as a software problem.

In the first picture could be difference in grain of the boards.

Second picture there is a change in grain direction between the top and facing side.

Third picture tells me that if this side of the cut is stepped and the other side of the cut has no steps then it is an alignment problem (see pics)

Forth picture I would say software or missing step from over feeding.

See less See more

difalkner

Discussion Starter · #9 · Nov 29, 2017

Wow, Mike! I appreciate the detail and explanation. It makes sense to me and was my first thought but when I checked the machine this past weekend I read it as perpendicular. To be clear, I have used the toolpath several times before tweaking it and making a minor change, so they aren't a 'use once and throw away' item.

On to alignment... how best to ensure the Z axis is completely perpendicular to the work surface spoilboard? What method do you use? Or if you can point me to something on YouTube or another site then I'll study what they have to say.

Thanks so much!
David

SplinteredDave

· #10 · Nov 29, 2017 (Edited)

Dave, I used a method that I borrowed from several Youtube videos... here a few that I used. All basically the same process. I purchased the mentioned dial indicator but I've seen other methods employed that are equally if not more effective.

Reactions: 1

MEBCWD

· #11 · Nov 29, 2017

David the gauges in the videos Dave points to work great and that is what I have ( the gadget guy way) or you can use the bent wire method like they suggest for drill presses.

difalkner

Discussion Starter · #12 · Nov 29, 2017

Thanks, Dave! I watched the first one and am seeing what I have that I can use since I don't have the tramming setup. I'll check in later...

I use the bent wire method on my drill press, Mike, but figured I want this a bit closer.

David

difalkner

Discussion Starter · #13 · Nov 29, 2017

This is what I came up with for tram setup. I think it should work ok once I get the glass level. Gotta' work on some songs for Praise Band practice so that's as far as I got but maybe when I get home, or at least first thing in the morning, I'll have a go at this.

Floor Wood Machine Hardwood Flooring

David

See less See more

Reactions: 2

CAD-Man

· #14 · Nov 29, 2017

I do not have the expertise of how your CAD system works but here is my question. Does you CAD software generate an output file that you then feed to you CNC system? If that is how it works, then the next question. Do you use that same code as a saved file each time you run the CNC? If that is the case you might try generating a new output file and feed that into the CNC.

What I'm driving at is there a possibility that your saved file has become corrupted and that's why the error is repeatable.

Sorry if I'm barking up the wrong tree. It's the 20 plus years in the CAD/CAE world.

CAD-Man

SteveMI

· #15 · Nov 30, 2017

I would consider what was said earlier about grain direction. Quick look at the parts and waste seem to indicate the poor result is happening on one side of the cut.

Three things I have found on problem woods and CNC are (1) use climb cutting, (2) do a "last pass" at full depth for final .004" to remove the lines and (3) clean the end mill. Cleaning the end mills and bits have proven to fix a lot, gunk on the cutter can cause all kinds of marks.

You have tried different ipm, but that would be my #4 thing to look at. Some woods like to be cut fast and others slow, actually different boards of the same species will act up at times.

I've even had to add a "lead in" on softer woods (walnut /cherry) to not have the end mill evidence of moving to depth.

Steve.

Reactions: 1

difalkner

Discussion Starter · #16 · Nov 30, 2017

CAD-Man said:
I do not have the expertise of how your CAD system works but here is my question. Does you CAD software generate an output file that you then feed to you CNC system? If that is how it works, then the next question. Do you use that same code as a saved file each time you run the CNC? If that is the case you might try generating a new output file and feed that into the CNC.

What I'm driving at is there a possibility that your saved file has become corrupted and that's why the error is repeatable.

Sorry if I'm barking up the wrong tree. It's the 20 plus years in the CAD/CAE world.

CAD-Man

No, Bruce, that's a good tree to bark up generally. In this case, though, I have regenerated the code many times. I do these about once every 6 weeks or so and I have usually learned something new in Fusion 360 that will allow me to cut faster, smoother, cleaner, etc. so I edit the CAD or CAM and create a new G-code.

When I first started cutting these I was cutting one tail in 12 minutes and wasting a lot of valuable Walnut. Fast forward a year and I'm cutting them two-up, wasting very little wood, and the total cutting time is 6 minutes for two - a quad-fold increase! Not that I'm in a production shop but hey, if you can speed things up by that amount without sacrificing quality then you're leagues ahead, present step-cut issue aside.

Thank you!

SteveMI said:
I would consider what was said earlier about grain direction. Quick look at the parts and waste seem to indicate the poor result is happening on one side of the cut.

Three things I have found on problem woods and CNC are (1) use climb cutting, (2) do a "last pass" at full depth for final .004" to remove the lines and (3) clean the end mill. Cleaning the end mills and bits have proven to fix a lot, gunk on the cutter can cause all kinds of marks.

You have tried different ipm, but that would be my #4 thing to look at. Some woods like to be cut fast and others slow, actually different boards of the same species will act up at times.

I've even had to add a "lead in" on softer woods (walnut /cherry) to not have the end mill evidence of moving to depth.

Steve.

1) Climb cutting is how this is cutting, 2) I don't do this but really should, 3) It's very clean.

In the year we have had this CNC I have cut hardwoods probably 95% of the time - Walnut, Maple, Cherry - and have cut some MDF and Pine one time. The hardwoods I'm cutting don't leave any gunk on the bit. It's interesting that you refer to Walnut as soft; I've always thought of it as one of the harder domestic woods. I almost always have a ramp cut to start my cuts into these woods, unless it's a small pocket and then I usually do a helix entry.

Thanks, Steve!
David

See less See more

tacomamacxtech

· #17 · Nov 30, 2017

Try cutting dense foam like they do at Boeing to prove out parts! They use tan foam made specifically for this purpose. You can get dense blue insulation foam at the "big box" stores in different thicknesses!

MEBCWD

· #18 · Nov 30, 2017

That looks like it should work fine David.

Now to throw a curve into the problem. If the guide rails or racks are not perfectly straight and aligned then you could have a spindle that is 90 degrees to the bed in the center but off 90 degrees where the rail or rack varies.

See less See more

difalkner

Discussion Starter · #19 · Nov 30, 2017

I'll only check it in the first 18" or so and will blindly assume the rest is perfect... :grin:

The really good thing is that I just talked to Nate at FLA and we both realized that when he sent me the upgrades last month he included a tramming mount for the spindle! I just have had zero time to take the machine down so I had not even looked in the box but today I am shutting everything down to install the tramming plate and get this settled so I can move on - got lots of work to do.

David

beltramidave

· #20 · Nov 30, 2017

difalkner said:
I'll only check it in the first 18" or so and will blindly assume the rest is perfect... :grin:

The really good thing is that I just talked to Nate at FLA and we both realized that when he sent me the upgrades last month he included a tramming mount for the spindle! I just have had zero time to take the machine down so I had not even looked in the box but today I am shutting everything down to install the tramming plate and get this settled so I can move on - got lots of work to do.

David

Wow, How do you rate? Nate didn't send me the tramming mount with my upgrades...

1 - 20 of 26 Posts

This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.