Jump to content

Welcome to eMastercam

Register now to participate in the forums, access the download area, buy Mastercam training materials, post processors and more. This message will be removed once you have signed in.

Use your display name or email address to sign in:

What is studder stepping?


LucasGC
 Share

Recommended Posts

Hi All,

I think I've figured further into why my machine is bouncing around all my corners. My machine has a tangency factor as well as an option to change the acceleration rate. I'm thinking other machines have something similar to a tangency factor to get them to move smoothly from point to point, but they're automatic calculations instead of something you have to change manually. Same with acceleration rate, I would think most machines tell the tool to only accel so much on a short line segment where as I'm thinking my machine (if i have it set at full speed) is trying to go from 0 to 10 to 0 on every line segment. I'm not trying to say my machine is worse than others, just trying to understand what i'm working with. 

Ideally I would be able to make any toolpath with a tangency factor of 1 and g800 acceleration.

I've been told "the only point of tangency is to fix bad geometry" so it's frustrating when I can make good geometry but the tangency messes it up.

This is why the tangency messes up - the lower values are suppose to slow down at the points so you can see it stepping through them, the higher values creates tangent arcs between points, and if you have a really high value creates arcs that skips the point altogether, this can make it unpredictable and sometimes jerks around corners causing bounce. 

The problem I'm having is that with a tangency value of 1 it's studder stepping around a curve that doesn't have very tight tolerance. If I increase the tangency it works smoothly, but jerks around some corners. I lowered it to .1 and expected to see it stepping through points, even if it moved very slowly, but it still shuddered through them. So even with the lowest tangency I can use, the tolerance is too tight.

I don't really understand why because my tolerances are not very tight. The toolpath I'm running is a highspeed pocket toolpath with a .125 corner rounding radius set in the high speed settings. I ran 12 tests, with 4 tolerances, .005, .001, .0005, .0001 and 3 tangency values, .1, 1, 10.

So, with .005 overall tolerance and .1 tang factor on a 2d toolpath, i really expected to see it step from point to point, which it did around the larger curves, but still shuddered around the smallest curve, which i assume is .125. This curve is made up of 10 line segments, none shorter than .009" ~ running at 100ipm 

Link to comment
Share on other sites

Not sure what brand of machine you're running, but most machines need a "high accuracy" mode to control acc./dcc. I run Mazaks at the shop I work at and this is controlled with a G61.1, KX. You should have a similar setting on your machine, code just depends on the brand. This will control the look ahead of the machine and how it handles speeding up and slowing down when the machine sees an arc.

 

If you can name the brand of machine you're working with someone might be able to provide the info you're looking for.

  • Like 1
Link to comment
Share on other sites
2 hours ago, manfowar said:

Not sure what brand of machine you're running, but most machines need a "high accuracy" mode to control acc./dcc. I run Mazaks at the shop I work at and this is controlled with a G61.1, KX. You should have a similar setting on your machine, code just depends on the brand. This will control the look ahead of the machine and how it handles speeding up and slowing down when the machine sees an arc.

 

If you can name the brand of machine you're working with someone might be able to provide the info you're looking for.

Thermwood, and I have had a crazy amount of problems with them but I don't want to say they're bad, I'm sure there are people out there that prefer them.

Or maybe it's a good machine for the price but you don't get as many features as you would with a more expensive machine.

So what makes it a high speed toolpath? If my machine can run scallop with a curve similar to a HSS curve, what makes those different? 

Unless scallop is high speed as well? What finishing toolpaths are not high speed?

Link to comment
Share on other sites

Sorry about the response time, I had to leave work early on Friday and wasn't done thinking about the problem yet.

 

After rereading your original post I figured there might be one other factor. Starvation. There is a machine limitation to how fast it can read code. This problem becomes worse when you have to drip feed. In the shop I work at we have some older Mori Sekis which can't run any of the HSM toolpaths faster then around 40 ipm because it will start to starve the machine and jerk around. This could be an issue if your machine can't read code fast enough.

 

On 8/24/2018 at 10:03 AM, LucasGC said:

So what makes it a high speed toolpath? If my machine can run scallop with a curve similar to a HSS curve, what makes those different? 

A complicated question. The definition of high speed toolpaths differ depending on what person/company you talk to. I personally think of High Speed toolpaths as anything that is taking full advantage of the machine/tool you're using I.e. D.O.C., SFM, Axial/Radial load, chip thinning, etc.

 

However, the definition is a bit convoluted.

 

A lot of the 3D toolpaths are high speed paths (except pocket), at least by Mastercams definition. Most of the 2d paths are not (except the toolpaths with a red and white arrow.) 

 

It sounds like your problem is narrowed down to one of two things. Either your machine is starving, or your post is not posting out the high speed options (if your machine is capable of them.)

 

Fixes:

 

Starvation: If your machine truly can't keep up with the code thrown at it there are a couple of options to help it out. Better settings optimized for longer code segments on your Arc Filter/ Tolerance page. You can slow down the feed rate to the point where the machine can actually keep ahead of the smaller segments of code. If you're running drip feed you could try to find a way to get the program loaded onto the machine. such as breaking the file into several programs.

 

High Accuracy: You can call the manufacturer of the machine and ask them if there is a better setting to control Acc./Dcc. Another bit of code I though of is G5 P2 which enables further look ahead then G61.1. You could also look through the manual to see if you can find any of these settings or maybe settings specifically for your machine.

 

Sorry if this isn't much help as I've never heard of the brand of machine you're working with.

  • Like 1
Link to comment
Share on other sites

Before i upgraded our machines to 2017 hurcos I was getting jerking and straight up bad moves on '97 milltronics controllers. I can jerk my Hurco but thats generally a smoothing default that i missed. Ive run 500ipm no problems. How old is the machine? Also, if its a gantry router that isnt going to help with rigidity

Link to comment
Share on other sites

It acts like it's starving, the geometry in that area just doesn't seem very "small" or like there are too many points. 

It's not 'technically' capable of high speed machining but they said as long as you don't have any small line segments that are consistently too small it should be able to run it.

I think this acceleration is something built in to a lot of machines for HSS purposes while on my machine I have to input it manually. 

I don't know if the tangency is built into other machines or if it's just a "fix" for something. I assume other machines must have some sort of tangency to make them move smoothly through points instead of stepping through all of them, but at this point I would rather see it stepping through.


I thought of a test to find my machine limits.

If I make a line with 5 inch segments broken down into .01, .005, .001 etc I can find the point where the machine stops slowing down through the points and starts bouncing through them, 

I can then make a slight arc with the smallest segment I can run without bouncing, and see if it can run the arc without bouncing. I assume as I increase the angle of the segment, I will have to increase it's length to get it to run smoothly, and if I can figure this out maybe I can beat this dang tangency factor

Link to comment
Share on other sites

There should be a much simpler way of handling breaking up the segments into the length you want. You can just draw a normal circle and do a contour path (or really any path.) Open up the Parameters: Arc Filter/Tolerance: Check smoothing settings: Check Use Fixed Segment Length: Then type the length you want in the box.

 

Looking through this page there may be some settings that will help you that I have never had to use before. What jumps out to me is the Override Arc Feed Rate. You could try decreasing the Max ID Decrease to something like 20 IPM so that instead of relying on the machine to take care of Acc./Dcc. your overriding the feedrate to something the machine can handle. 

  • Like 1
Link to comment
Share on other sites
22 hours ago, manfowar said:

There should be a much simpler way of handling breaking up the segments into the length you want. You can just draw a normal circle and do a contour path (or really any path.) Open up the Parameters: Arc Filter/Tolerance: Check smoothing settings: Check Use Fixed Segment Length: Then type the length you want in the box.

 

Looking through this page there may be some settings that will help you that I have never had to use before. What jumps out to me is the Override Arc Feed Rate. You could try decreasing the Max ID Decrease to something like 20 IPM so that instead of relying on the machine to take care of Acc./Dcc. your overriding the feedrate to something the machine can handle. 

The arc feedrate would probably help, but for a scallop toolpath I need to be able to control what it's doing through line segments since it can't have as many arcs. 

Right now if I make a scallop toolpath with .005 overall tol and nothing else it still shudders like it's starving for code, but with such a loose tol I wouldn't expect it to. 

And when I run verify it doesn't look great but when i run the toolpath it's still jittery and makes it worse than what i would see in verify

scallop.mcam

Link to comment
Share on other sites

Lucas,

I took a look at the file. I can see that in some parts of the path it has extremely jagged edges, this is a result of the pencil toolpath which really isn't designed to be surfacing a surface like this. You mentioned in your post something about scallop so I'm guessing this might not be the path you're using. Even the bigger and better machines out there can't handle the sudden change in direction that the toolpath you had is showing. Most machines only slow down when they see arcs and try to maintain feed on line segments. Your machine may be doing the same thing. I did try a couple of different paths against the part that I was confidant would do a bit cleaner job. First off is Equal Scallop which is a new toolpath in 2019. I saw that the file you posted was in 2018 so I'm not sure if you have that yet. If you don't you could try the Raster path, this is slightly less efficient then the scallop and takes some kindy funky parameters to make sure it doesn't skip over large portions of near vertical walls, but it provides you with basically straight lines instead of arcs all the way across the path, only changing direction when it loops off of the part. Hopefully keeping the tool from stuttering while in contact with it. 

Link to comment
Share on other sites

What thermwood model are you having these problems with? We are about to scrap two oid 67s, not fast enough in my opinion. Our rule of thumb with them is only use raster, waterline and horizontal toolpaths otherwise its gonna be a bad day. Also what type of work are you doing? Using their macros over inputting tan and acc values?

  • Like 1
Link to comment
Share on other sites
On 8/30/2018 at 6:52 AM, manfowar said:

Lucas,

I took a look at the file. I can see that in some parts of the path it has extremely jagged edges, this is a result of the pencil toolpath which really isn't designed to be surfacing a surface like this. You mentioned in your post something about scallop so I'm guessing this might not be the path you're using. Even the bigger and better machines out there can't handle the sudden change in direction that the toolpath you had is showing. Most machines only slow down when they see arcs and try to maintain feed on line segments. Your machine may be doing the same thing. I did try a couple of different paths against the part that I was confidant would do a bit cleaner job. First off is Equal Scallop which is a new toolpath in 2019. I saw that the file you posted was in 2018 so I'm not sure if you have that yet. If you don't you could try the Raster path, this is slightly less efficient then the scallop and takes some kindy funky parameters to make sure it doesn't skip over large portions of near vertical walls, but it provides you with basically straight lines instead of arcs all the way across the path, only changing direction when it loops off of the part. Hopefully keeping the tool from stuttering while in contact with it. 

Thanks for looking at it, i've tried a few paths and like the way the pencil looked, but now I am working with the hybrid toolpath. Could you maybe screenshot an area that was particularly jagged? I'd like to know what to look for. I did try raster and still noticed some studdering, i caved today and bumped the tangency to 10 and it's smoothed out a lot of it, i bet if I went back to the raster I'd be able to get it to work but i did see the walls weren't looking great like you said. 

3 hours ago, Sharn said:

What thermwood model are you having these problems with? We are about to scrap two oid 67s, not fast enough in my opinion. Our rule of thumb with them is only use raster, waterline and horizontal toolpaths otherwise its gonna be a bad day. Also what type of work are you doing? Using their macros over inputting tan and acc values?

A new model 70, plenty top speed but I don't know if I'll ever get it there. it can definitely cut straight lines just fine...

Mostly sculptural surfacing on low density tooling board. Using what macros?

 

 

Is there anything with the machine that could cause studdering? If the tolerance of the machine is off more than it used to be, could that make it shaky, or just not as accurate?

Link to comment
Share on other sites

Here is a picture (Pencil Stutter) of the sharp edges I spoke of, I took a screen shot and outlined what I was talking about in red. If you zoom in on these sections in MC you can see the sudden direction changes.

 

This is a result of the way Pencil is processed. It is offsetting internal corner radii. I have attached a pdf of the offsetting I'm talking about. When you look at it, the left side is the original line that I drew and as the lines are offset in eventually a radii becomes a sharp corner. This is why pencil isn't really the best at surfacing a whole part. It's more designed to pick out corners left behind by other paths.

 

Hope this helps.

 

Caleb

PENCIL STUTTER.BMP

Sample Offset.pdf

  • Like 1
Link to comment
Share on other sites
1 hour ago, manfowar said:

Here is a picture (Pencil Stutter) of the sharp edges I spoke of, I took a screen shot and outlined what I was talking about in red. If you zoom in on these sections in MC you can see the sudden direction changes.

 

This is a result of the way Pencil is processed. It is offsetting internal corner radii. I have attached a pdf of the offsetting I'm talking about. When you look at it, the left side is the original line that I drew and as the lines are offset in eventually a radii becomes a sharp corner. This is why pencil isn't really the best at surfacing a whole part. It's more designed to pick out corners left behind by other paths.

 

Hope this helps.

 

Caleb

PENCIL STUTTER.BMP

Sample Offset.pdf

you've helped a ton, i appreciate it all. 

new development: I block stopped the tool during a curve that it was bouncing through.

I stepped from point to point and could still see it shake. it was not the same shake and did not happen every time.
I found four lines that i stepped back and forth between that usually caused shudder but not everytime:

G01 X3.1849 Y6.7586 Z1.2835 C24.5141 B-9.648

G01 X3.1636 Y6.7551 Z1.2804 C23.1278 B-10.2792

G01 X3.1415 Y6.7507 Z1.2784 C21.7406 B-11.9283

G01 X3.1219 Y6.7459 Z1.2759 C21.9351 B-12.1107

Even if these points are not tangent, I don't think I should be seeing it studder when I step through them? Or could this happen if I have too high acceleration?

This really got me thinking it's a machine problem, got recommended to check for lash. Haven't done that before but I'll let you know when I get the results.

Link to comment
Share on other sites
On 8/31/2018 at 9:45 AM, manfowar said:

Here is a picture (Pencil Stutter) of the sharp edges I spoke of, I took a screen shot and outlined what I was talking about in red. If you zoom in on these sections in MC you can see the sudden direction changes.

Also I definitely see the sharp changes in these areas, but the shaking is happening all over the part, starting at the first plunge

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

  • Recently Browsing   0 members

    • No registered users viewing this page.

Join us!

eMastercam - your online source for all things Mastercam.

Together, we are the strongest Mastercam community on the web with over 56,000 members, and our online store offers a wide selection of training materials for all applications and skill levels.

Follow us

×
×
  • Create New...