Machsupport Forum

G-Code, CAD, and CAM => G-Code, CAD, and CAM discussions => Topic started by: RocketmanOU on February 14, 2012, 08:58:59 PM

Title: Arcs not cutting smoothly
Post by: RocketmanOU on February 14, 2012, 08:58:59 PM
I profiled a part this morning that should be relatively simple, but the arcs are coming out as what seem to be either line segments or much smaller arcs. I don't see anything in the g-code that should cause this, so I'm pretty confident it's not a Cam issue (I'm using CamBam). Any ideas what might be causing this and how I could correct it? Thanks!

(http://i927.photobucket.com/albums/ad111/RocketmanOU/Machining/ConnectingLinkSmall.jpg)

(http://i927.photobucket.com/albums/ad111/RocketmanOU/Machining/DSC06381Small.jpg)

(http://i927.photobucket.com/albums/ad111/RocketmanOU/Machining/DSC06379Small.jpg)


G-Code:

G20 G90 G91.1 G64 G40
G0 Z0.5
T5 M6
( Outside Profile)
G17
M3 S1000
G0 X-6.1827 Y-0.4475
G1 F10.0 Z-0.03
G1 F30.0 X-1.4882
G2 X-0.6681 Y-0.7575 I0.0 J-1.24
G3 Y0.7575 I0.6681 J0.7575
G2 X-1.4882 Y0.4475 I-0.8202 J0.93
[...repeated again and again and again with decreasing z...]
G0 Z0.5
M5
M30
Title: Re: Arcs not cutting smoothly
Post by: Overloaded on February 14, 2012, 10:42:12 PM
First thought is, what is the resolution of the machine ?
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 14, 2012, 11:14:04 PM
Quantified how? My motors are 10000 pulses per inch, running Mach3 at 45 kHz kernel frequency on a very powerful  (3.2 GHz quad core, 4 GB RAM) computer. I have glass scales on the axes so I can watch real measurements, and I can get precise motion down to about 0.001" or so when I'm using backlash compensation. Backlash in the machine is only 0.004" or so anyway though. For the record, I cut the rough cut with backlash compensation on, then did the final pass with it off to see if it made any difference - nada.
Title: Re: Arcs not cutting smoothly
Post by: Overloaded on February 14, 2012, 11:23:20 PM
The electronic resolution at .0001 is fine enough, but .004 backlash is huge.
Have you had better result with arcs prior to these ?
What type of machine is it ? Router ? Mill ?, ? ? ?
The extent of y in the toolpath is about 1.75", what is the dia. of the tool ? or, what is the rad. of the rod ends ?
Title: Re: Arcs not cutting smoothly
Post by: RICH on February 15, 2012, 06:22:42 AM
The straights along the arc are rather consistant. Would be interesting to have you measure them , then draw the flat inside  or along the diameter
and see how much deviation there is from the arc. I am going to guess it's related to the backlash and the deviation from the arc is close  or relates to the
backlash.  Whats the different in measurement across flats as compared to arc?

Not sure how the BC  is implemented during CV mode. ???
May be all wet......... ;)

RICH
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 15, 2012, 08:02:30 AM
The outside arc radii are 0.75" and 0.375", and the inside radii (the fillets) are 1.5". I used a 1/2" endmill for roughing and a 1/4" stubby endmill with large shank for the finish pass. It's a LMS mini mill with the CNCFusion ballscrew package. These are some of the first parts coming off the machine, so I can't speak to whether this is something that just cropped up or not.

The length of the flats are .160" on the outside of the large diameter (0.75") and 0.120" on the outside of the smaller diameter. They're also not flat, now that I take a close look. They're definitely arcs in and of themselves.

If this were due to backlash, I would expect to see a seam any place an axis changes direction, but only there. In other words, I should see variation at these points, and these points alone:

(http://i927.photobucket.com/albums/ad111/RocketmanOU/Machining/ridgesSmall.jpg)

Outside of the backlash, the axes are very consistent - example: I make a g0 move from x0 to x-3.7500, then from x-3.7500 to x-3.5000. I now zero the glass DRO, and I can make any move in the positive direction and be pretty much dead on (within a couple of ten thousandths, anyway).
Title: Re: Arcs not cutting smoothly
Post by: Overloaded on February 15, 2012, 08:29:40 AM
You might try CV/ES with BC On/Off just to see if there is a difference.
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 15, 2012, 08:49:33 AM
As I said, I tried CV with BC on and off, and there was no difference. I can give exact stop a try if you think it would help.
Title: Re: Arcs not cutting smoothly
Post by: Dan13 on February 15, 2012, 10:30:18 AM
Doesn't look like backlash to me. I would, however, guess that it is a mechanical problem in one (or both) of the axes. If you have belt transmission, for instance, a run-out of the pulley could cause something like this.

To make troubleshooting easier you could hold a pencil in the spindle and draw the tool path on a paper.

By the way, the transition points you specified in the above drawing are not all correct. The 2 diagonal lines should be removed as there is no direction change in either axis.

Dan
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 15, 2012, 10:53:02 AM
Doesn't look like backlash to me. I would, however, guess that it is a mechanical problem in one (or both) of the axes. If you have belt transmission, for instance, a run-out of the pulley could cause something like this.


My motors are direct drive, and the Lovejoy couplers are nice and snug. Plus, if it were a problem such as that, I'd also see ridges (or at least feed changes) on long, straight sections, and I'm not seeing that at all. Per part inspection, the feed on the straight sections is remarkably uniform.

By the way, the transition points you specified in the above drawing are not all correct. The 2 diagonal lines should be removed as there is no direction change in either axis.

Ah, good point. I was thinking in velocity and not displacement.
Title: Re: Arcs not cutting smoothly
Post by: Dan13 on February 15, 2012, 11:09:43 AM
Is the arc segmented all around? And segments are absolutely equal in length (both on same arc and different radius arcs)? Can't be seen on the photo.

Mach3 sometimes does do strange things. Have you tried restarting the PC?

Have you successfully cut arcs before?

Dan
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 15, 2012, 11:27:42 AM
Arcs are indeed segmented all around. As I said before, the measurements are different on different radius arcs, but the size change isn't proportional to radius. I've restared mach3, restarted the computer, tried a couple of different things. Same results.
Title: Re: Arcs not cutting smoothly
Post by: Dan13 on February 15, 2012, 11:42:27 AM
Sorry I missed this:

The length of the flats are .160" on the outside of the large diameter (0.75") and 0.120" on the outside of the smaller diameter. They're also not flat, now that I take a close look. They're definitely arcs in and of themselves.

Outside of the backlash, the axes are very consistent - example: I make a g0 move from x0 to x-3.7500, then from x-3.7500 to x-3.5000. I now zero the glass DRO, and I can make any move in the positive direction and be pretty much dead on (within a couple of ten thousandths, anyway).

Can you extend this test. Zero Mach3 and glass scales (after taking out any backlash) and then do short moves of 0.01" up to a total of about 0.15" and compare the numbers in Mach3 and the glass scale at each point and see if they correlate. Repeat on both axes.

Dan
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 15, 2012, 02:30:00 PM
OK, I did exactly as you asked - I started at 0, then I did g0 x0.010, then g0 x0.020, etc, and on the y axis as well. Here are the results:

g0xy
0.0100.00960.0100
0.0200.01960.0200
0.0300.02940.0300
0.0400.03940.0400
0.0500.04960.0500
0.0600.06000.0600
0.0700.07060.0700
0.0800.08020.0800
0.0900.08980.0900
0.1000.09920.1000
0.1100.10880.1100
0.1200.11860.1198
0.1300.12860.1300
0.1400.13860.1398
0.1500.14940.1498

By the way, and I'll put this in bold, I WOULD PAY GOOD MONEY FOR AN EXPLANATION OF HOW BALLSCREW MAPPING WORKS. Seriously, if someone would explain it (even just the basics) to me, I'd be happy to put together a nice article detailing its use and publish it for everyone. Seriously.
Title: Re: Arcs not cutting smoothly
Post by: RICH on February 15, 2012, 05:02:10 PM
EXPLANATION OF HOW BALLSCREW MAPPING WORKS

I don't recall anyone posting about it . Just looking at the screen it would seem that  one could find some deviation and then have Mach account for it.
I have profiled ball screws and in doing that you find and graph the deviation of the axis movement as to perfect movement ( your comparing to a know calibrated standard).
Sort of what you did in reply #13 but you do it over the total range of the screw. As a example the screw may gain , deviate, or lose and not necessarily in a linear fashion and that includes both travel directions of the screw.

RICH
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 15, 2012, 05:24:26 PM
Oh, I get that much - how it works in Mach3, however, is a mystery.  When you enter the correction points, do you enter the correction you want to apply, or the actual coordinate read on the external DRO?  I try to enter correction points, but nothing happens. Perhaps this is a bug in that it only works in metric? It defaults the screw lengths to 1300, and the graph of corrections is in increments of 0.1. If this is a small bug, it would be a great help if it could be corrected, as I do everything in inches.
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 15, 2012, 06:03:44 PM
Poking around on my own, the graph is simply poorly scaled for English units. I went through and entered a series of points along the x-axis by doing incremental g0 moves, then entering what the point reads on the DRO, clicking add correction, then save curves. You can't see any changes on the graph, but they're definitely stored in there. I entered about 40 points along the x-axis, then went back and did some checks - it seems that the map made accuracy decidedly worse. I'm guessing I'm using it wrong, but I dunno. You can definitely hear the motor speed changing based on location along the screw when you rapid jog back and forth.

In any case, this doesn't change anything about the segmented arcs I'm getting on this part. I tried exact stop this afternoon, and it didn't make a difference. Still getting the same behavior.   ???
Title: Re: Arcs not cutting smoothly
Post by: Dan13 on February 16, 2012, 04:36:20 AM
In any case, this doesn't change anything about the segmented arcs I'm getting on this part. I tried exact stop this afternoon, and it didn't make a difference. Still getting the same behavior.   ???

I think the answer lies right there, in the numbers of your test. X axis is pretty much obvious, heaving an absolute deviation of 0.002", which is enough for those segments to be seen. I would guess that if you did the test over a longer distance you would see a distinct repetitive nature of the numbers. Entering the numbers in Excel would make it easier to visualize.

Since you're not using a belt drive, you have to start looking elsewhere in the driver mechanism. For instance, what is the motor? Could be a stepper motor with poor accuracy - unevenly spaced poles or magnets on the rotor. Since your Y axis is much better than the X, you could try swapping the motors.

Dan
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 16, 2012, 08:56:48 AM
I'll give that a shot. I'm not sure I agree, given that, for the most part, absolute deviation in the x is under 0.0005" (with several spots up to 0.0014" max), and incremental deviation is even less. Also, recall that the steps that I'm seeing are regular regardless of orientation in x and y, and are dependent on radius of the arc.

In my continuing quest to eliminate backlash, I just placed an order for some stainless helical couplings from McMaster - should be here tomorrow. I'm currently using the lovejoy spider couplings that came with the cncfusion kit, which are far from perfect.
Title: Re: Arcs not cutting smoothly
Post by: RICH on February 16, 2012, 04:51:32 PM
Quote
lovejoy spider couplings


Have a pic of these as the name is not familar?

RICH
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 16, 2012, 06:47:23 PM
http://www.ebay.com/itm/CNC-Stepper-Motor-Lovejoy-Coupling-3-8-X-1-4-spider-/330548952567 (http://www.ebay.com/itm/CNC-Stepper-Motor-Lovejoy-Coupling-3-8-X-1-4-spider-/330548952567)

Basically two hubs that squeeze a rubber spider, which allows some misalignment while transmitting torque. The squeeze of the spider and any space between it and the coupler ends results in backlash.
Title: Re: Arcs not cutting smoothly
Post by: RICH on February 16, 2012, 07:36:25 PM
If you can get the alignment right get rid of those suckers and hard couple.
Also watch out for the Flexible Beam Couplings ( all metal by the way). They are not cheap. I used them on the lathe and then all of a sudden
it was experiencing backlash / inaccuracies ( .001 to .003" and not repeatable). Short story, the coupling over time just became week and with rather little torque ( say 40 in oz )
applied to it it allowed twisting with out motion. Drove me nuts until I found it and would never even have thought that it was the culprit.

RICH
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 16, 2012, 08:11:27 PM
What material coupler did you use? Aluminum helical couplers are total crap, and wear out rather quickly if used out of spec. The ones I have on the way are stainless, and rated far beyond what I'll need for torque. I worked in a manufacturing plant where we used these for a number of different operations - some with good results, and others with horrid results. The stainless couplers are indeed not cheap, but it's a lot easier to spend the money once than to worry about grinding your bearings to bits over misalignment that you can't even effectively measure. They're plenty stiff, and enjoy a fatigue limit that aluminum can't achieve.
Title: Re: Arcs not cutting smoothly
Post by: BR549 on February 16, 2012, 09:23:05 PM
The coupler is not the factor here the SPIDER is what makes the coupling good or bad. We always had good luck with the ORANGE spiders.

I have studied this over and over again. I have never seen Mach3 cut an arc in this fashion whne it was not line segmented code. BUT the error is TOO consistant in nature to be mechanical or so it would seem. It repeats itself exactly over ALL the stepdowns in Z.

It would be interesting to see it DRAW a circle based on ARC as large as you can(no cutting loads)  and see IF the error is still there.

I do know IF you ZOOM in on the toolpath you will "see"  segments BUT that is just the display. OR is it????

(;-) TP
Title: Re: Arcs not cutting smoothly
Post by: RocketmanOU on February 17, 2012, 02:01:29 PM
Got the new couplers installed this morning, x backlash is down to 0.0012", y down to 0.0024". Definitely an improvement. With BC on, I see no variation in position from Mach3 to DRO. I'll hopefully get to cut some curves this evening, and we'll see what happens.
Title: Re: Arcs not cutting smoothly
Post by: Ariel_HT5_John on February 23, 2012, 01:57:11 PM
I have had this problem. It was nothing to do with the machine, the machine was a Bridgeport VMC500 withe a Heidenhein TNC530. it was caused by the tolerance used in the Cam package. I reduced the tolerance then it was fine.

Backlash will show up as mismatch on direction change.
Title: Re: Arcs not cutting smoothly
Post by: Overloaded on February 23, 2012, 02:03:30 PM
That makes perfect (symmetrical) sense !
Thanks for posting the suggestion.
Russ
Title: Re: Arcs not cutting smoothly
Post by: Dan13 on February 24, 2012, 02:08:01 AM
It's different in this case, I am afraid, as he has arcs programmed in the Gcode and not segments.

Dan
Title: Re: Arcs not cutting smoothly
Post by: Conundrum68 on July 24, 2012, 07:02:58 PM
Got the new couplers installed this morning, x backlash is down to 0.0012", y down to 0.0024". Definitely an improvement. With BC on, I see no variation in position from Mach3 to DRO. I'll hopefully get to cut some curves this evening, and we'll see what happens.
I know this is an old thread, but did this ever resolve?
I am experiencing the same exact problem.
Title: Re: Arcs not cutting smoothly
Post by: Ariel_HT5_John on July 25, 2012, 03:08:41 AM
If you are cutting an arc say from 12 o'clock to 3 o'clock is it smooth in any position? i have noticed on my machine its better in some places than other. Between the 1 & 2 position is the worst.
Title: Re: Arcs not cutting smoothly
Post by: Conundrum68 on July 25, 2012, 12:20:36 PM
If you are cutting an arc say from 12 o'clock to 3 o'clock is it smooth in any position? i have noticed on my machine its better in some places than other. Between the 1 & 2 position is the worst.
It seems to always be present during arcs. For a reality check I ran the same part with LinuxCNC and the results were exactly the same. Facets in the same places. So unless they are using the same exact stepping routines, I am going to have to look into my mechanics. Always a PITA. Guessing it has something to do with the system surging due to either timing belt, slightly off-center timing pulley, or possibly a wobbly lead-screw. The parts are accurate within .002, but the facets are bad.
Title: Re: Arcs not cutting smoothly
Post by: Graham Waterworth on July 28, 2012, 05:08:47 AM
Are your slideways sticking or too tight on the gib , this can cause a buildup of pressure that releases in a jolt, you can see this on a DTI clock set against a block with the backlash removed first if you jog in very small moves in the same direction, the clock will not move and then suddenly jump.

Only my two penath,