Circle problems

Mach3 and Mach4 questions, tips and topics can be posted here
Post Reply
Leeroy6392
1.5 Star Member
1.5 Star Member
Posts: 32
Joined: Mon Jun 17, 2019 9:39 pm
Location: Tulsa ok

Circle problems

Post by Leeroy6392 »

I am having issues with flanges I am cutting out. The first 60 or so did not have any issues other than my programming on the first couple now after 50+ it has started making a egg shaped flange and bolt pattern follows the same egg. I have checked the steps/ calibrated the axis. I have ground rod attached to the table to limit interference, matched speeds on acceleration and travel speed between x and y. I’m just lost any help is appreciated.
Attachments
78E2ED39-D50A-4CFC-B601-A17550895CEA.jpeg
robertspark
4.5 Star Elite Contributing Member
4.5 Star Elite Contributing Member
Posts: 1806
Joined: Mon Jun 12, 2017 6:43 pm

Re: Circle problems

Post by robertspark »

something mechanically worked loose (pinion) check your axis in the direction of the smaller (squashed ) circle.... gcode does not change.... the machine will still be "trying" to output the same steps to both axis
Leeroy6392
1.5 Star Member
1.5 Star Member
Posts: 32
Joined: Mon Jun 17, 2019 9:39 pm
Location: Tulsa ok

Re: Circle problems

Post by Leeroy6392 »

All three motors are tight, guides, y axis and torch mount. I thought it might have been torch lead pulling so tightened and still draws in and out on the 45 degrees but it does measure out to only be on one side.
robertspark
4.5 Star Elite Contributing Member
4.5 Star Elite Contributing Member
Posts: 1806
Joined: Mon Jun 12, 2017 6:43 pm

Re: Circle problems

Post by robertspark »

Picking at straws now..... what cnc software are you running? (Mach3, Linuxcnc???)

Have you shut down the PC, restarted and tried again to see if the issue is consistent?

It may be a memory thing not being flushed and something is wrong with the planner.... (like I said its just a guess.... but it can be only one of a few things..... mechanical, signal transmission (is what is being sent being recieved correctly), drives or motors (have you tried swapping axis / motors around if that is possible .... does the problem move (may be a motor thing / axis thing then), or its a planner thing (unlikley.... but who knows)

Try to break the problem in 1/2.... swap the motors around on the drives ./.... does the problem move? Swap the drives around..... does the problem move....?

Sorry out of ideas maybe someone else will have a brainwave...
weldguy
4 Star Elite Contributing Member
4 Star Elite Contributing Member
Posts: 1929
Joined: Fri Apr 17, 2009 11:48 am

Re: Circle problems

Post by weldguy »

Have you done a dry run with the plasma off and the shop quiet to see if you notice a hiccup, strange noise, or jerky motion at some point in the design?
Leeroy6392
1.5 Star Member
1.5 Star Member
Posts: 32
Joined: Mon Jun 17, 2019 9:39 pm
Location: Tulsa ok

Re: Circle problems

Post by Leeroy6392 »

I have ran with everything off in the shop including the plasma it still seemed to run fine never noticed any issues. The part are not jerky it’s still smooth all the way around (not like motor I popping out of the gear).
Yes I have turned it off since the issue occurred bad thing was cut 3- 6x12’ sheets over 2 days went to make the parts then realized they were misshaped. They are all consistent in location direction the south east side is long while the north east side is short. I will see what I can do on swapping motors.
Bucket8577
1 Star Member
1 Star Member
Posts: 23
Joined: Wed Apr 03, 2019 6:19 am

Re: Circle problems

Post by Bucket8577 »

I would do some more testing before swapping motors.

The first thing I would do is do some test cuts in some of the scrap pieces. Simple squares. Measure them and see where the error is.

I would also re-check everything mechanical for the axis that has the error.

One more thing. Has anyone edited the g-code by hand? Maybe re-compile it and give a run.
Newbie to plasma cutting.
User avatar
acourtjester
6 Star Elite Contributing Member
6 Star Elite Contributing Member
Posts: 7770
Joined: Sat Jun 02, 2012 6:04 pm
Location: Pensacola, Fla

Re: Circle problems

Post by acourtjester »

You do have slippage in the mechanics, when cutting large circles the transition between the X and Y is very small and hard to see it happen.
this will not get better :Sad
slipping.JPG
slipping.JPG (34.01 KiB) Viewed 1343 times
DIY 4X4 Plasma/Router Table
Hypertherm PM65 Machine Torch
Drag Knife and Scribe
Miller Mig welder
13" metal lathe
Small Mill
Everlast PowerTig 255 EXT
User avatar
WyoGreen
4 Star Member
4 Star Member
Posts: 897
Joined: Tue Mar 04, 2014 8:36 pm
Location: Cheyenne, Wyoming

Re: Circle problems

Post by WyoGreen »

You might try sending commands to move the torch in the 2 directions and measuring the distance moved versus what was commanded to move. Perhaps one of the axis is getting tight and can't accelerate as fast as it used to and is losing steps.

Steve
Precision Plasma gantry
CommandCNC Linux controller w/Feather Touch & PN200 hand controller
HT-45 plasma cutter
Plate Marker
Router
Laser
robertspark
4.5 Star Elite Contributing Member
4.5 Star Elite Contributing Member
Posts: 1806
Joined: Mon Jun 12, 2017 6:43 pm

Re: Circle problems

Post by robertspark »

are you running servos or stepper motors?
Leeroy6392
1.5 Star Member
1.5 Star Member
Posts: 32
Joined: Mon Jun 17, 2019 9:39 pm
Location: Tulsa ok

Re: Circle problems

Post by Leeroy6392 »

I’m running stepper motors. But the circle is half correct then I has one qtr pushed in 1/8”-3/8” and the other pushed out 1/8”-3/8” I did drop to the inside of the scrapped piecesand it was only out 1/8”
Leeroy6392
1.5 Star Member
1.5 Star Member
Posts: 32
Joined: Mon Jun 17, 2019 9:39 pm
Location: Tulsa ok

Re: Circle problems

Post by Leeroy6392 »

Ok now I just ran a program parts not horrible but has done a lot better. Traversing the table and brining it back to front of the table the y and a motor out of alignment 5/16” could one be getting weak? Thanks for all the help so far by the way.
Bucket8577
1 Star Member
1 Star Member
Posts: 23
Joined: Wed Apr 03, 2019 6:19 am

Re: Circle problems

Post by Bucket8577 »

Leeroy6392 wrote: Wed Jul 10, 2019 8:17 pm Ok now I just ran a program parts not horrible but has done a lot better. Traversing the table and brining it back to front of the table the y and a motor out of alignment 5/16” could one be getting weak? Thanks for all the help so far by the way.
Sounds like you have a gear or pulley spinning slightly on a shaft.
Newbie to plasma cutting.
Leeroy6392
1.5 Star Member
1.5 Star Member
Posts: 32
Joined: Mon Jun 17, 2019 9:39 pm
Location: Tulsa ok

Re: Circle problems

Post by Leeroy6392 »

Just slid the motors back gears are tight could something not be communicating to both motors throwing off the steps? Or do I have a weak motor that needs changed.
weldguy
4 Star Elite Contributing Member
4 Star Elite Contributing Member
Posts: 1929
Joined: Fri Apr 17, 2009 11:48 am

Re: Circle problems

Post by weldguy »

Try cutting a giant L in the lower left corner of a sheet and using a large square be sure the L you just cut is square. If the left and right side of the gantry are not even you cannot cut square and can cause all kinds of problems like this. May need to re-align the left and right sides.

If your losing steps then you will not return back to the exact same position you started at. Mark your starting position on the rails with tape or sharpie and do a dry runt to see if your returning to where your starting.
User avatar
acourtjester
6 Star Elite Contributing Member
6 Star Elite Contributing Member
Posts: 7770
Joined: Sat Jun 02, 2012 6:04 pm
Location: Pensacola, Fla

Re: Circle problems

Post by acourtjester »

I have found the motors having a hardened shafts can be a problem I file or machine 2 flats that match the set screw location on the pulley/gear used in the drive assemble.
DIY 4X4 Plasma/Router Table
Hypertherm PM65 Machine Torch
Drag Knife and Scribe
Miller Mig welder
13" metal lathe
Small Mill
Everlast PowerTig 255 EXT
Leeroy6392
1.5 Star Member
1.5 Star Member
Posts: 32
Joined: Mon Jun 17, 2019 9:39 pm
Location: Tulsa ok

Re: Circle problems

Post by Leeroy6392 »

Update: still having issues
1. Checked motors and the have a flat for the set screw
2. Installed new motors( same that it came with)
3. Tightened motor spring to allow movement but not enough to come out of the rack
4. Adjusted acceleration down throwing stuff at the wall to get it to work
5. Turned everything off including lights in the shop still messed up so not em interference from what I understand

What’s on my mind on possibilities:
1. I was sold a piece of junk that only worked properly for a couple of months
2. Stepper motors are not large enough
3. Linear guides are poor design from others I have seen
4. Mach is lagging on direction changes or motors are
Attachments
CAB97F37-FC4D-4DBB-AE51-D7F7774F1BB8.jpeg
A688848E-3DAE-4F8A-BAC9-00F9BA5C792F.jpeg
97683C88-DC30-4FFD-9792-4C239C809966.jpeg
User avatar
Nacs Fab
3 Star Member
3 Star Member
Posts: 329
Joined: Tue Dec 04, 2012 4:11 pm
Location: Columbia, MO

Re: Circle problems

Post by Nacs Fab »

I have started using red lock-tite on my pinion shafts as well as set screws. Its a bitch to change pinion but it never slips. I would guarantee you have a slipping pinion somewhere.
Shop Droids Plasmabot 3.0 DIY 4x8 with water table
G540 with LCTHC by candcnc
Hypertherm 1250 with Duramax Torch Upgrade
all built... by me...
Leeroy6392
1.5 Star Member
1.5 Star Member
Posts: 32
Joined: Mon Jun 17, 2019 9:39 pm
Location: Tulsa ok

Re: Circle problems

Post by Leeroy6392 »

But if it has machined flats it shouldn’t be moving. I just removed them to check it and they are snug as can be.

What motors do you run? These were only $35 a piece that just seems cheap to me.


And if your running similar setup can I see your motor tuning page? I have no idea what I’m doing in those just making sure the motors are sounding as quiet as possible.
cuttinparts
3 Star Member
3 Star Member
Posts: 351
Joined: Tue Sep 30, 2008 3:40 pm

Re: Circle problems

Post by cuttinparts »

I've seen pinions slip even on motors with flats. If the setscrew in the pinion gear backs off even a little bit the motor shaft will rotate inside the pinion gear until the backed out setscrew contacts one side of the flat on the shaft. When motor changes direction the shaft turns but the pinion doesn't until the setscrew hits the the other side of the flat. I'm with Nacs Fab and after seeing this more than once started using Loctite on the setscrews.
Leeroy6392
1.5 Star Member
1.5 Star Member
Posts: 32
Joined: Mon Jun 17, 2019 9:39 pm
Location: Tulsa ok

Re: Circle problems

Post by Leeroy6392 »

I want to thank you all for all the help so far I can not tell you how much I appreciate it. That being said I have put loctite on the set screws. I have also went to a heavier spring on my motors and have not witnessed either slipping out of the rack. But I have noticed my x axis motor has a different step count than my y axis even though they are the same motor and track. Could my a motor have different steps than my y motor?if anyone is willing to talk on the phone I would appreciate it as well or anyone in the Tulsa area I would pay to have you stop by and check it out.
beefy
4.5 Star Member
4.5 Star Member
Posts: 1504
Joined: Fri Jan 18, 2013 3:19 am

Re: Circle problems

Post by beefy »

Leeroy,

can you post the gcode file you are using, plus your Sheetcam job file, plus your DXF. If those are proprietary to you, can you at least make some file/gcode that gives the same results and post that. If doing a marker pen "dry run" test (mentioned below) gives your the same incorrect shape then you don't have to waste steel on a real test cut.

Also, what version of Mach are you using.

What motion tests have you done so far ? For example there's the been the suggestion of a square, a letter L, etc. I would do any such tests as "dry" run, i.e. without cutting. Instead mount a fine marker pen to your Z axis, ideally in a fine fitting tube so the marker pen can drop/rise with any changes in plate height.
You can use the measure feature in Sheetcam to check the distances of the various sides of any test shapes, and confirm with a tape measure that your axis moved that distance. You could even single step through your gcode, confirming with the tape measure that the axis moved the correct distance each time.

At this point in time I think you do NOT have a slipping axis issue. Reason is your circle appears to come back accurately to the start point. If an axis was slipping, I'd expect the start/finish points of the circle would not meet. I'm assuming ALL of your start/end points are meeting accurately. I would be surprised if an axis would slip so gradually and consistently to create the oval you are getting, then meet back at the start point. Just can't see it happening.

When I first built my table I did a dry run "torture test" for lost steps. I made a basic XY run file with circles, lots of zig zags, whatever crazy shapes I could think of. This run lasted over 10 mins at rapid speed (10,000 mm/min), and I think 1500 mm/s/s acceleration, it looked like the table was having a nutter attack. At the end the Z axis position went exactly back to the start point indicating I had not lost a single step.

Keith.
2500 x 1500 water table
Powermax 1250 & Duramax torch (because of the new $$$$ync system, will buy Thermal Dynamics next)
LinuxCNC
Sheetcam
Alibre Design 3D solid modelling
Coreldraw 2019
Leeroy6392
1.5 Star Member
1.5 Star Member
Posts: 32
Joined: Mon Jun 17, 2019 9:39 pm
Location: Tulsa ok

Re: Circle problems

Post by Leeroy6392 »

The l ( I made a framing square) came out fine. It does return to a good finish point which is what has me so confused. The tap file I use to load into Mach 3 would not upload so I’m back out in my shop trying some different stuff.
On which Mach3 version where would I find that?
I also tried running some complex files when the square turned out good, thinking I had it fixed. It did not turn out well lol.
I am trying the guide rails now seeing if backing them off on tension maybe it’s just randomly causing it to bind up?
User avatar
WyoGreen
4 Star Member
4 Star Member
Posts: 897
Joined: Tue Mar 04, 2014 8:36 pm
Location: Cheyenne, Wyoming

Re: Circle problems

Post by WyoGreen »

Leeroy,
I've been reading thru this thread, and read the post where you mentioned your motor step per settings were different. Normally, the 2 Y axis motors (Y and A) would have the same settings, other than the rotation direction being set different. The X and Y could have different settings if there is a difference in the rack and pinion. If they have the same parts, then they also should be the same. I'm not sure if you mentioned it or not, but have you measured the distance traveled X & Y wise versus the distance commanded?
Precision Plasma gantry
CommandCNC Linux controller w/Feather Touch & PN200 hand controller
HT-45 plasma cutter
Plate Marker
Router
Laser
Leeroy6392
1.5 Star Member
1.5 Star Member
Posts: 32
Joined: Mon Jun 17, 2019 9:39 pm
Location: Tulsa ok

Re: Circle problems

Post by Leeroy6392 »

Yes during calibration they move correctly but even when I just drive it north and south with the keyboard(with a lot of directional changes) it will get off. I have also noticed in moving between 2 sets of holes( out of 30) on each part it will sound like the motor is bogging down.
Post Reply

Return to “Mach3 & Mach4 CNC”