Torchmate loses position

[tigger_na]

Dear Forumites

I’ve been having problems on and off with my machine losing position, and this time around I have managed to isolate the problem.

Attached is the troubleshooting file, g-code and actual cut. The problem comes in when the program reaches the apex of the cut and goes down again, line 28 in the code. If I remark line 28, the cut works perfectly, so this isn’t an issue with sticking or loose gears, grounding or any other more common causes. The problem is repeatable, it will cut perfectly every time with line 28 removed, and loose position every time it isn’t, irrespective on where on the table I cut.

Has anyone encountered, and solved, this problem before? I'm using a Routermate (with a heavy gantry) and TM version 4.0.61.

Best regards
Arthur


(File created using Torchmate DXF Import)
(Import File: Moria_Repair_Cut_2.dxf)
(Import Date: 12/21/2017)

M06 T1
G00 X23.6420 Y19.3882
M50
G03 X22.7681 Y18.6252 I1.5953 J-2.7094 F3000.00
G03 X22.4255 Y15.3237 I2.4173 J-1.9193
G01 X24.8887 Y10.7168
G02 X27.1148 Y6.1172 I-55.0888 J-29.4998
G01 X27.3175 Y5.5494
G01 X27.3736 Y4.7968
G01 X25.5220 Y3.3174
G01 X24.8180 Y3.0958
G01 X20.6842 Y2.6981
G01 X20.6757 Y0.0151
G01 X6.6938 Y0.0000
G01 X6.7025 Y2.7375
G02 X2.7916 Y3.3345 I1.7683 J24.6945
G01 X1.1432 Y4.1556
G01 X0.6848 Y4.5059
G01 X0.0295 Y5.9727
G01 X0.0000 Y6.6678
G01 X2.8845 Y12.6182
G01 X4.0332 Y14.6063
G01 X13.6726 Y30.4669
(G01 X14.0473 Y30.4580
G01 X22.4255 Y15.3237
G01 X22.8342 Y14.8059
G03 X26.4900 Y14.1072 I2.3133 J2.1903
M51

G00 X0.0000 Y0.0000

 

[LECS-Chad]

Rotate your file and see if you get the same result. With the older system some "angles" would cause the head to accel but not be able to "catch" itself for a motion change. Looks like you found it.

 

[Dnmeistr-LECS]

You can also modify your Continuous Contouring under Configuration > Machine > Feedrate Ramping, default is 70 try dropping the speed down to see if that fixes the issue.

 

[tigger_na]

Thank you, gents.

Chad, if this is the issue, and it may well be, then I'd need to know what moves, exactly, cause this. I have a program than can evaluate (and change, within predefined limits), angles and line/arc lengths. I use it to eliminate tiny moves that I always thought caused this. This cut, with a length of 0.014", is exactly the type of move it is designed to catch and delecte. I didn't run a clean-up on this file, since it wasn't very large to begin with. Would it be possible for you to test this in the factory, with my settings, on one of your controllers? Needless to say, this behaviour can, and did, completely ruin larger, read expensive, cuts.

Dnmeistr, will do. I currently have 106 dialed in, quite a lot higher than the recommended 70.

 

[tigger_na]

Dnmeistr, reducing the continuous contouring feedrate to 1,780mm/min did the trick, thank you.

Rotating the part by 90° also worked.