Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: dodiano on March 24, 2009, 11:09:27 PM
-
Hi,
Want to introduce myself my name is Roberto Soriano I have a fully licensed MACH 3 software which was bought with my CNC machines from 3DROUTERS.COM long gone I think... Well I have been cutting several small parts very basic shapes in 5mm Acrylic...
I was cutting today and everything was running normally... When I took the finished piece out of the table and measured it was 1/16th of an inch shorter in width... I double checked my drawing ins Autocad Saved in DFX re-measured and everything was fine... Well after trying to figure out what was happening I simply re-draw the parts again from scratch... And convert it via LAZYCAM from DXF to GCODE and we carved the pieces into a piece of wood to measure if they where correct... And of course it was this time no mistake and carved into the wood...
So we started cutting again in the Acrylic and once the piece was finished we took it out and Guess what this time it was 1/16th of an Inch Bigger in width... I mean it had already been setup and carved into wood once the only thing I did was Edit the GCODE for Feedrate speed but no mods to anything... This happened to me in MACH 2 couple of times any ideas?? I will appreciate any help!
Regards,
Roberto Soriano
-
Robert, it sounds like you may have been skipping steps to start with, compensated for it in the drawing and now it is too big. That or possibly you have a mechanical slip somewhere. Were both parts in the same place on the table when you tested them?
Brett
-
Hi,
Yes actually the Wooden part where we carved the first run was where we placed the plastic to cut it out we didn't even move the ZERO it was just right after we cut the first run we started the second one in the Acrilic that is why is so weird...
-
Hi Roberto,
As Brett has already said it sounds like missing steps to me. Remember that lost steps in one direction will become gained steps in the opposite direction. I suggest that you check for any binding on the axis travel as well as the leadscrews.
Tweakie.
-
Hi,
Yeah we have checked the machine and the screws are fine for all Axis... I still don't understand what you mean by Missing steps?
Thanks.
Roberto
-
Hello,
By missing steps they mean that your step motors cant keep up with the load, and they loose steps.
Were your cuts in acrylic more aggressive than in wood? (higher feed, bigger depth of cut?)
Fernando
-
Now that you mention it I think that I cut with faster speed rates and feeds and also the depth of cut... Yeah that might be the problem. So you suggest cutting with lowe depths and Feedrates? Should that solve the problem?
Regards,
Roberto
-
probably best dropping the accel and velocity a bit in motor tuning if your motors are missing steps.
Hood
-
You can try what hood suggested, since that is what you should really aim for, to limit accel and velocity rates in tuning so that your motors never miss steps.
But yes, that is what i suggested lower feedrates and depths of cut. Stepper motors lose their torque when increasing their speed. so if you are making a faster cut, you would have less available force to make that cut, and if on top of that the depth of cut is bigger, you need even more force for that.
Fernando
-
Hi,
Ok I did the following:
- Checked measurements in AUTOCAD
- Erased the previous G-CODE files
- converted once more the DXF files to G-CODE
- Reduced the Feed rate and the depth of the cut...
- Cut the part again and I got 1/16 error again?
Could it be a software issue? Maybe Autocad?? Th thing is my Height is perfect but the lenght is what is giving me all the trouble... And a part previously cut is perfect and did it with the faster feed rates and depths??
Ideas? Thanks guys by the way!
Regards,
Roberto
-
might be a stupid question, but i might as well ask it. Are you compensating for tool diameter?
Post your gcode so we can check it.
Fernando
-
Yes I am compensating, Actually I draw the offset within AUTOCAD so it is compensated for the tool it is weird casue one side is larger... We have 2 programs one that cuts holes for the parts and the other that cuts the actual part itself... Still don't know what is worng it does the same thing for the holes? it throws me a hole with the same 1/16th error? Weird thing is that with other files there is no problem only this specific one??
Regards,
Roberto
-
Might be the loosing steps thing.
Post your gcode
Fernando
-
Just another thought - when you are converting your DXF files and using tool offset, if the AutoCad lines are not all joined and a small gap is encountered the offset can revert to the other side of the line. (I know what I mean here but I am not sure I have explained this very well). I think it is called joining polylines.
Tweakie.
-
Good observation tweakie, I did understand you ;)
But i think by offsets he meant just drawing the offset in autocad, not using tool offsets within mach.
Fernando
-
Yeah I'm doing the offset within autocad... Basically if I draw a square I draw another one that is 0.15 bigger for tool compensation... Sorry I stayed Jurassic this way but on MACH 2 i did it that way... How you set the offset of 0.15 in MACH 3?? I did check the file and the lines are joined...
-
Are the DROs in Mach showing the true sizes?
Hood
-
DROS??
-
Digital read outs... its the place where the axes positions are displayed. X, Y, Z, A, etc...
-
These
-
Well they are showing the true Sizes... the 3 Inches in lenght of the bigger part being cut is done perfectly is the width the one that is giving me all the trouble, the part measures 7 15/16 inches x 3 inches ... Will post the Code in a sec not in the machine right now but they are going to send it to me.
Regards,
Roberto
-
Ok here is the code:
N1 (File CAJAVER2PERFIL )
N2 (Default Mill Post)
N3 (File Posted in Mill Mode)
N4 (Wednesday, March 25, 2009)
N5 G90 G80 G40 G91.1
N6 G0 Z1.0000
N7 X3.9762 Y-0.0985
N8 M3
N9 G1 Z-0.0600 F8.00
N10 X-0.1425 F8.00
N11 Y3.2015
N12 X8.0950
N13 Y-0.0985
N14 X3.9762
N15 G0 Z1.0000
N16 X8.0950 Y3.9948
N17 G1 Z-0.0600 F8.00
N18 Y5.0135 F8.00
N19 X-0.1425
N20 Y3.9948
N21 X8.0950 Y3.9948
N22 G0 Z1.0000
N23 M5 M9
N24 X0.0 Y0.0
N25 G40 G80
M30
Regards,
Roberto
-
ON the axis giving the error Double check the coupler where the motor Mounts to the Ball screw shaft. MAKE sure they are TIGHT.
(;-) TP
-
What diameter cutter are you using?
Fernando
-
Your code is giving you these dimensions. The Y axis is moving a total of 3.3inch so if you have a 0.3 dia cutter(I use metric but still seems a strange size even for imperial ? ) it will be exactly 3 inch however the X is moving a total of 8.2375 so with a 0.3 dia cutter you will end up with 7.9375.
Hood
-
Roberto,
How overeall width of the piece that you are cutting.
Can you post the gcode of the piece? In fact, one each with the different feed rates.
Don't think Autocad is your problem.
What program are you usiing to generate the gcode?
RICH
-
Robereto,
I assume this is the piece.
Looks like it's cut just the way you coded it. The width is 7 15/16" (7.9375").
Red is cut piece, yellow are z moves ( circle is 0.300 cutter), dimensions in white
RICH
Hey Hood,
All my re-sharpened mills are weird sizes post resharpening them. ;)
-
Hi Guys,
Yes the 0.3 difference is the tool offset so as Rich says those are my measurements so they are good... But I am getting 1/16th inches added there somehow... Gcode is generated by LAZYCAM.
Any ideas? Will try checking the machine but all looks perfect!!
Thanks
-
Sorry I need to read properly, I was thinking the part was meant to be 8 inch and it was cutting 7 15/16 . I gotta learn to read :(
Hood
-
Following up on TP's posting you could also check for backlash in the leadscrew, leadscrew moving in the thrust bearing, leadscrew nutplate loose in it's mounting or anything else that has worked loose on the axis drive chain.
Tweakie.
-
Will check the machine and let you know... It is properlly coded so I guess it is the machine!!
Will let you know.
Regards,
Roberto
-
Mount a ball point pen on the gantry, remove any z moves from the code. run the program a few times
onto a piece of paper. No mess this way and you'll have a visual on repeatability. Then address what seems bad.
Just a economical thought.
RICH
-
Hey Guys! It is working!! I dissarmed the axis and set it up again and it is working we cut the parts today!!
Thanks for the help!!
-
;D
Tweakie.