Thread: Machine Problem - Rockcliff
I'v got an issue with my machine, i cant work out what it is thats giving me this issue but im hoping you will know or know somthing i'v missed off my check list.
Ok i'v checked:
1. Measurements in my drawing
2. Toolpath settings in my drawing
3. Tool settings in my drawing
Ok so thats what i have checked pre Mach3.
In Mach3 i have checked:
1. Motor tuning (1280, 200, 50), drivers setup for 1/8th step
2. Zero All Axis's, REF All Home updated to new zeros
thats what i have checked pre machine (Rockcliff)
On the machine i have checked:
1. Rails + Bearings (levels, square's)
2. Drivescrews + Nuts (delrin drive nuts all nice and tite, leadscrews dont seem to have any issues.)
3. Checked Calibration (machine moves 20mm psyically when moved/jogged 20mm on the screen for mach3 (DRO?) )
My problem is:
The part is 3mm under size all round, the center hole should be 43mm but its coming out at 44mm (note this is only 1mm error), the bottom of the part (bit that fix's to the machine) should be 80mm across, its coming out at 77mm.
Any ideas guys? I thought maybe i'v not set the right end mill size in the design but thats set to a 3mm End mill and that is what i have in my spindle.
I'm telling the toolpath to cut on the "outside" of the vector, i really cant think of anything else i can check.
Please any info i can provide you with please ask and ill reply more or less instantly with an answer. The GCODE is posted below if you would like to look at that.
Cheers, Lee :confused: :( :( :confused:
Last edited by Lee Roberts; 15-02-2009 at 08:14 PM..Me
Just out of interest how accurate are your measurements?
The error you have for the centre is less than for the base - approximately you've got a 2.5% error for both.
(If) the error is really getting larger as the machine moves it suggests that there's a scaling issue.
However you say you've measured a jog of 20mm and measured 20mm displacement (bearing in mind that an error or 2.5% would only be 0.5mm error).
I'm not that familiar with G codes, I assume there's no scaling in it, tbh I simply use whatever comes out of the Vectrix software as is and never have any issues.
If you have any more info that might help narrow it down, but I'd start by getting as accurate a measurement as you can, that way we can decide if the error is a scale or a displacement. If it's a displacement it's likely to be a tool, a scale could be software or setup in Mach3.
Worth bearing in mind also that an end mill always cuts slightly oversize.
How many passes do you do to cut out the part?
If you use multiple passes can you see any steps in the part or does it look ok?
looking at the gcode, it doesn't look like you run a finish pass.
You'll get tooling flex so you should have the software leave about .5mm or less
for the finish pass, it can take the last pass at full depth and it will clean up any
ridges left from the multipasses during roughing.
Something else to check - is the circle round or oval? It looks very much like you have set the X axis up with the wrong data. If the circle is round then the Y axis is also out. You have not said whether the piece is the right height in the Y axis. Apologies - just looked at your post again and you say it is 3mm out all round. As mentioned by Kwackers, some accurate measurement is needed to determine if the error is cumulative. i.e. gets larger the further the axis travels. If this is indeed the case, you will need to double check your leadscrew pitch, stepper steps and steps per mm.
Checked the code and the inner circle is 40mm and the outer block is 83 so they are right given a 3mm tool.
Got to be scaling, setup or machine flex.
If the centre hole is larger than it should be, but the outside is smaller then it can't be a scaling problem. Is the machine loosing position under cutting load conditions. Try taking a very fine cut or try modling board or something soft to machine to see if that is OK. Also check calibration over at least 200mm movment.
It almost sounds like its programed to cut center line instead of edge..... but if that were true then your inside would be just as off as the outside and you've already checked that.
I'm wondering if your not losing position under load like captainresonator has mentioned. You would think though if that were the situation though you'd not have a uniform error.
What size of guide rails did you use?
What type of lead screws did you use (metric/inch)
What units are you using? (you answered that already = mm)
What CAM tool are you using? ie post processing of CAD into GCODE
What are you using for the spindle motor? (how heavy is it?)
Can you post photo of the machine? Showing the x-z axis mount in particular.
My hunch is that there is a lot of movement in the tool in the Y plane. This is due to over large tolerances in the bearing blocks (I have the same problem after 18 months) and the X-axis rail mounts. The MDF bearing blocks compress with vibration and lubricating oil thus allowing for a very large movement in the tool typ. 5mm
So to prove the hunch ... park the spindle in the middle of the X plane. Put the tool into the spindle motor. Drive the tool down to just above the table. Turn EVERY thing OFF. (or risk loosing fingers). Pull on the tool and see what movement there is and what force is needed to move it. Look for flexing in the guide rails, movement in the bearing blocks, look for sleeves falling out of their mounts (oh yes!).
where the light was kept in the Dark Ages
By dragon in forum Boxford LathesReplies: 1Last Post: 28-12-2014, 08:27 PM
By Eugene in forum Computer SoftwareReplies: 2Last Post: 27-01-2011, 06:00 AM
By xchipx in forum Heiz CNC RoutersReplies: 2Last Post: 03-04-2009, 12:08 PM
By Steve-m in forum Artsoft MachReplies: 9Last Post: 12-01-2009, 04:38 PM
NEW MEMBER: RockcliffBy Rockcliff in forum New Member IntroductionsReplies: 3Last Post: 05-08-2007, 01:32 AM