LinkBack URL
About LinkBacks
You probably need to calibrate your axes. Hoss2006 has a video on YouTube that shows how to do it: [nomedia="http://www.youtube.com/watch?v=OTHdgk12EcI"]YouTube- Mach 3 Steps Calibration[/nomedia]
bob
| | |
I'm having trouble with my 135 with differences in the X and Y axis.
For example, if I try to mill a 1" circle the X axis dimension is correct but the Y axis dimension is -0.020" less. I've written a simple program to cut parallel slots, measured their dimensions and found the same problem, ~0.020" per inch incorrect in the Y direction but the X axis dimensions are correct. I looked at the G code and the data is correct but for some reason the machined Y distances are incorrect.
I had a similar problem on another piece of equipment and I was able to adjust the number of pulses/inch the drive motors were given but I don't know if I can do that with this unit. I'm running Mach3 software.
Any help would be appreciated.
Tom
You probably need to calibrate your axes. Hoss2006 has a video on YouTube that shows how to do it: [nomedia="http://www.youtube.com/watch?v=OTHdgk12EcI"]YouTube- Mach 3 Steps Calibration[/nomedia]
bob
Bob,
Thanks for the tip. The videos were very informative but when I did the setup he outlined I found there were both pretty close (+0.002 for the X and -0.001 for the Y).
What I did find more interesting is his video on backlash. He showed a cut that was supposed to be a circle but was eliptical, exactly my problem. He went on to say he adjusted the backlash but it wasn't clear how or how he measured the backlash in his sytem.
Do you have any idea how that's done?
I have to say the support from Novakon is horrible!! No responses to e-mails or voice mails and this is not the first time they've been so unresponsive. They only have one guy - I think - and he's completely overworked.
Any help would be appreciated.
Tom
Tom,
There are few things that contribute to backlash that you can check quickly:
Ensure that the coupling is clamped tightly to the motor and the ball screw. If there is a bit of play there it will manifest as backlash.
Ensure that the bolts that mount the ball screw to the table or saddle are tight and that there is no play there.
If your gibs are out of adjustment that can also cause backlash. Make sure both the x and y are properly adjusted since play on the one axis can show up in the other. i.e. if your x gibs are loose, there can be play in the y axis.
I don't have a Novakon so I can't help you with specifics but this should give you something to look at.
bob
I'd suggest doing everything already listed - that's all good starting steps to track down backlash. I'd also suggest simply going through all the nuts and bolts on the machine and ensuring they are all tightened down solid from shipping. The the first thing I'll be doing when my NM-200 gets here in a few weeks.
My X2 had horrible backlash initially and people here helped me a great deal to get it all sorted out. It's also come loose a few times though use and crashes and had to be re-adjusted. I first noticed the problem the same way you did - not-quite circular circles.
With that said, measuring backlash is pretty simple - mount a solid edge on your table somewhere. I just use the edge of a toe clamp but most people use a vise. Chuck up a dial test indicator in the spindle. Pick an axis to test - let's assume X for this. Move the machine via your controller software (I'm assuming Mach) until the indicator is touching the edge of the vise/clamp. Don't reverse directions while doing this! Once you've bumped forward and are registering a bit on the indicator, zero out the axis in Mach. Now go the opposite direction away from the vise/clamp in steps of 0.001 or less. Keep doing it until you see the dial indicator move and by how much. Subtract what the DRO in Mach says you moved from what you actually see on the dial indicator and that is your backlash in the X axis. Repeat for Y and Z. You then go into Mach's configuration and set the backlash compensation for each axis and run the test again. It's important to do this AFTER the mechanics are all sorted out and tightened down otherwise it's all going to need to be redone.
Personally, my backlash/inaccuracy came from two things - loose gib screws on the Z axis and the little ballscrew mount that attaches under the table for the X axis. No idea what it's called, but it has a tendency to work loose on my mill over time so every few weeks when I mount a fixture I do a quick backlash test to see if it's still zeroed out. Once you have it dialed in for your machine, if it changes you can make the assumption that something has worked lose as wear shows up very slowly. In my case, it's things inevitably working itself lose and needing some TLC.
Either way, good luck to you!
-Mike
they have more than one guy, but i think they are overworked at the moment. khai is out of town at the moment.Originally Posted by tas99
theres 3 easy things to check here.
1 is for play in the Y axis ball screw mount. basically, move the axis one direction, then move back and measure how far it actually went.
if theres no backlash, the next thing is to check if its moving the right distance. there is a tool in mach3 to do this. i think its called calibrate ball screws. this will give you the correct number of steps to enter.
the last things is if you see that the circle looks "flattened" on 2 sides. if this is the case, go into ports and pins and check "sherline mode". this seems to be a issue with gecko's microstepping and pops up once in a while. i assume the control is gecko at this point, i imagine the error could happen on other drives.
For reference, IIRC, backlash on my nm135 was as follows:
X .0007"
Y .0013"
Z .006"
This was done before I snugged up the gibs, which will help things. These values seem pretty good to me; good enough to not bother with Mach3's possibly-risky-possibly-buggy backlash compensation. The Z is the least critical, unless you're doing 3D mold work or something.
the z axis backlash tends to be twist in mounts for the ball screw in the back. my little mill had alot more than that, and it was generally not a big issue - even with 3d contouring in aluminium. if you needed ultra critical precision.. well, i imagine you wouldnt be going for the budget mill segment.
I certainly appreciate all the suggestion and I'm exploring each of them.
One of my questions is how to get access to the 135s Y axis ballscrews, etc. They only sent me a "Draft" manual and there are no pictures in it so it's hard to reference. It looks like I have a lot of covers to remove in the back to get access. For those with a 135, is that correct?
How do you measure the backlash?
Even though I haven't got access to the Y assy yet everything seems pretty tight and there's little or no play in the table. I've written some short programs for simple cuts that I can measure and they error seems to always be the same; no matter which direction and which order the cuts are made. I would have thought if it were a backlash problem there would be some variation but each set og cuts is always exactly like the otheres. Thsy are wrong but exactly the same.
I don't think I'm asking for too much accuracy. I'll attache a sample circle and as you can see there's about 0.025" variance and that's not acceptable at all. I got tited of burning up aluminum so I'm testing in MDF but the results are exactly the same. I've tried that circle with a climb and a regular cut and it comes out the same. The numbers are diameter at that point.
Any other suggestions?
Tom
looks flat. try setting mach3 to "sherline mode" in ports and pins and try the cut again.
I'd try the Sherline thing as that's super easy to check. As for:
My post above - #5 - goes into detail on exactly how to measure backlash on a mill.How do you measure the backlash?
-Mike
Mike,
Thanks again. I forgot you mentioned how to do the backlash measurement and I'll do it first thing tomorrow as well as switch to Sherline.
This thing has me totally frustrated and if it weren't so heavy I'd throw the thing out the door.
But thanks anyway.
Tom
Posting Permissions
| | |
