Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: dagdaj on February 21, 2007, 12:46:02 PM
-
Good morning all. Your posts have been so helpful to me in getting me to the point where I can at least cut some semblance of a part. But now I'm stuck.
problem: circles and arcs not 'smooth'
symptoms: motor/motion appears to 'chatter' and the cut edge appears 'rough' and 'wavy'. It also appears as if a step is maybe being 'lost?' in that the geometry is off. When the cut is finished, the end of the path is not exactly where it should be.
cutting: just cutting .125" HIPS, using a good carbon O-flute bit with neutral spiral, using Mach 3
Here's what I know and have done so far:
-Cuts straight and clean along x and y axis. Only seams to be a problem on arcs or when x and y are working in rapid succession (generally).
-I did several diagonal cuts about 10 inches long and didn't noticeably see the problem.
-Then I cut a few squares rotated 45 degrees and did see the problem.
-The problems don't seam to manifest themselves in the same parts of the toolpath each time.
-The problem appears to 'muted' when I switch to Exact Stop, but is still clearly visible.
-I have tuned motors by increasing and decreasing acceleration with no noticeable results.
-I have adjusted feed rates on the cut. A faster rate appears to improve it slightly, but it's clearly still a problem.
-I've moved the cuts around the table to make sure it's not isolated to a particular section.
-Changing direction of the toolpath has no impact.
-I have checked backlash...though I don't know if what I did was right. I took a depth mic and set it up and sent the machine back and forth on the x and y and made sure it came back to the same place. I appear to be very lucky in that backlash does not appear to be a problem.
I have learned quite a bit from these forums, and have learned a fair amount about this problem through thread searches, but I'm still stuck. And then there's the element of the standoff between the mechanical, electrical, controller, and software elements. That makes it hard to sort out where to start troubleshooting.
First off...if anybody has any ideas or Mach3 settings to fiddle with...I'd be more than grateful.
But, equally important, if anyone can point me in a troubleshooting direction that could help me eliminate culprits, I'd be more than grateful as well.
-
Check to make sure that you have don't have any of the 'CV Control' helpers checked in 'General Config'. Also turn off the 'CV Feedrate' on the 'Settings' page (this is not CV/Exact Stop on/off, G64/G61 does that) .
I have had these cause some hesitation at times.
-
I'm not sure you are getting good readings from the backlash test. If you are jogging it off and then back to the mic. you are only checking repeatability. Checking your backlash should be no problem if you have an indicator. Put the indicator on something solid. Jog your axis until you move the indicator to a reading of about half of its travel. Write down the reading on the indicator. Jog the axis in the other direction by a set amount, say .010" and see how much difference the indicator shows. If the indicator says you moved .005 and Mach says it has moved .010 then you have .005" backlash.
Brett
-
Well now...now I know for sure how to measure backlash...instead of assuming I knew. 8)
I followed Chaoticone's (THANKS!!!) steps. Found none.
I checked Scott's suggestions.
I think there's something wrong with a chain or bearing or something. When I 'ref all home' the machine/motors make a tremendous amount of growling and other racket when homing the X axis. Also it seams that when I make very small jogs along the X axis...I get all sorts of extraneous growls, clicks and rattles that I don't get on the Y and it generally sounds fine when it's a prolonged cutting or jogging movement across the X axis.
When I cut air circles with no material and no power to the router, the whole thing has moments where it shakes and shimmies like Elvis on ecstasy. Interestingly enough the shaking and shimmying seams to coincide with that now-familiar growling and rattling. Then it runs smooth for a bit with no shaking and no grinding. Then back again.
I think my Dad who 'rebuilt' this thing needs to revisit some of the mechanics. I'm not feeling it as a software thing or setting.
Does that seam to make sense?
-
;D
it shakes and shimmies like Elvis on ecstasy
;D
What kind of machine is this?
Brett
-
Dagjad.
It does sound like a mechanical problem.
I guess the jerking and growling shows up more when it's cutting air, becaouse you don't have a router screaming - I would guess the machine is still jerking and growling.
If you've got this far, you can fix it. Can you post pics of your ugly circles. That can tell a lot. And get you dad round - He might spot it straight away. Fresh eyes sometimes do
Jim
-
It's a big semi-old Cam/Tech that had it's brain and some of its guts transplanted with a new motherboard (Delagrange board and various motors) and is now being run by Mach3.
We just looked at it again.
We're thinking it might be a problem with a motor. It appears to be the source of the shimmy. If we manually 'shimmy' the gantry, the drive and that motor holds steady. Easy enough to check...we've got another motor.
*clicks cross fingers emoticon*
James...oh noes...we don't want Dad around. It's me, my brother, and sometimes my dad. Ever watch American Chopper? I'm the Mikey character...albeit a little smarter. My brother is Paulie only way smarter...and Dad is clearly Senior but way smarter and not a total meany jacka$$. From the start he's been saying it's a software/motherboard/controller problem. I may not be a machine/machinist/engineer guy, but I'm a computer guy. Computers are dumb...and they're always dumb in the same way. This isn't being dumb in ways that computers are dumb.
My brothers got a few ideas, we'll try those and see where we get. If not, I can revisit with pictures.
I should collect all the archival posts that have been tremendously helpful to us in getting us this far.
There's a cr@pton of them.
-
Does your X axis have two motors, one on each side? If so, they may be in a bind. They may be trying to torque the gantry to square it up. Just a thought.
Brett
-
Update:
I've enlisted more eyes. At this point we've checked the drive/chain/belts and tension on both sides of the table.
And have been able to identify that the 'jump/chatter' appears to occur in the same general area of the path.
We have still yet to replace the motor before we have to cave.
But, does anybody know any consultants in the St. Louis area?
-
I have sent you a personal message.
Brett
-
Just posting to say thanks to all those who offered help ('specially the Chaoticone!) and to followup with the resolution on the extreme offchance another noob comes along and has the same problems. I've sorted out a ton of issues reading posts where you all have pointed towards resolution. Solved another little thing today as a matter of fact.
Anyway, we put in a board from a different manufacturer. Saw the same thing...and other worse problems. Finally put in 3 cards/controllers...one for each axis from yet another vendor. Got the same exact results. Tried using KCam instead of Mach. Same problems (magnified).
When my brother dialed down a current setting on the board (we're not totally sure specifically what he did yet) it dramatically improved. We attempted to do the same thing on board 2 as it had pots to do that. But it didn't work. The first board just had an elaborate system of resistors and capacitors that weren't very adjustable. Anyway, the setting we dropped was actually set to a much lower setting than the board manufacturer said to put it at.
Either way we're pretty sure that it was a torque thing. That seams to be totally consistent with everything we experienced. Loud motors...almost sounding like they are straining under certain conditions. As we fiddled with it more and more the problems were isolated to times when the motors were accelerating or decelerating. And...that would explain why the higher feed rates (less torque) seamed to mitigate it some. Lowering that current clearly shifted us dramatically into a place where we could find a sweet spot on that curve.
Apparently we just couldn't find a good spot on the motor tuning curve with that setting where it was.
The better news...we were able to adjust the resolution for microsteps and move from 250 steps/inch to 500...further improving our cut.
We're in business now.
-
Your welcome, you did it all though. Glad you are running. Keep us posted.
Brett
-
Your welcome, you did it all though. Glad you are running. Keep us posted.
Brett
"were running"
I think we have a problem with one of the controller cards now. Whenever the thing moves Y- it refuses to move X+. You can see Mach is giving it the signal and it starts to move...and then stops.
We flipped the pins on the inputs and the same problem manifest itself then in the other axis. Flipped those back and then flipped the output wires to the motors....same problem in that other axis. So it's not the motors. Same results on KCam.
We're starting to learn how to troubleshoot.
Good thing Sr. knows where there's a supply of these things at the electrical junk dealer. He was talking about buying them all and dealing them on Ebay. He's clearly bored in his retirement.
-
We're in business now.
I'm sorry, I thought you were runing. Can you post your XML file for us to look at?
Brett