The whole point of the exercise is to learn how to use LCNC and the CAM software. If I didn't have the Mill then I wouldn't need to send anything to the laser shop in the first place.
I held the work firmly this time and ran into the same issue again. Too many photos so I created this document. I thought I posted it to a member but not to the group. http://www.autoartisans.com/mill/Creating_a_solenoid_mount_with_LCNC.pdf Also last night I replied to a personal email but not to the group. Here's the gist of what I've found: It doesn't look like it's the G-Code or the material. I run the profiling G-Code after a Home All. It finishes at the tool change position. Then G0 X0 Y0 I find the X is now at -0.005 and Y at -0.001 as reported by my Shumatech DRO-350. Then I change tools a few times and each time back to G0 X0 Y0 and it's still at -0.005 and Y at -0.001. Run the whole G code profiling code again again and now G0 X0 Y0 has X when it's supposed to be 0.000 and was -0.005 is now -0.036 . . . In fact, a bit more work just now shows that the X axis is losing position in the positive(?) direction. So on each return to X0 after a bunch of jogging back and forth (even without backlash enabled) shows as high an error as -0.156". And it's cumulative increasing in the negative count direction. That wasn't there when I first changed from the US Digital Encoders to the CUI encoders. So either the encoder is slipping on the shaft or the way the wires are run is creating noise that changes the encoder values. I did try to use the STMBL drive instead of the HP_UHU with Henrik Olsson's processor module but could never really get it tuned as well as the HP_UHU. The toothed belt pulleys are keyed to the shafts. The only non-keyed item is the CUI encoder. I suppose if it was slipping that could cause the problem. But more likely noise. But then why not random enough to cause positive errors too. In either case, jogging back and forth starts changing where it thinks X0.000 is based on the Shumatech DRO position report. And that would cause the incorrect milling. The Y axis is stable. In fact, the HOME all and then move to X0 Y0 (G54 coordinate 0,0) is the same every time and matches the Shumatech DRO. Until I start jogging the X axis back and forth. I removed backlash, slowed max speed to the same as X and reduced acceleration. Still errors. I need to connect to the HP_UHU serial port and start up the terminal program to see what the drive reports. Luckily, other than the issue of where to put it and the associated wiring, I have an Bergerda AC servo I could swap in place of the DC Servo. And it takes up less space. Way less space, same RPM, same Torque. Controller is larger though. So I may try that first to eliminate hardware. Which was the whole point (expensive) of replacing the US Digital encoders with CUI encoders. John > -----Original Message----- > From: Roland Jollivet [mailto:[email protected]] > Sent: August-04-21 8:19 AM > To: Enhanced Machine Controller (EMC) > Subject: [Emc-users] Something went wrong. > > I would send the .dxf to the laser shop and collect the cut-and-bent parts > in a few days. Some parts are not worth machining when there's a far easier > practical solution. > > Roland > > > > On Tue, 3 Aug 2021 at 23:31, John Dammeyer <[email protected]> wrote: > > > That's a good suggestion John F. Thanks. I did do some more playing > > around and it's clear the part flexing and the backlash both were at fault > > especially with the plunge to the next depth. > > > > And that brings up another issue. One of my pet peeves with electronics > > project magazines is they are great at a schematic and either point to > > point wired or PC board but very little energy is spent on describing > > various ways of mounting or installing in a cabinet. Especially with the > > concept of Human Factor Engineering which is the practice of making > > something easy to use or even intuitive. Like an ESTOP button is always a > > red mushroom Not a toggle switch. > > > > Same goes with work holding. Youtube has tons of videos that show a > > tiring sequence of a milling cutter sprayed with coolant making chips for 3 > > minutes with 3 commercials interjected, one every minute. But very little > > on work holding. > > > > The next part I am making is shown in the attached screen shot rendering. > > My raw material is in the second photo. So the question is about work > > holding and how or what features of LinuxCNC can be used to make this > > easier. > > > > I can use my band saw to create the initial width and split it into two L > > shapes. But after that I start to have problems, due to lack of experience > > I think, on how to firmly hold it and mill the stuff with a 5mm and 6.35mm > > (1/4") cutter. > > > > Suggestions? > > > > Thanks > > John > > > > > > > -----Original Message----- > > > From: John Figie [mailto:[email protected]] > > > Sent: August-03-21 11:59 AM > > > To: Enhanced Machine Controller (EMC) > > > Subject: Re: [Emc-users] Something went wrong. > > > > > > Speaking of backlash. My tiered old Bridgeport has about 0.002" of > > backlash > > > in the ballscrews. So if circular interpolation is used there is a small > > > bump at each 90 degrees of the circle. But if I make the finish pass > > first > > > clockwise and then repeat counter clockwise the imperfections are much > > > smaller. I know from experience with my first CNC machine that's built > > that > > > actually used leadscrews with lots of backlash the clockwise followed by > > > counter clockwise method was remarkable compared to a single direction > > > final pass. > > > > > > John > > > > > > On Tue, Aug 3, 2021, 10:10 AM jrmitchellj <[email protected]> wrote: > > > > > > > You might try the run again without the backlash compensation in LCNC > > to > > > > get a feel of what it is actually doing. > > > > > > > > > > > > --J. Ray Mitchell Jr. > > > > > > > > > > > > ?I predict future happiness for Americans, if they can prevent the > > > > government from wasting the labors of the people under the pretense of > > > > taking care of them.? > > > > > > > > THOMAS JEFFERSON > > > > > > > > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
