@David the swarfer, I have problems with arcs on my machine. I use Mach3. I think I had, a few years ago, a similar problem when I had my PhlatPrinter III. I now use different breakout board and driver. I can't recall what the solution was then and/if it was the same problem. That's what it should look like And that's what it looks like. The problem is on the Y axis.
Hiya This is a weird one (-: My first thought was about incremental vs absolute IJ mode. GRBL only does incremental so all the code is written in that mode, but Mach3 can do absolute mode as well though you wojudl normally get a mismatched radius error if it is incorrectly set. You can turn on the output for the incremental setting in the Feature Options, which will make sure that your Mach is in the correct mode for every file you give it. or you can use one of Mach's many menus to change the settings, or give an MDI command of G91.1 to set it. I ran your code through my machine (bCNC + GRBL) and it ran perfectly so it is definitely a Mach3 thing, though I don't think mechanical even though I do see 2 cut tracks on that lower arc. That arc segment that failed has a J offset of 0.0027mm so I am wondering if Mach is correctly reading all 4 decimal digits? Does Mach have an error log we can look at?
Thanks, David. I think I found the answer though. In SketchUp I exploded the circles, after I increased the segment to 72 from 24 and generated the gcode again. This time with no arcs and I had the same problem. I ran the machine again, but this time faster and I had no problem with it. I played with other gcodes and continued to have problems. I got to the sad realization that I have problems with the stepper driver. Yes, I'm using the TB6600 drivers. I didn't use the driver board form my PhlatPrinter since it has only 3 drivers on it and I need four for my machine. I have already other drivers coming from China. Should be here, in Canada, in about 30 days.
Well...I hope I didn't screw up again. Mine are DM542's. I saw a post on CNCZone from a someone in the UK that said he has those and has no problems with it.
Love my DQ542's. The best move a made was way back at the beginning was to get rid of the DRV8825's and switch to these. The M542's should be comparable. My motors are so quiet, cool, and happy.
Well...the DM 542's didn't help. Here is how the story developed. I'm using Mach3 I started with the Y axis having two NEMA 23 stepper with about 270oz/in. When I got the problem seen in the OP I thought that the steppers are too week for the Y axis and I changed them to 2 NEMA 34 4.2N/m steppers. The problem still persisted. I then thought that maybe my cheap TB6600 drivers are at fault and bought the DB542 drivers. Obviously the problem is somewhere else. I tried changing acceleration up and down to no effect. I tried changing the feed speed to no effect. Every time it gets to the same spot as can be seen in the OP the Y axis stalls, makes a grinding noise while the X axis continuous and then the Y axis comes back on line until it reaches the other side of the circle. I have the problem when I cut wood AND when I do air cut. Somehow Mach3 has a problem with the arc command at that spot and I don't know what parameter to reset to get it to play nice. Bottom line is I now have three steppers and four drivers that are most likely perfect for my machine but will sit in a drawer with no love given to them!
This is why I now have a lathe. I had a Blackbox sitting for a year waiting to be appreciated along with some 270 oz steppers that are being replaced with the high torque steppers on my original CNC.
But I still have to find out why I have this problem since my CNC is practically useless without tackling this problem. Anyone has an idea what setting needs adjustment?
If it is not software then it is hardware.... changing acceleration and maxspeed down will increase power, but you say that did not help, so there is a really sticky spot in the Y travel. A grinding noise indicates missed steps. Ergo Y is getting stuck. Have you tried putting the raw material further forward or back so that the 'problem spot' is a) in a different place on the cut b) not in the cut area at all? c) air cutting, ie router off and bit not in contact with anything? maybe rubberband a Sharpie to the router so you can see where it goes.
Yeah, I did all of that. Since the Y axis is missing steps on that g-code spot, rerunning the g-code without resetting Y 0 gets me a different physical position on the Y axis. Also, I can turn the lead screw by hand all the way up and down the Y axis with just 3 fingers with no problem. It is doing that at the same spot no matter where on the Y axis it is located. And it is doing it not only on this g-code, it does in on other things too where I have small circles. I thought I've mentioned that I did air cutting too but maybe I didn't. Anyway, I did air cutting over and over and I have the same problem. It must be something with Mach3 and maybe the BoB. That's the only thing I didn't change yet in all this trying to fix the problem and I know Mach3 worked fine on this computer for years with my PhlatPrinter III. The only other thing I can think off is that the slaving of the two Y axis stepper is not welcome to Mach3. That's really the only difference between the PhlatPrinter and this machine. Yes, on the PhlatPrinter I used a different driver/BoB board but I already changed the drivers on this machine. Just to make sure, the setting in Mach3 for the A and the Y axis are identical unless I'm missing some setting that is not in the Motor Setting tab.
ok, I have got it.... I think drivers need a small delay between the change in the direction pin and the next step pulse. I believe Mach3 has a setting for this.... somewhere so what I think is happening is that at that point X is reversing direction which is causing either a long or short pulse on Y, which is allowing Y to stall. Make the pulses longer What is the step/direction pulse width? and if there is a direction change delay, halve it and test, then double it and test. somethign gotta work (-: also for more info - maybe inverting step and direction will do it? Time between change of direction pin and step pin · Issue #103 · grbl/grbl
Unfortunately it didn't help. I tried all the way from 1 to 19 and it didn't make a difference. I'm working now on converting it to GRBL. I'm having problem with it but I'm sure I'll get it solved and once I do I'll see if the same code will run OK in GRBL or not.
I ran the same test program I did on Mach3 on GRBL using OpenBuild Control and it ran perfect, just as expected. Which means that as I thought, it's a setting in Mach3 that is not plying nice with the g-code. Alas, I don't know which one. I bet that if I reinstall the NEMA 23 steppers and the TB6600 drivers it will also run perfect. Well, thems the breaks. Maybe I'll find a use for them in the future, but only maybe .
