CoreXY setup problems with movement
-
I guess it's not impossible that the Y driver has a fault that causes it to step too much. If so, that would be the first time I have heard of that fault happening. The way to test it would be to connect the Y motor to the E1 motor output, and put this command in config.g:
M584 X0 Y4 Z2 E3
The M584 command must come before any M906 and M350 commands.
I will hopefully get a chance to try this also tonight. I will post the results as well as the other suggestions.
Looks like I’ll be busy testing if I can get some time to get to it.I appreciate all the suggestions and help from everyone with my problem. It’s not the way I wanted this to go with the new board, but having the support sure does help.
-
It is possible that you have a defective stepper driver but I suggest you run the checks of motor movement that I linked to in my first post above, just to make sure that it isn't a configuration issue.
I went through that before joining the forum as part of the troubleshooting.
The one thing I didn’t do was to measure to see if it was actually diagonal by measuring. I was looking for general direction movement. Also when commanding just one axis to move, as in X +10, the X and Y both had movement.
I'm just a tad confused by that. When you run the G1 S2 commands that I linked to, they test each individual motor so movement should be at 45 degrees. You shouldn't need to measure anything to see if it's diagonal as it ought to be obvious. So when you run G91 then G1 S2 X10 F3000 what do you get in terms of X and Y movement (direction that is, not distance)? Also, what X and Y movement do you get when you run G91 followed by G1 S2 Y10 F3000?
I will have to reinstall the Duet Board to do it again.
I might not have time this evening to uninstall the ramps setup and install the Duet board and test. I will try though and report back here.
At least I was able to use my printer last night with the ramps setup, and see that mechanically, it is functioning properly.
I really hope it’s something simple in the programming and not a board issue. The web interface is nice and the motors are quieter.
Even the first board I used on this printer, a cheap Anet V1 board, worked great. It’s the parts printer I used for some parts.
Redid the testing on from the link.
I do get each motor turning like it is supposed to. I can clearly see only one motor turning and it goes diagonal.When I run G1 S2 Y10 F3000, I get +X -Y movement.
-
I would also point out that you should also delete your Config-overide file and maybe comment out the M501 at the end of your Config.g until you have this sorted?
Doug
I don't see the Config-overide file in the list of files through the system editor, plus I did comment out M501 for now.
-
I guess it's not impossible that the Y driver has a fault that causes it to step too much. If so, that would be the first time I have heard of that fault happening. The way to test it would be to connect the Y motor to the E1 motor output, and put this command in config.g:
M584 X0 Y4 Z2 E3
The M584 command must come before any M906 and M350 commands.
I don't know if I should be happy that I found the problem, or pissed off that the board is the problem, lol.
I tried your suggestion and I had the same results, nothing different. So common sense says, if the Y-axis didn't make any difference, try the X axis.
With the X Axis connected to E1and setup, everything is working exactly like it's supposed to. Except for the endstop, but I wasn't too concerned about the endstop just yet. Finding the root of the problem was what I was after.
You mentioned it would be the first time you see this, well of course, I had to be the first, haha.
Now that we figured out that the X-axis has some kind of problem with it on the board, what is my next step to get the board replaced?
-
I'm sorry you appear to have received a faulty board. Our current test procedure doesn't check the speed of the motor outputs. I'll add this to the requirements for the automated test equipment we are developing.
If you purchased the board direct form duet3d.com, see https://www.duet3d.com/warranty. Otherwise, please contact your supplier.
-
I'm sorry you appear to have received a faulty board. Our current test procedure doesn't check the speed of the motor outputs. I'll add this to the requirements for the automated test equipment we are developing.
If you purchased the board direct form duet3d.com, see https://www.duet3d.com/warranty. Otherwise, please contact your supplier.
I purchased the board from the authorize Canadian supplier, Spool3D. I will contact them and see how they want to handle it.
-
Contacted the supplier this morning, they have already shipped a replacement and sent me the tracking number. I also filled out the warranty form as they requested.
In about a week or less I should have the replacement and we can try again, lol.
-
I little update.
I received the new board today. Installed it, ran through the configuration, tested the X and Y movement. It worked as expected. I tweaked the Z steps, calibrated the extruder and ran auto PID tune on the hotend and bed. That’s further than I’ve ever been with the first board.
Figured I would load a calibration cube that I had printed on that printer previously with the ramps board.
Z axis wants to stay at the Home position. I can manually change it by 1mm while printing, but as soon as it gets to the next layer, it goes back to home. I’ll do some searching and see what I can figure out tomorrow.
-
Are you able to jog the Z axis using the buttons in the web interface after homing, and does it move in the correct direction?
Bear in mind that on a machine with the bed moving in the Z direction, the Z up buttons should move the bed down, and vice versa. This is because +Z means "increase the distance between the bed and the nozzle".
-
Yes. After homing, using the web interface, I can jog Any of the Axis and they move in the correct direction. I figured if something was backwards, jogging it would show it. Also, when it goes to the next layer, it’s not like it tries to move the 0.2mm that it’s supposed to according to the file, it’s backwards, plus it moves till it hits the end-stop.
-
That sounds really weird. Can you post a few lines of the gcode file you are trying to print, showing the G1 Z move on layer change.
-
Please confirm that you read and understood this bit of my previous reply:
Bear in mind that on a machine with the bed moving in the Z direction, the Z up buttons should move the bed down, and vice versa. This is because +Z means "increase the distance between the bed and the nozzle".
-
That sounds really weird. Can you post a few lines of the gcode file you are trying to print, showing the G1 Z move on layer change.
I will try that if I get a chance tonight after work.
-
Please confirm that you read and understood this bit of my previous reply:
Bear in mind that on a machine with the bed moving in the Z direction, the Z up buttons should move the bed down, and vice versa. This is because +Z means "increase the distance between the bed and the nozzle".
On the Octoprint web interface with ramps setup, UP did make the bed move down.
On the Duet interface, when I hit +10-Z or + 100-Z, it does move the bed down by how much I command. The opposite is true, when I command -10mm-Z it moves the bed up 10mm.
When I home the Z axis, it does move up till it hits the end stop, as it’s supposed to.Could still be something I setup wrong, but I’m sure it will get figured out.
-
Curiouser and curiouser. So moving the head manually moves the Z axis in the correct direction. i.e. a negative move goes toward the nozzle and a positive move goes away from the nozzle. Likewise when you home the bed, it moves towards the nozzle then triggers the switch. But, what you are saying is that when you print, Z moves are reversed. It doesn't make sense.
Unless…...If for some reason the gcode file contains relative Z moves but the Duet is expecting absolute Z moves (as set up by G90 at the start of config.g). Then a relative move of plus 0.3 from say a height of 1.8mm will send it to the absolute position of 0.3mm from home instead of 2.1mm (1.8 plus 0.3).
One quick question. When you move the head manually and do say Z+10 which should move it 10 mm away from the nozzle, is that before or after you have homed the printer? If it's before you've homed it, can you home the printer then try Z+10 again and see if you get the same behaviour? Just wondering if there is something amiss with your homing files that changes anything.
-
Curiouser and curiouser. So moving the head manually moves the Z axis in the correct direction. i.e. a negative move goes toward the nozzle and a positive move goes away from the nozzle. Likewise when you home the bed, it moves towards the nozzle then triggers the switch. But, what you are saying is that when you print, Z moves are reversed. It doesn't make sense.
Unless…...If for some reason the gcode file contains relative Z moves but the Duet is expecting absolute Z moves (as set up by G90 at the start of config.g). Then a relative move of plus 0.3 from say a height of 1.8mm will send it to the absolute position of 0.3mm from home instead of 2.1mm (1.8 plus 0.3).
One quick question. When you move the head manually and do say Z+10 which should move it 10 mm away from the nozzle, is that before or after you have homed the printer? If it's before you've homed it, can you home the printer then try Z+10 again and see if you get the same behaviour? Just wondering if there is something amiss with your homing files that changes anything.
Possibly the G-code file movement is different. I will investigate when I get a chance.
Z+10 works before or after homing.
Another thing is while printing, I added +Z movement from the web control and it did move, but on the next layer, Z moved all the way home, not 0.2mm as per the G-code. -
I think we need to see the GCode file you are trying to print.
-
I think we meed to see the GCode file you are trying to print.
I agree. I will try and get that posted as soon as I can. It has to be something simple, because with this replacement board, everything else was setup nice and quick and it all works.
-
Quick update.
I had the power off during the day while I wa# at work. Decided to verify the homing files for G90 and it was there. Verified the Gcode and it was also there.
I decided to upload the file again and it worked properly. I do have some extruder tweaking to do, but it’s working.I’m going to call it some kind of glitch from the setup and changes I made, and a complete power down cleared it up.
If there ends up being a problem, I’ll dig into it again, but as for now, I can’t fix what isn’t broken.
This weekend, I will try and get it tuned up and see if it pops up again.
