Hi all,
Building on previous issues I seem to having recently, I am still struggling to get any code to run correctly, nothing has changed software or hardware wise but the x and y axis seem to have a mind of their own.
I managed to get a video this time as its difficult to explain, the operation you see should be a 2d contour of a roughly rectangular shape with heavily radiused corners, it seems to look ok until it hits the first corner then it just has a meltdown.
https://youtu.be/i-yBlTxEtzo
There is an operation before this cutting a couple of shallow slots using the exact same 2d contour settings which runs fine.
Most basic checks have been done, cards formatted, stepper temp ok, continuity checked, code looks fine in NCviewer and Fusions preview, also changed to Guffy’s post processor as suggested in previous post.
The issue has to be somewhere between the sd and the Ramps, it feels like its just missing 90% of the gcode and then jumping to the random coordinates it does get, but I don’t have any parts spare to just swap things out and see if it works, any help or advice would be appreciated.
Did you put the origin in the same place as the other op? Only time I’ve seen this behavior before is when running negative coordinates through the dual endstop firmware. It can do all kinds of bonkers crap.
No i don’t think so, it was part of a setup of 3 operations, to cut out the slots, then the contour and a finishing pass, so the origin is the same for all of them isn’t it, as its set in the setup not the operation itself.
Yes, if they’re all in the same setup, the origin will be the same. It was a long shot, but I had to ask because I haven’t seen that kind of motion any other way.
If path 1 follows the same contour as path 2 and they are in the same setup, I’m stumped.
I sure hope somebody else has some ideas.
I appreciate the input, i’m all out of idea’s apart from rip out the brains and start again so anything is step forward from where I was.
Maybe its a fault with the lcd board as the SD is being read from there, but I can’t rule it out until I check with a known good one, I’m just hoping someone will pipe up and point out I’ve done something wrong in fusion and avoid buying a load of new hardware.
jeffeb3
August 2, 2020, 3:08pm
7
Can you try with v1pi or repetier host?
jeffeb3:
Can you share the gcode?
I certainly can, but i don’t have a pi, which I know is shameful, but its is the truth
Code
;Fusion 360 CAM 2.0.8624
; *** START begin ***
; *** SECTION begin ***
; *** SECTION begin ***
; *** SECTION begin ***
; *** STOP begin ***
I put it through NCviewer and, to me it looks fine, but obviously something is being lost somewhere.
jeffeb3
August 2, 2020, 4:01pm
9
It does look fine in ncviewer. I have not idea why the G2/G3 commands might be giving you trouble, but they are still in there. I thought the workaround before was to take those out?
We can look closely at the firmware you have, and maybe update it to fix the arcs. Or we can focus on the LCD and try to come up with another way to send gcode (to eliminate noise as the source). Or we can focus on CAM to remove the arcs.
jeffeb3
August 2, 2020, 4:02pm
10
I wouldn’t think dumping the electronics and buying new stuff is going to help anyone until we know where the problem is.
1 Like
Sorry, yes you are right, but removing the G2/G3 commands only stops the wild movements, the code is still not being communicated to the steppers correctly,
I took another vid of just the G1 commands and it just moves back and forth in x, but it does move slightly in Y maybe an 1/10 of a rotation under power, so its not like its completely dead and the multiple z movements are new, no idea what that is.
https://youtu.be/x4CDiwVXgPQ
I didn’t mean to sound overly dramatic or defeatist when I say replace the electronics, It just seemed to me that with the code being ok and the cabling and steppers all testing ok, surely that only leaves whats lies in between, a comment in the last topic mentioned that re flashing Marlin was unlikely to make any difference, so in my head that only left the board or the lcd at fault somehow.
I have an SKR 1.3, with a touch screen in a project 3d printer, is there any reason I can’t strip that out, swap the drivers over and try that?
Also thanks for the help, I don’t think I mentioned that before.
jeffeb3
August 2, 2020, 5:26pm
12
Can you post the gcode without the arcs?
jeffeb3
August 2, 2020, 5:27pm
13
What controller do you have now, and what firmware version do you have?
jchaager
August 2, 2020, 5:38pm
14
Have you double checked all the stepper wiring connections? I have had problems in specific parts of the axes range of motion due to flexing of the connectors causing intermittent breaks in the connections and stuttering in the movement.
Sorry, distracted by dinner and kids.
Marlin-MPCNC_Ramps_T8_16T_LCD_32step_DualEndstop
I guess I had better check this, but putting a “;” in front does make it invisible doesn’t it, as in I just changes all the G2/G3 for ;G2/;G3, just in case I made a rookie error?
Code
;Fusion 360 CAM 2.0.8624
; *** START begin ***
; *** SECTION begin ***
; *** SECTION begin ***
; *** SECTION begin ***
; *** STOP begin ***
I think so, I cant say 100% there is not a loose connection, I pulled each cable and checked continuity on each wire, also you can jog the machine faultlessly over the whole axis.
But a good call, thanks for the suggestion.
jeffeb3
August 2, 2020, 7:37pm
18
It removes the arcs, but it doesn’t replace them with straight lines. So it won’t work right, it will just skip the arcs and move onto the next line.
I think you should try to update Marlin. Sorry if I suggested otherwise a while ago. That might fix the arcs.
The ramps build hasn’t been keeping up, but the next Marlin config does have it back. You can try the new stuff here:
Specifically:
https://github.com/V1EngineeringInc/MarlinBuilder/releases/download/502/V1CNC_Ramps_Dual-2.0.6-src.zip
ok, great I will do that asap, hopefully I will get time in the week, but on the weekend if not and I’ll let you know how i got on.
1 Like
jeffeb3
August 2, 2020, 7:53pm
20
No rush from me :). Let us know how it goes.
