« Back to “Debugging a Smoothieware bug”

Comments for “Debugging a Smoothieware bug”

This is something I noticed a year back on the oscilloscope. If you run a CNC program which has small, tiny lines(blocks) at relatively higher speed, soothie doesn`t execute it properly. Even with closed loop drives.

— Kunal on 13.05.2015 at 01:18 (UTC) 

Thanks for publishing this information I am finding it very helpful, dealing with similar issues trying to add additional axes of operation to smoothieware.  That has actually gone pretty well except for playing back gcode which seems to hang after about 10-20k worth.

  It's a shame your fix wasn't better received I think it's a good idea.  Seems to be SOP though, recently I submitted a fix for a dreaded 'simplify3d exported files cause random crashes in smoothie' bug, fixed it completely and they shot it down because apparently it is a feature for F0 to be interpreted as 'move at rate 0, so never ever finish this move'.  Then they told me they already fixed the bug anyways even though I could test and see that it was very clearly in effect before my fix and gone afterwords.  In case anyone else is wondering, that particular problem is F0 being specified with g0/g1 commands.

  Admittedly that's ridiculous gcode and should never be produced (looking at you, simplify3d), but it seems unreasonable to let it crash/hang/freeze the firmware.

— Matt on 08.07.2015 at 23:14 (UTC) 

Hi Petteri,

an amazing analysis! Great work! I just begun to hack the Smoothieware Firmware. I was wondering, is this issue still in the firmware or was it fixed some time ago?

many thanks.
best regards

— Chris on 14.01.2016 at 10:34 (UTC) 

Yeah, the bugs are still present in the original branch:
https://github.com/Smoothieware/Smoothieware/issues/635
https://github.com/Smoothieware/Smoothieware/issues/617

I have my own fix here, but the mainline smoothie project didn't want to use it:
https://github.com/PetteriAimonen/Jerkware/commit/5f4ee095a5c031207e609d2816e2f2aeabe54e92

— Petteri Aimonen on 14.01.2016 at 10:40 (UTC) 

thanks for the quick reply! I am working on a fork of Smoothieware for my fabbster 3D printer. Very simlar hardware, same processor. https://github.com/acurana/Smoothieware/tree/port/fabbster-board-rev-1.1
If you have a GCode file, so that I can easily check if it is occuring, would you mind sending it by email?

— Chris on 14.01.2016 at 12:03 (UTC)