Page 1 of 1
Stopping initial retract?
Posted: Sun May 29, 2016 11:47 pm
by lignumaqua
I'm running DC42 V1.12a (2016-05-10) on a Duet with a Rostock Max V2. All seems to be working fine. I have one question, the last line in config.g is
Which, I believe, is supposed to just set the default temperatures for the main extruder. However it also causes a retraction which messes up the first layer as the print starts with no filament.
I also tried
but it made no difference.
Is there a way to stop this G10 command causing a retraction, or am I supposed to add an equal and opposite extrude command to undo what it did? Or maybe this is indicative of some other setting I have wrong? Thanks!
Re: Stopping initial retract?
Posted: Mon May 30, 2016 12:35 am
by IMBoring25
Do you have an M563 P0 D0 H1 somewhere in config.g?
It shouldn't be interpreting the G10 command as a firmware retract unless the P parameter is missing and other than checking for the M563, the only other thing I can think of that might be confusing it would be that I've always set the temperatures (S and R parameters) and tool offset (X, Y, and optionally Z parameters) in separate G10 commands. I would expect 0,0 is the firmware default, so you might not even need those parameters.
Re: Stopping initial retract?
Posted: Mon May 30, 2016 1:01 am
by lignumaqua
IMBoring25 wrote:Do you have an M563 P0 D0 H1 somewhere in config.g?
It shouldn't be interpreting the G10 command as a firmware retract unless the P parameter is missing and other than checking for the M563, the only other thing I can think of that might be confusing it would be that I've always set the temperatures (S and R parameters) and tool offset (X, Y, and optionally Z parameters) in separate G10 commands. I would expect 0,0 is the firmware default, so you might not even need those parameters.
Thanks for the response. Actually, I think I've sent you off on a wild goose chase here.
I now think it's nothing to do with config.g and is instead a problem with my initialization code in the main gcode file. I have this section:
Code: Select all
M140 S60
M190 S60
T0 ; Select T0
M104 S205 T0
M109 S205 T0
I added the extra 'T0' because of
this problem where M109 wasn't waiting until the head reached temperature. The suggestion there was to add an explicit T0 to make sure that Head 0 was the selected tool. I suspect that the retraction is an unintended consequence of that extra T0. Perhaps Duet is seeing the T0 as a tool change (even though there are no other tools) and is therefore performing the retraction. I suspect I have a setting to turn that off somewhere in my slicer. Sorry for the misleading question. I need a single-step option in g code interpretation so I can see precisely which command is causing which action!
Re: Stopping initial retract?
Posted: Mon May 30, 2016 1:38 am
by IMBoring25
That makes more sense.
That's an appropriate thing to have in your start g-code. If you're selecting T0 when it was previously inactive, check your tpre0.g and tpost0.g files on the SD card, as the Duet will run those macros before and after the tool change, respectively.
Re: Stopping initial retract?
Posted: Mon May 30, 2016 9:11 am
by geneb
My T0 command is at the end of my config.g file - I've got the same issue where it's not heating the hot end after I upgraded to 1.12a from 1.09k.... I wonder if the T0 in config.g is being ignored.
g.
Re: Stopping initial retract?
Posted: Mon May 30, 2016 10:45 am
by lignumaqua
geneb wrote:My T0 command is at the end of my config.g file - I've got the same issue where it's not heating the hot end after I upgraded to 1.12a from 1.09k.... I wonder if the T0 in config.g is being ignored.
g.
I think the T0 has to be immediately before the associated heating command, that's how I got it to work. I had to add it to the slicer configuration starting script. (I'm using Simplify3D). The head now heats up (and waits) correctly.
Code: Select all
G28 ; home all axes
M140 S[bed1_temperature]
M190 S[bed1_temperature]
T0 ; Select T0
M104 S[extruder0_temperature] T0
M109 S[extruder0_temperature] T0
G32 ; Auto level
However, now I look at this again, the T0 happens before the auto leveling, and the bogus retraction happens *after* that. Sigh... I need to keep digging. It must be something S3D is adding at some point.
Re: Stopping initial retract?
Posted: Mon May 30, 2016 10:57 am
by Xenocrates
There are G code simulator apps for both phone and computer. I mostly use the CNC milling/lathe ones, but I know printer versions exist, and they usually have single step support. Maybe give that a shot?
Re: Stopping initial retract?
Posted: Mon May 30, 2016 12:17 pm
by lignumaqua
BTW - the move to Duet has made real improvements to the print quality of my Rostock Max V2. These are both from the same machine and same gcode file. Only difference is the Duet. Left is stock machine, right is Duet.
[img]
http://gsak.net/stats/screenshots/2016- ... -14-38.png[/img]
Re: Stopping initial retract?
Posted: Mon May 30, 2016 1:43 pm
by Aflac
I've been having a similar problem. Except mine happen at the end of a print, it completely unloads the filament at the end of every print. It did not do this on the smoothie board before I swapped and I did not change my slicer settings
Re: Stopping initial retract?
Posted: Mon May 30, 2016 4:01 pm
by IMBoring25
The discussion of config.g confused me. If it's doing it at the start of a print, that's normal. I think all the slicers assume you have the nozzle primed and do a retraction move before the non-printing move to the first printing move. Most people use a skirt to get the nozzle nicely primed again before printing the stuff that matters. If it's causing you problems I guess you could do an unretract at the end of your start gcode, but you may have to find some way to deal with the extrudate.
Aflac, do you have M82 or M83 in your config.g? If you go through your entire gcode are there any M82s or M83s in it? Which one comes last prior to the command that unloads the filament? Are there any G92s?
Re: Stopping initial retract?
Posted: Mon May 30, 2016 11:53 pm
by Jimustanguitar
I had similar oddities, but at the end of my prints.
The difference in behavior between relative and absolute positioning on the Duet got me at first. The end GCode that I've used in Cura for years will cause the extruder on my Duet machine to retract 5mm from absolute 0 at the end of a print, so it backs up for several meters to where I started and then retracts 5mm from that...
I dug into the GCode a little bit, and it seems like my Repetier electronics zero the extruder count when you switch to relative positioning, and the Duet does not. I had to add a G92 E0 command right after the G91 in the end GCode to get things to behave the same.
Note, I'm talking about adding these commands to the end GCode in my slicer, NOT adding them to the config.g file on my Duet's SD card.
Re: Stopping initial retract?
Posted: Tue May 31, 2016 2:54 pm
by lignumaqua
IMBoring25 wrote:The discussion of config.g confused me. If it's doing it at the start of a print, that's normal. I think all the slicers assume you have the nozzle primed and do a retraction move before the non-printing move to the first printing move.
Yes, that's happening but, for some reason, I'm getting a double retract which is then so long that the normal skirt isn't long enough with small parts to prime the nozzle. I can fix it by reducing the normal retract amount (I currently have it set to 7mm for PLA).
Re: Stopping initial retract?
Posted: Wed Jun 01, 2016 8:38 am
by dc42
We recommend using relative extrusion with RRF, because it avoids rounding errors, avoids issues with resetting the extruder position to zero, and is IMO more logical. However, absolute extrusion should work too, provided that any prime or retract commands you have in tool change files, pause.g, resume.g etc. have M83 commands before them (the default ones that I provide do, but that wasn't always the case).
If any of you think there is a problem with using absolute extrusion, such as double retraction, please post the section of gcode that you think is being executed incorrectly, along with your tool change files, the pause.g/resume.g if the problem occurs when pausing or restarting, and a description of what is happening.
Re: Stopping initial retract?
Posted: Sat Jun 11, 2016 12:46 am
by bob64
Aflac wrote:I've been having a similar problem. Except mine happen at the end of a print, it completely unloads the filament at the end of every print. It did not do this on the smoothie board before I swapped and I did not change my slicer settings
http://forum.seemecnc.com/viewtopic.php?f=111&t=10023
Re: Stopping initial retract?
Posted: Sat Jun 11, 2016 8:54 am
by Aflac
bob64 wrote:Aflac wrote:I've been having a similar problem. Except mine happen at the end of a print, it completely unloads the filament at the end of every print. It did not do this on the smoothie board before I swapped and I did not change my slicer settings
http://forum.seemecnc.com/viewtopic.php?f=111&t=10023
Thanks, that is exactly what I was about to try....then some hardline on my water cooling loop cracked, got water on muy video card and so now I have done computer repairing to do before I can get that going.
Re: Stopping initial retract?
Posted: Sat Jun 11, 2016 10:52 am
by Xenocrates
Aflac wrote:
Thanks, that is exactly what I was about to try....then some hardline on my water cooling loop cracked, got water on muy video card and so now I have done computer repairing to do before I can get that going.
Ouch, that's got to smart. Was the computer running at the time?
Re: Stopping initial retract?
Posted: Sat Jun 11, 2016 5:44 pm
by Aflac
Xenocrates wrote:Aflac wrote:
Thanks, that is exactly what I was about to try....then some hardline on my water cooling loop cracked, got water on muy video card and so now I have done computer repairing to do before I can get that going.
Ouch, that's got to smart. Was the computer running at the time?
Sure was, screen went all crazy and everything, drained my loop cleaned the video card thoroughly, put everything back together and its working just fine now, i guess thats the advantage of distilled water, non conductive.
but I have added the M83 to my end gcode and am trying a print now, hopefully it works well
Re: Stopping initial retract?
Posted: Sat Jun 11, 2016 6:12 pm
by Xenocrates
Aflac wrote:Xenocrates wrote:Aflac wrote:
Thanks, that is exactly what I was about to try....then some hardline on my water cooling loop cracked, got water on muy video card and so now I have done computer repairing to do before I can get that going.
Ouch, that's got to smart. Was the computer running at the time?
Sure was, screen went all crazy and everything, drained my loop cleaned the video card thoroughly, put everything back together and its working just fine now, i guess thats the advantage of distilled water, non conductive.
but I have added the M83 to my end gcode and am trying a print now, hopefully it works well
I'm surprised distilled water stayed non-conductive, given the dust on parts. Glad it's revived, and I hope the cleaning included a dunk in 99 iso. I've seen some of those loops come apart, and I'm always horrified to see it (Partially because I have dealt with industrial hydraulic parts, and almost every time I see an issue, I see things that if the water cooling company bothered to use parts that are actually worth what they charge, that issue wouldn't have happened.)
But that's getting rather off topic. I also hope the problem is solved.
Re: Stopping initial retract?
Posted: Sat Jun 11, 2016 7:51 pm
by Aflac
Well the m83 seems to have solved my problem so that is good. And yeah I had used hardline for my cooling loop initially, And that was a mistake, I've had a few leaks from the hardline, and never anything from the flexible, I'm all flexible now. Not as pretty but less leaky