Cura seems to reset extruder value at E1568 over and over

[mtnbkrrick 0]

The E value resets many times in the gcode when it reaches E1568.  It leaves defects in the parts, like Zits every time it does it.  Any thoughts on a solution? Thanks

 

 

Ex:

 

G1 X516.79 Y466.91 E1556.83231
G1 X514.395 Y470.753 Z3.957 E1557.68409
G1 X511.581 Y474.999 Z3.958 E1558.64227
G1 X508.972 Y478.729 Z3.959 E1559.4985
G1 X508.648 Y479.175 E1559.6022
G1 X505.931 Y482.842 Z3.96 E1560.46069
G1 X502.772 Y486.875 Z3.961 E1561.42434
G1 X499.855 Y490.401 Z3.962 E1562.28515
G1 X499.499 Y490.823 E1562.389
G1 X496.484 Y494.274 Z3.963 E1563.251
G1 X496.113 Y494.682 E1563.35473
G1 X493.003 Y498.054 Z3.964 E1564.21761
G1 X489.414 Y501.735 Z3.965 E1565.18468
G1 X486.789 Y504.306 Z3.966 E1565.87584
G1 X486.387 Y504.687 E1565.98003
G1 X483.705 Y507.201 Z3.967 E1566.67151
G1 X480.973 Y509.654 Z3.968 E1567.36217
G1 X480.557 Y510.02 E1567.4664
G1 X477.772 Y512.413 E1568.1571
G92 E0
G1 X474.512 Y515.092 Z3.969 E0.79372
G1 X471.638 Y517.354 Z3.97 E1.4817
G1 X468.269 Y519.893 Z3.971 E2.27524
G1 X465.309 Y522.024 E2.96132
G1 X464.857 Y522.34 Z3.972 E3.06506
G1 X461.879 Y524.389 E3.74503
G1 X458.347 Y526.704 Z3.973 E4.53941
G1 X455.273 Y528.63 Z3.974 E5.22177
G1 X451.691 Y530.769 Z3.975 E6.00655
G1 X448.552 Y532.556 E6.68599
G1 X448.068 Y532.82 E6.7897
G1 X444.902 Y534.534 Z3.976 E7.46692
G1 X441.712 Y536.175 Z3.977 E8.14171
G1 X441.218 Y536.421 E8.24552
G1 X438.004 Y537.991 Z3.978 E8.91837
G1 X434.267 Y539.714 Z3.979 E9.69244
G1 X431.015 Y541.136 E10.36008
G1 X427.236 Y542.694 Z3.98 E11.12898
G1 X423.954 Y543.97 Z3.981 E11.79136

[tinkergnome 927]

Cura resets the e-coordinate on every layer AFAIK - and i guess what you noticed here is something similar for the "spiralize" mode. I'm pretty sure that's not the root cause of the shown problem though...

 

The recommendation in this video may be the solution:

 

 

[mtnbkrrick 0]

Thanks for checking.  I doubled value and it still shows up in the gcode.  Spiralize mode or standard, same thing...Its definitely not every layer, which is even weirder.  Its almost like it hits a volumetric limit for the E value and resets throughout the whole code. 

 

Thanks again for offering up a possible solution.  You never know what small change can help sometime.  Ill keep looking.

 

 

[mtnbkrrick 0]

Where in the programming code of Cura would be a good place to start to change when the E value resets via a G92 E0 in the Gcode.  I can reduce the number of times it restarts if i reduce the extrusion multiplier to lets say %.  Its definitely hitting some limit then restarting somewhere. I just don't know why.

[burtoogle 513]

3 hours ago, mtnbkrrick said:

Where in the programming code of Cura would be a good place to start to change when the E value resets via a G92 E0 in the Gcode.

 

This is where it decides to reset the extruder.

 

https://github.com/Ultimaker/CuraEngine/blob/cf47e6dd3115d1998b93047ba6617ace1cf8badf/src/gcodeExport.cpp#L861

[mtnbkrrick 0]

burtoogle,  Thanks.  That is exactly what Im looking for.  Am i able to modify that in a text file in my cura installation, or do I need to recompile a "custom" cura just for me?

[burtoogle 513]

That is part of the CuraEngine which is a C++ program that does the actual slicing. It would need to be recompiled.

[ahoeben 1,948]

You could try switching to relative extrusion instead of absolute extrusion (see the Special Modes category). This will effectively remove the extruder resets, but has precision issues of its own.

[mtnbkrrick 0]

Thanks everyone.  I have a couple questions then.  When I switch to 1% extrusion it takes a lot longer to get to the threshold where it needs to reset.  Can the code just multiply the that e value again by 100% at the very end of the code?  Maybe create a Virtual E or something? 
 

Other thought, when I run the parts through prusa slicer or s3d, these do not show up?  Would I be safe to assume they will have precision issues as well but are not addressing the issue?

 

Thanks again and I really appreciate your help and thoughts here. 

[mtnbkrrick 0]

Would anyone be willing to recompile the latest version of cura for me and put no limit to the extrusion distance for me?  

I was looking into what type of programs i would need to do this, and seems out of my league.  lol. I have attempted but am coming up short. 

 

Thanks!

[mtnbkrrick 0]

On 1/21/2020 at 1:48 AM, burtoogle said:

That is part of the CuraEngine which is a C++ program that does the actual slicing. It would need to be recompiled.

How long and would you be willing to help with recompiling the latest Cura version but with an unlimited e value?

 

Sorry to have to ask.  Greatly appreciated.  

[burtoogle 513]

Can you not use relative extrusion as was suggested above?

[mtnbkrrick 0]

20 minutes ago, burtoogle said:

Can you not use relative extrusion as was suggested above?

We are using a large format printer that uses industrial controls and it runs using absolute and is the preferred method for the motion controller.  I didn't know if it would be a fast thing to do or not.  No worries if not.  Thanks

 

[ahoeben 1,948]

It would also be possible to write a postprocessing script that would rewrite the E-values and remove the G92 commands. The upshot of that solution is that it will work with future updates of Cura.

[ahoeben 1,948]

3 hours ago, mtnbkrrick said:

We are using a large format printer that uses industrial controls and it runs using absolute and is the preferred method for the motion controller.

 

Have you *tried* it though? Your large format printer should still support the M83 command:

https://marlinfw.org/docs/gcode/M083.html

[mtnbkrrick 0]

47 minutes ago, ahoeben said:

It would also be possible to write a postprocessing script that would rewrite the E-values and remove the G92 commands. The upshot of that solution is that it will work with future updates of Cura.

Yes.  We are in the process of doing that now.  That makes sense for future updates though.  You are right.

Thanks!