Unwanted retraction UM2

[MJJ56 0]

Hi,

When I print a model sometimes there is a unwanted retraction (even when retraction is NOT enabled). There is no reason for the retraction, printing is going very well at that point. My question is: what triggers the retraction. Second: where can I find the code in the Marlin firmware that perform the retraction?

I changed the PTFE coupler some weeks ago and I must say the above described perfomance is not noticed anymore. But I wat to know if it happened during a print, because at that point the print is obviously weak.

Thanks for an answer,

Jaap.

Edited May 7, 2017 by Guest

[DaHai8 74]

There's not really a piece of code in the Marlin Firmware that does retractions. You won't find a 'retract filament' method or subroutine.

Retractions are done in the g-code using the move commands: G0 (rapid) & G1 (normal)

specifying a negative value (when using relative moves), or a lower value than the previous one (for absolute moves), for the Ennn Parameter is what does a filament retraction.

The g-codes are created by the slicer (Cura) and sent to the printer. All the firmware does is read the movement command and then rotate the motors the correct direction and speed.

Hope this helps.

[MJJ56 0]

DaHai8,

I totally disagree your with your answer. There is NO retraction in the the g-code. Even when the first layer is printed there is a sometimes a retraction. Sometimes printing the brim. See https://ultimaker.com/en/community/16596-retraction-issue

there is a video example I found today.

If I lower the maximum E-current from 1300mA to a lower value (1000mA) then this retract is far more often. So my conclusion is that the Marlin firmware does this.

So your hope that this helps is: NO.

My hope is now that you agree with me and/or someone else can help. Anyway thanks for your time and answer though.

Regards Jaap.

Edited May 8, 2017 by Guest

[DaHai8 74]

DaHai8,

I totally disagree your with your answer. There is NO retraction in the the g-code. Even when the first layer is printed there is a sometimes a retraction. Sometimes printing the brim. See https://ultimaker.com/en/community/16596-retraction-issue

there is a video example I found today.

If I lower the maximum E-current from 1300mA to a lower value (1000mA) then this retract is far more often. So my conclusion is that the Marlin firmware does this.

So your hope that this helps is: NO.

My hope is now that you agree with me and/or someone else can help. Anyway thanks for your time and answer though.

Regards Jaap.

 

With the little information you provided, my answer is factually correct. You have something else wrong with your printer then - like your stepper drivers are overheating as now you provide additional information on power settings. But again, retractions done within Marlin are controlled by g-code ONLY. There is no 'retraction' routine.

Sorry I tried to help...good luck

P.S.

Edited May 9, 2017 by Guest

[zerspaner_gerd 382]

Sounds like skipping!

https://ultimaker.com/search?search=skipping

[neotko 1,417]

Indeed steppers overheat can increase the amount moved, making electronics miss steps (as in missed forever and never done).

Skipping does actually a noise of over stress on the feeder that is usually quite easy to spot as 'what tha hell was that weird noise!'

Also. Steppers under torque and without torque, move more or less. Specially on old stepper chips like the ones um2 has. So there's a direct effect of moving under torque and without torque that affects positioning and makes less extrusion than retraction.

To learn more about that very interesting fact:

http://hackaday.com/2016/08/29/how-accurate-is-microstepping-really/

That's why (unrelated to this topic) I'm starting to use negative retraction distance restart. Making the blobs almost invisible (with proper calibration and tests)

[zerspaner_gerd 382]

Indeed steppers overheat can increase the amount moved, making electronics miss steps (as in missed forever and never done).

Should not it be better if you lower the current strength, and not worse!

 

If I lower the maximum E-current from 1300mA to a lower value (1000mA) then this retract is far more often. So my conclusion is that the Marlin firmware does this.

[neotko 1,417]

 

Indeed steppers overheat can increase the amount moved, making electronics miss steps (as in missed forever and never done).

Should not it be better if you lower the current strength, and not worse!

 

If I lower the maximum E-current from 1300mA to a lower value (1000mA) then this retract is far more often. So my conclusion is that the Marlin firmware does this.

 

Indeed lowering current should help. I installed heatsinks a year ago in all my stepper chips um2.1.x boards and never had that problem again (did happen in summer).

Also the only thing that comes to mind is a bug on the retract/extrude Gcode. But I always use reprap mode so dunno. Also marlin and amount of amps ain't related at all like @zerspaner_gerd says.

Edited May 10, 2017 by Guest

[neotko 1,417]

Also. Post the gcode (zip no copy paste) and share it on a dropbox or something to check it.

[MJJ56 0]

The answer of DaHai8 pushed me towards the A4988. I think (but I can't find a description) that the A4988 reverses the direction if the maximum current is reached.

I tried the following. Printed a object (small cube) with the maximum current for the E of 1300mA. Result: no problems, no retractions, just fine.

Then I lowered the E current to 900mA. With the same g-code: problems, sometimes a rectraction.

So I agree with DaHai8.

 

There's not really a piece of code in the Marlin Firmware that does retractions. You won't find a 'retract filament' method or subroutine.

 

If I read the datasheet of the A4988 well overheating or maximum current result in the same behavior??

As I said before changing the PTFE coupler had a great positive impact. I even programmed a PTFE used timer in the Marlin firmware, so I know how long I'm using the coupler.

What I dislike most is the fact that I don't know that is "skipping" happened, since I don't stay with the printer all the time

So DaHai8, sorry, yours answers meant a lot!!

Edited May 10, 2017 by Guest