Cura 15.04 retraction when using "Print one at a time"

[shurik 87]

The issue might have been arisen previously, but this forum makes the communication difficult, to put it gently.

So here is my trouble.

WIth Cura 15.04 on Windows 7, when there are several parts to be printed using the "Print one at a time" setting, the first one prints OK. Then, the skirt of the second part starts, but the plastic barely comes out at the beginning of it, catching up usually on the 1st layer of the second part.

As the number of the parts goes up, the problem gets more and more severe. Usually, no more than 2-3 parts can be put on the glass to get printed successfully.

It looks like Cura doesn't calculate the retraction for each new part correctly and not extruding it enough at the beginning of each new part.

Is there a way to deal with it?

[shurik 87]

Anyone?

[shurik 87]

Overwhelming. It's so sad to see what this forum became...

[Dim3nsioneer 557]

Hey @shurik... I swear I see this thread for the very first time! I see, you feel like that again (sorry, I couldn't resist... ):

It happens I just had to play around with the one-at-a-time feature... to me the gcode looks fine. There is a G92 E0 (zeroing the extrusion), then a G10 (retraction), then only G0s without extrusion, then a G11 (priming) and then reasonable numbers for the E-coordinate.

Maybe something bad happens in Marlin. If (for whatever reason) the G10 is executed before the G92 E0, one would exactly get the issue you seem to have.

To be honest, I rarely print more than one object in one print job and if I do, I print it all at once to have reasonable layer times. That doesn't mean you're doing something bad but I just do not have experience with it...

Edited August 28, 2015 by Guest

[shurik 87]

Yeah, was trying to insert exactly this animated gif some time ago but the forum engine didn't support the animations.

I never play with the gcode (and do not even understand it), but could these settings play a part in the problem:

?

[Dim3nsioneer 557]

No, they should not. This is the style of the direct print window. I guess you print from the SD card on your UM2 and don't use it via USB with RepRap gcode flavor, do you?

Let's tag @gr5 as he knows much more of the Marlin insides than I do. He might be able to give us some information about if such an effect as I described above could be possible.

[shurik 87]

I do print via the SD card only.

@Daid - Maybe, you can help here?

The forum is a mere shadow of former self... :(

[nallath 1,120]

Daid is not very active these days, partially due to personal issues. I'm guessing this is not so much a g-code issue. If it is a g-code issue, disabling retraction should fix it,but if it's fixed by disabling retraction it can still be a hardware issue. Retraction squashes the material which influences the extrusion as the feeder cant give the same amount of pressure.

[shurik 87]

Thanks @nallath for the reply.

It looks mostly like a bug in the code generation algorithm. With each subsequent part, the material is being pushed forward for less and less at the beginning of it, as compared to the real need. Or, it might be retracted a bit too much at the end of the previous part.

In any case, it should be somewhere in Cura logic and it doesn't seem to be too complicated to fix.

What is really strange that the issue was not pointed out previously by anyone. It is persistent and can be easily reproduced. Any way to resolve this please?

[nallath 1,120]

Could you check if the same thing happens with 15.09.81 Cura? We've stopped development on 15.04, so it could be that the issue is already fixed in the slicing engine (I don't work on that, so I don't have any idea what changed / was fixed in the past months)

[eldrick 65]

Go to Expert settings, and set the Skirt Line Count to a larger number.

[gr5 2,173]

Sorry I have tag notifications turned off.  The new forum is half the volume of the old forum but that is still incredibly busy.  In fact most questions that get unanswered could probably be blamed on too much activity.

I stopped using print one at a time because it would often make a mess at the start of the second (or third...) print during the priming phase.  I only ever tried it on UMO - never UM2.

I had to print 25 bracelets at a time so I just wrote my own custom gcode.  The part that does retraction is buggy - G10/G11 gets confused; some operation resets it - not sure if homing does this or maybe the G92 E0 command messes it up... but something messes up G10/G11.

It would be nice if you could experiment with the gcode and figure out what exactly the bug is in marlin so that we can have Cura work around the bug.

some of these bugs may have been fixed in more recent *firmware* so you might want to try loading the latest firmware for your machine.

Edited September 17, 2015 by Guest

[shurik 87]

Hi @gr5 and the the rest of the gang, thanks for the info!

I will try some tests once I'll have some time for this. I do not know the GCode and have not much time or desire for it at the moment.

Are you sure the problem is in the firmware and not in Cura?

What is the latest stable firmware that can be used? The newer Cura versions are of the same quality as the forum, which is really annoying but that is another story.

[gr5 2,173]

I'm only certain that on my machine which has very old firmware there is some serious bugs in G10 and G11 that come up when you start a new print. By the time you do the second set of retract/unretract it works fine but the damage is often already done. Also I'm certain that those bugs were addressed (but I never verified that they were truly fixed). There also is/was a bug regarding how far it retracts at the end of the print versus how much it primes at the start. This I believe was also fixed over a year ago.

[gr5 2,173]

The firmware hasn't been changed much in the last 6 months so anything 15.04.* or newer (including the beta which should have the same firmware) is all good. Although I recommend tinkergnome's version of the firmware.