Cura 15.04.6 or Cura 2.1.2 That is the question

[prb4 6]

Well @prb4 if you want someone to fix this, please post an STL and the 2 gcode files created so someone can analyze what is different and maybe make an improvement to the latest cura.  Really the entire gcode files need be posted as that will include any scaling you did, multiple objects, and will include all settings.  You can't post them here so post them in dropbox or some other web site that you have control over (preferably not something that will only host the files for a month).

 

OK, thanks for the offer.

I can easily create an stl and 2 gcode files however I don't currently know how to supply them to you so it might take a few days for me to sort it out.

[gr5 2,173]

dropbox is free.

[diettelo 5]

Am i Drunk? i make the new Firmware Update from Cura 2.1.3, Printer is showing 15.04.7 RC5

Is this correct? Because before i´ve got 2.1.0 Firmware on it.

[jotwe 0]

Am i Drunk? i make the new Firmware Update from Cura 2.1.3, Printer is showing 15.04.7 RC5

Is this correct? Because before i´ve got 2.1.0 Firmware on it.

 

I wanted to ask the same question. I was pretty surprised to find a 15.04 version number after upgrading my UM2 to an UM2+ with Cura 2.1.3 und OS X. But it seems to be normal then. Could somebody please confirm this?!

[nallath 1,120]

 

Am i Drunk? i make the new Firmware Update from Cura 2.1.3, Printer is showing 15.04.7 RC5

Is this correct? Because before i´ve got 2.1.0 Firmware on it.

 

I wanted to ask the same question. I was pretty surprised to find a 15.04 version number after upgrading my UM2 to an UM2+ with Cura 2.1.3 und OS X. But it seems to be normal then. Could somebody please confirm this?!

 

Yeah it's correct. Right before the launch we noticed some problems with the firmware version we built for 2.1.3 (which should be the same as 15.04.07, save for the version number). As we didn't want to delay, we just used the 15.04.07 firmware for 2.1.3

Sorry for the confusion.

[jotwe 0]

Yeah it's correct. Right before the launch we noticed some problems with the firmware version we built for 2.1.3 (which should be the same as 15.04.07, save for the version number). As we didn't want to delay, we just used the 15.04.07 firmware for 2.1.3

Sorry for the confusion.

 

Ok, thanks!

[prb4 6]

Well @prb4 if you want someone to fix this, please post an STL and the 2 gcode files created so someone can analyze what is different and maybe make an improvement to the latest cura.  Really the entire gcode files need be posted as that will include any scaling you did, multiple objects, and will include all settings.  You can't post them here so post them in dropbox or some other web site that you have control over (preferably not something that will only host the files for a month).

 

OK, sorry it has taken me so long to sort out the files.

However if anyone is interested in looking into this here are some files.

There are three files:

- The original stl file of a cuboid 30x30x10 mm

- The gcode file sliced using Cura 14.09

- The gcode file sliced using Cura 15.04.6

I have the firmware from 15.04.9 installed on the Ultimaker 2+

Cura 14.09 said the print would take 25 min and it actually took 25 min 48 seconds

Cura 15.04.6 said the print would take 24 min and it actually took 32 min 17 seconds

I used the same profile for both prints. Sliced with 14.09 first and loaded the profile into 15.04.6 from the gcode created by 14.09.

The basic settings are:

0.1mm layers, 0.8mm shell, 0.8mm top and bottom, 25% fill, 50mm/s, Brim.

The files are in dropbox

stl file

https://www.dropbox.com/s/2kxl3khhnsw05mm/30x30x10%20cuboid.stl?dl=0&oref=e&n=206000559

 

14.09 gcode

https://www.dropbox.com/s/eqh62hm52mb14f5/14_09%20cube.gcode?dl=0&oref=e&n=206000559

15.04.6 gcode

https://www.dropbox.com/s/elu6nuuei1bx7jg/15_04_6%20cube.gcode?dl=0&oref=e&n=206000559

Many thanks to anyone who can shed some light onto this....

[gr5 2,173]

Oh! But what about cura 2.X? Development on Cura 15.X is dead. No one at UM is going to look into this unless there is a problem with 2.X. Sorry!

[gr5 2,173]

Wait - maybe I misunderstood. Cura is the slicer. But maybe you are talking about the firmware? I just read your original post before I posted above but I don't know if you said anything about firmware. The firmware is called "marlin". Are you printing the same gcode file on different versions of Marlin and getting different times? This seems extremely unlikely. More likely the slicer is different.

[prb4 6]

Wait - maybe I misunderstood.  Cura is the slicer.  But maybe you are talking about the firmware?  I just read your original post before I posted above but I don't know if you said anything about firmware.  The firmware is called "marlin".  Are you printing the same gcode file on different versions of Marlin and getting different times?  This seems extremely unlikely.  More likely the slicer is different.

 

I am talking about Cura, it's a slicing problem.

Firmware is identical

Exactly the same problem occurs in 2.X and 15.X

I did run some tests with 2.X and as far as I could tell the times were the same as 15.X.

i.e. 2.X takes much longer to print than it should

[neotko 1,417]

It probably it's because they are doing many new stuff that it's planned to improve the quality, give them some time they will really surprise you. Good software takes time and ultimaker it's really pushing the software forward.

Who knows maybe I might even move from s3d to cura

[nallath 1,120]

Who knows maybe I might even move from s3d to cura

 

That would be the day

[prb4 6]

It probably it's because they are doing many new stuff that it's planned to improve the quality, give them some time they will really surprise you. Good software takes time and ultimaker it's really pushing the software forward.

Who knows maybe I might even move from s3d to cura

 

I see no improvement in quality

Only a massive increase in print time (up to 100% longer).

This bug has been around for a long time.

I occasionally try a new version of Cura to see if it has fixed the problem only to be disappointed time and time again to find that the print times are extremely long.

As I said in an earlier post I think the problem is related to the way retractions are handled.

This could be a red herring however visually it seems that on versions post 14.09 (ish not sure which exact version the fault was introduced) the print head pauses for a long time during a retraction compared to 14.09 which hardly pauses at all.

...and before anyone asks the retraction settings are identical!

[prb4 6]

All gone quite again eh!

I've uploaded files illustrating the problem as requested (see my earlier post).

I am surprised someone at Ultimaker isn't interested in significantly improving their print times?

[nallath 1,120]

It's not so much a matter of being interested, it's a matter of having to do a gazillion things that have higher priority and already too little time to do them.

[johank 7]

 

@prb4: It seems the 14.09 gcode does not have retractions enabled.

[nallath 1,120]

Enabling retraction doesn't save any time. I don't even think there are much retractions in this print.

Changing the top/bottom does have a dramatic impact. Cura 2.1 uses wall speed of 40, outer /inner wall of 30 and top/bottom speed of 20.

Setting those to 50 will improve print speed from 46 min to 24 min.

[prb4 6]

@prb4: It seems the 14.09 gcode does not have retractions enabled.

 

That's interesting but are you sure?

I ask because the printer definitely looks and sounds like its doing retractions.

If it isn't retracting I wonder what it is doing?

[prb4 6]

Enabling retraction doesn't save any time. I don't even think there are much retractions in this print.

Changing the top/bottom does have a dramatic impact. Cura 2.1 uses wall speed of 40, outer /inner wall of 30 and top/bottom speed of 20.

Setting those to 50 will improve print speed from 46 min to 24 min.

 

OK that's interesting. I'll definitely look at that and see if that is where the delay is.

I can understand that reducing the speed for the walls and the top/bottom should improve the print quality. However isn't there a problem with nozzle temperature?

Normally if I were to print at 20 mm/s instead of 50 mm/s then I would reduce the nozzle temperature.

I am interested in this because I can see the potential to increase the speed to 100 mm/s or even more and keep the wall and top/bottom speeds at say 50mm/s and not loose any quality but reduce the time. However I am concerned about what the nozzle temperature should be with such an approach?

Is there perhaps a recommended limit on the difference between the speeds set?

[nallath 1,120]

We are working on setting temperature based on flow (higher flow -> higher temp). But it's a complicated problem. The "slowness" of the heater is not helping and the PID controller doesn't like it if it's target temp is changed all the time.

[geert_2 557]

@prb4: It seems the 14.09 gcode does not have retractions enabled.

 

I use Cura 14.09 on two computers (haven't found time to try later versions...), and on both retractions were enabled by default. I heard and saw our printers doing retractions even before I knew what retractions were, when I began 3D-printing.

[neotko 1,417]

We are working on setting temperature based on flow (higher flow -> higher temp). But it's a complicated problem. The "slowness" of the heater is not helping  and the PID controller doesn't like it if it's target temp is changed all the time.

 

On that experiments did you notice problems on the color finish? I get really noticeable changes on the actual color uniformity of the print from 5-9C and also the print changes from matte to gloss at some temperatures.

[johank 7]

That's interesting but are you sure?

I ask because the printer definitely looks and sounds like its doing retractions.

If it isn't retracting I wonder what it is doing?

 

@prb4: Looking through the gcode you posted named 14.09 I can only find a single G10 (retrection) command in the code at the very end. Also the extrusion values are as far as I can tell increasing, never decreasing (only checked for non-strictly increasing in the mathematical sense).