Help with apparently too hot printing
-
Make your PA 0 and test - for a direct drive feeding a Dragon, i would expect it to be below .1 - for reference, my most recent print using a volcano hotend with a hemera is ~.01 with PLA. Per Duet's documents - .4 is where they'd expect a medium-length bowden setup would be. https://duet3d.dozuki.com/Wiki/Pressure_advance
This stands out as the primary issue - PA is something that needs to be tuned, not guessed, and is unique to the filament and sometimes even the colorIf overextruding - Overhangs would exhibit "drooping" behavior because there's more material to droop - I've seen that behavior on my own prints, and corrected it by dialing back the multiplier a few %.
Re: vibrations
Sorry to sound so abrasive - I run a bedflinger that has a 5kg bed assembly and I am now running 5k accel for production prints. My extrusion toolhead moves are 330+mm/s (.25mm height, .48width track, ~35mm3/s) If you want to remove the defects, slow your accel or stiffen your motion. I use two linear rails for the bed, 9mm belt, and a huge nema 17 to accomplish accurate and beautiful prints at these speeds+feeds. All I'm trying to say is that its not inevitable that you'll have vibrations or issues with large beds or "high speeds" - every machine has its limits somewhere. I can't print on this machine 15k accel like my corexy platform can, but its nowhere near "slow"
-
@luke-slaboratory said in Help with apparently too hot printing:
Make your PA 0 and test - for a direct drive feeding a Dragon,
done
Sorry to sound so abrasive - I run a bedflinger that has a 5kg bed assembly and I am now running 5k accel for production prints. My extrusion toolhead moves are 330+mm/s (.25mm height, .48width track, ~35mm3/s) If you want to remove the defects, slow your accel or stiffen your motion. I use two linear rails for the bed, 9mm belt, and a huge nema 17 to accomplish accurate and beautiful prints at these speeds+feeds. All I'm trying to say is that its not inevitable that you'll have vibrations or issues with large beds or "high speeds" - every machine has its limits somewhere. I can't print on this machine 15k accel like my corexy platform can, but its nowhere near "slow"
My printer is a CR10 S5 V rollers and not a very rigid frame. I have trouble relating to a bed slinger with a rigid frame.
-
The next print is done and it's not a pretty picture.
A temperature of 170C appears to be the lowest possible temperature. There appears to be a bit of a reduction in model strength and there is considerable (but very thin) spider webbing inside the model. Based on that, the print temp should be in the 180, 185C, 190 range.
The bad news, while the crap on the outside is reduced, it is still there. Also, the random seam location didn't help.The only thing I have left is the change to the PA (from 0.4 to 0). Beyond that the next thing is to add support to the edges.
Next print will be with the PA changed and support added. I will have to fix my z probe first though.
-
For further info, Cura has acceleration set to 500 and jerk is 8
-
@jens55 said in Help with apparently too hot printing:
My printer is a CR10 S5 V rollers
I cannot find that on the Creality site. Does it go by another name?
Have you ever gotten good prints out of your printer?
Can you do simple things like calibration cubes at slow speeds?
Frederick
-
@fcwilt, it's a fairly old model and probably not available any more.
Here it is from Amazon.com:The printer produces good prints if you accept artifacts related to slinging the heavy bed. Yes, I can do better if I go slower. Yes I can do calibration cubes.
Even the current print job is ok except for the huge overhang. I am not concerned with the mechanical vibration. I am in the process of building a Jubilee for jobs that need a better finish.
I have a normal 300x300 CR10 and it is set up with ABS and it prints the same model just fine but it uses the original Creality controller (the S5 uses a Duet2) -
@jens55 said in Help with apparently too hot printing:
@fcwilt, it's a fairly old model and probably not available any more.
Here it is from Amazon.com:The printer produces good prints if you accept artifacts related to slinging the heavy bed. Yes, I can do better if I go slower. Yes I can do calibration cubes.
Even the current print job is ok except for the huge overhang. I am not concerned with the mechanical vibration. I am in the process of building a Jubilee for jobs that need a better finish.
I have a normal 300x300 CR10 and it is set up with ABS and it prints the same model just fine but it uses the original Creality controller (the S5 uses a Duet2)Thanks for the info.
The difference between 300x300 and 500x500 is very significant. The Y axis drive system may not be robust enough.
Also looking at the picture in the link you provided the printer looks like it is going to suffer from vibration issues.
If you can generate good prints at slow speeds but not high, try cutting back on the acceleration and jerk settings in your config.g file to very modest levels but keep the print speed up - just as a test.
And for basic PLA use an extruder temp of 190, a bed temp of 60 and turn off PA and the like.
Frederick
-
@fcwilt said in Help with apparently too hot printing:
The difference between 300x300 and 500x500 is very significant. Y axis drive system may not be robust enough.
Oh yeah .... a significant difference! The Y axis is robust but that isn't going to help in the overall vibration issue.
Also looking at the picture in the link you provided the printer looks like it is going to suffer from vibration issues.
Yup!
If you can generate good prints at slow speeds but not high, try cutting back on the acceleration and jerk settings in your config.g file to very modest levels but keep the print speed up - just as a test.
As I mentioned, acceleration (via Cura) is 500 and jerk is 8 so pretty low values. I am at 70 mm/sec for printing speed including walls. I could bring the outside wall speed down to half ... which is what it was when I started this journey (and which didn't help)
Well I am printing with supports at the corners and with PA turned off now so tomorrow morning will be interesting ....
-
@jens55 can you record a little video showing that acceleration? Maybe it is still too high, but the numbers doesn't mean anything to me because i don't own a bedslinger.
-
@mikes, I have never recorded a video with my phone and have no clue what / who / how / where / why .....
For now I will want to see how tomorrows print looks like (with 0 PA and support in the corners) and I might run another print with the same parameters (ie no PA) and no support. -
@jens55 another thing to check is that the filament must be dry. If you leaved the spool outside a dry box for more than one week you probably have to dehydrate it. Wet filament produces more stringing and flow inconsistencies.
-
@mikes, the filament did indeed have a moisture problem which resulted in very low adhesion of the first layer to the build plate but it was dried and works fine again.
-
@jens55 said in Help with apparently too hot printing:
@fcwilt said in Help with apparently too hot printing:
The difference between 300x300 and 500x500 is very significant. Y axis drive system may not be robust enough.
Oh yeah .... a significant difference! The Y axis is robust but that isn't going to help in the overall vibration issue.
Well it's hard to tell from the picture but it looks like it is belt driven using one belt - perhaps 9mm. And it may have some sort of belt tensioning device.
Given there may be 1000mm or more of belt that represents a pretty good spring which means accurately positioning a heavy bed (at speed) is going to be somewhat of a challenge.
Without any hands-on experience I would suspect that 70 mm/sec is rather optimistic.
Frederick
-
Well it's hard to tell from the picture but it looks like it is belt driven using one belt - perhaps 9mm. And it may have some sort of belt tensioning device.
Given there may be 1000mm or more of belt that represents a pretty good spring which means accurately positioning a heavy bed (at speed) is going to be somewhat of a challenge.
Without any hands-on experience I would suspect that 70 mm/sec is rather optimistic.
If its a quality 9mm belt it should be fine. I run 9mm belt on my bed which is 5/16" mic6 with an AC heater and magsheet setup - its about 5kg all said and done and works fine over 300mm/s at 5kaccel. I bet the v-wheels aren't in the best orientation for the heavier bed load which could add, but I heavily suspect PA is responsible for the issues currently at hand.
-
My latest test print have me out of ideas and filament (from the roll I have been using so far)
I can report that pressure advance definitely was an issue !
I can report that the switch to random seam shows only a tiny 'hickup' at any of the 4 corners. I will definitely have to run PA calibration though as a PA of 0 clearly shows PA issues. I always planned on calibrating that so that's not a huge surprise.
If I print with support turned on and the overhang angle set to 50 degrees, I get support at just the four corners. This improves the outcome but with contact between the point of the corner and the support being ever so tiny and missing on occasion, I still had one corner that had some bending up causing problems.Things are vastly improved over the original model when all this started.
So, where to go from here ...... since I used up the roll I was working with, I will try a different roll from a different manufacturer (or at least a different brand name) and see what happens.
-
Glad to hear you are making progress.
I've never used PA but I'm thinking I may experiment with it.
Frederick
-
@fcwilt you are missing out my friend! Even on direct drives, I've found a small amount helps
-
@fcwilt, I started out with a Bowden setup (CR10) and PA is a noticeable (significant) improvement for that (even though I know people who use no PA). My second printer was the CR10 S5 which has a Bowden tube of probably close to 800 mm. It would be nuts to expect reasonable prints from that without PA.
I am starting to work with direct drive extruders now and PA can help for the discerning person but isn't absolutely essential. It might be essential if you are doing soft stuff like TPU but haven't tried that yet. -
Final update:
Different filament did the same thing with curling at the points of the corners.I had previously reported that support at the corner edges helped but that one corner still had problems. I changed the X/Y/Z distances between support and the model in Cura which got me a solid support and no curling.
To sum it up, the two big issues where pressure advance and support. I am sure I printed without support to at least 60 degrees of overhang before but maybe I didn't and maybe the shape of this model causes issues earlier but setting the support angle at 50 degrees and having the support close enough to the edge fixed all the issues. While the side angle was 45 degrees, the edge where two 45 degree sides meet must be just under 55 degree overhang.
While I did go down to 180C from my normal 200C nozzle temp for PLA, I do not know if that was actually necessary given that support was the big issue. It did help when there was no support.Thank you to everybody that helped me sort through this issue !!!
-
Well well well ..... I think I jinxed it when I said 'final' update.
So I kept thinking about the blower fan I use for model cooling. Yes, it works, yes I can reduce the output and it turns nicely even at 25% .... but I just couldn't leave well enough alone.
I removed the fan so I could gauge the output a bit better and there clearly was output. Put a new fan on to compare - same thing.
Still wasn't happy so hooked up a power supply, hooked up one of the fans and got the same output.Contemplated life a bit and decided that still wasn't good enough!
I dug out yet another new fan and hooked that up .... the output was probably 4 times as much air !!!!I will attempt to print one more of the fence post caps but this time with proper air flow for cooling.
Boy will I be pissed off if it prints perfectly even with no support!
I will also get myself a new batch of better quality fans (this time ball bearings) !I can't believe that there is a good chance that I spent 3 days or so printing because of a fan issue, the very first thing I checked and suspected!
All fans were marked as 24V and I use 24V but either two of the fans were 36 or 48V or defective (while behaving just fine except for low rpm)