3dmaker4u
-
Posts
102 -
Joined
-
Last visited
Content Type
Forums
Events
3D Prints
Posts posted by 3dmaker4u
-
-
If you refer to the tiny holes, here and there, they are most probably due to a bit too low flow. It is less visible on lower layer heights, but appear for 0.2 mm (and for 0.15 mm a bit). Perhaps increasing the flow just a bit will "complete" the external perimeter, just before the retraction starts. If there are other holes, I might be looking for the wrong thing or in the wrong place
To benchmark with Marvin, well, you may have a look here
and/or open a hub on 3D Hubs and you'll be tested 
And improvement, perhaps printing overhangs with 0.1 mm - more fan, less speed, whatever...
-
The holes are perhaps "top layer" style, and I can bet they appear on the upper part of Marvin. More perimeters and more top layers
Here is what I used (Slic3r), but please take into account I printed with ABS (started at 240C and dropped to 210C after first layer):
; layer_height = 0.2
; perimeters = 7
; top_solid_layers = 4
; bottom_solid_layers = 2
; fill_density = 0.2
; perimeter_speed = 10
; infill_speed = 10
; travel_speed = 100
; nozzle_diameter = 0.39
; filament_diameter = 1.70
; extrusion_multiplier = 1
; perimeters extrusion width = 0.40mm
; infill extrusion width = 0.40mm
; solid infill extrusion width = 0.40mm
; top infill extrusion width = 0.40mm
; first layer extrusion width = 0.40mm
-
Higher layer height (it will be ok even with 0.2 mm) , full fan, print slower (I did 10 mm/s
. -
This "effect" on the bottom layers should ideally not appear only if you have a perfectly leveled bed, which is not heated, but has perfect adhesion and you print with filament that do not weight anything. Pretty ideal, huh?
In practice, due to a combination of the factors above, the first few layers (1-3) will exceed a bit the designed contour of the piece you print. How much "a bit" depends on the aggressiveness of the settings (e.g. temperature of the heated bed).
Depending on the kind of piece I'm printing, I use a 0.25-0.5 mm chamfer (especially for mechanical parts I design) or opt for filing the piece after printing (particularly if I cannot alter the model), while some other parts I left them as they are
-
Holes appear on "top" surfaces due to insufficiently thickness of the top layers in respect to the fill density used. Sparse infill requires thicker top, while using a higher infill may work with thinner top. Since you do not normally require a strong figurine, increasing fill density is not necessary, but this would in turn require to increase the top thickness.
I'm not normally using Cura, but I observed that the thickness of the top is coupled with that of the bottom, which is not very useful either. It means that if you increase thickness of the top, the same will be on the bottom, but this will unnecessarily prolong the print time and waste material. In Slic3r (well, I know not so many people like it, but I got used to) top and bottom thickness is decoupled (you can specify separately the number of layers at the top and at the bottom).
Now, the actual values of the two fill parameters in Cura depend on some other factors as well, such as layer height, temperature, speed and cooling. Shell should not count too much since the slicer is adding some additional 100% infill when the curvature of the model requires. But, it will still count for the top thickness that was set.
I'm printing figurines in ABS with 0.2 mm layer height, 40 mm/s at 230C, and I'm using 30% infill and 4 top layers (i.e. 0.8 mm). When using only 3 layers and 30% infill, or 4 top layers and 20% infill, I got sometimes holes.
Printing at 0.1 mm layer height would probably require more top layers, though the total height might be lower (e.g. 0.5-0.6 mm). This will add to the total print time, but there will be no issues with holes.
More aggressive local cooling may help as it will do with the second issue mentioned in the initial post: overhangs. However, cooling might impact bed adhesion, particularly if the print time is long and/or the footprint is large. If cooling do not suffice, you need support (not so bad, since you're going to finish the object anyway). If cooling result in margins lifting from the bed, try a 2-3 mm brim around the model.
-
Very nice trick with the needle.
The only issue is about how to make sure the nozzle hole is peefectly clean. Of course, one may assume that by appropriately retracting the filament, it will get to the clear metal. Well, may be, or maybe not.
Anyway, do not use sharp stuff of tougher metal to clean!
-
Well, I think 180C is far too low. 1 mm/s as the extrusion speed, counted at the filament level, not the extruded thread, for a 3 mm filament and a 0.3 mm nozzle (just to be simpler to calculate) is equivalent to... 100 mm/s printing speed! So, skipping is no wonder.
Actually, I have to admit I forgot the UM is working with 3 mm filament ;( My bad, and apologize for any inconvenience.
For 1.75 and a "regular" 0.5 mm nozzle, 1 mm/s of extruder speed is just 12.25 mm/s printing speed. Even for a 0.3 nozzle is less than 35 mm/s.
I think the methodology should be adjusted to be independent of the filament diameter.
So, let's assume we will extrude so that the printing speed is 20 mm/s (which is actually proved to work fine for 1.75 mm filament and 0.4 nozzle).
Then, going backwards, for 3 mm filament and assuming a 0.4 mm nozzle, the extruder speed should be set to 0.36 mm/s or 21 mm/min.
I didn't check George's graphs, but 180C might still be too low. But if does not skip at this speed, it's ok.
-
Digital caliper (0.01 mm) should suffice. Also, it is advisable to measure using the narrow part (at the tip of the outside jaws). This will avoid errors due to imperceptible curls in the thread. However, the caliper should be in good mechanical condition. Look against a reasonable light source to check if the jaws close uniformly along the edges, and if there is angular play.
-
Here is a tip for measuring the nozzle diameter:
- set hotend temperature to a lower value than normal, but high enough to have good extrusion (see below)
- set extrusion speed at 1 mm/s (60 mm/min) or even lower.
- extrude a few centimeters
- measure the extruded thread
The low speed is aimed at minimizing die swell.
The low temperature is to allow for faster cooling and minimize the deformation due to the extruded thread weight. However, the thread should be smooth and glossy. If it is rough, the temperature is too low.
If the host application does not allow manual control of the extrusion speed, but allow you to send commands, then try:
G92 E0
G1 E30 F60First code will bring the extrusion length to the origin. It is not anymore relevant if there are absolute or relative lengths.
Second will simply extrude 30 mm with 60 mm/min.
A funny story is about me having discovered that I printed for almost one year believing a nozzle I have is 0.5, to find it was actually 0.4 mm
I thought the method is wrong (I found it in reprap world) and test with 0.4 and 0.5 drill bits. It was indeed 0.4 mm 
- set hotend temperature to a lower value than normal, but high enough to have good extrusion (see below)
-
Let's call it a cylinder - a tube.... - it looks like it could be close to 0.4mm.
There are great chances that what you obtain is... normal.
I assume your printer has a 0.4 mm nozzle. So, you expect to get 5 circles to get the 2 mm wall thickness.
Well, I don't know how Simplify3D deals with extrusion width, but it could do the same as Slic3r does, by adding a very small fraction of 1 mm to 0.4 mm and gets, say, 0.41 mm. With such width there is no more space enough for 5 threads.
Now, if you had 2 perimeters on the outside and inside, in-between most probably you wont get anything.
If this is the case, one solution would be to use a single perimeter, so that the slicer could try to apply the infill between the walls. Once again, don't know about your slicer, but could be specific tricks for it (as there are in Slic3r).
-
... I normally use 1.61 for my wall thicknesses. IE 4 times extruder width. I use Slic3r for slicing and if I specify 1.60 it only gives me 3 times extruder width! ...
To get rid of this behavior, set the nozzle diameter to 0.39 and the issue is all gone. Please note that Slic3r use the extrusion widths to calculate the flow rates, hence this trick will not affect the printing parameters.
When you print objects with thick walls, there will be no problem to alter the wall thickness by 0.01 mm, but if they are thin (e.g. 1.6 mm!) you may end up with... 2 thinner walls
-
Oops! I just saw an error. The speed in my previous post is NOT extrusion speed, but rather the... printing speed! Or the retraction speed for some
Then, you have to calculate the actual extrusion speed that is obtained as follows:
Notations:
Df = diameter of the filament
Dn = diameter of the nozzle
Ps = printing speed
Es = extrusion speed
Then:
Es = Ps * ( Dn / Df ) ^ 2
For example: Df = 3 mm, Dn = 0.4 mm, Ps = 40 mm/s results in Es = 40 * 0.01778 = 0.711 mm/s or 42.67 mm/min.
I apologize for any inconvenience.
-
My experience with white PLA (sourced from RepRapWorld) is it oozes more than other colors at the same temperature. So it requires lowering the temperature. As a consequence it might be that the printing speed has to be lowered as well.
However, I don't think this is the main, or at least the only, reason for the "effect" you get.
What do you mean by "the Bowden tube moving"? Is it just shaking due to reversal, or it moves back and forth together with the filament. The later shall not happen, or else the tube is not... Bowden. I mean for it to properly function, it is mandatory to be well fixed at both ends. Any play will reduce the efficiency and precision.
-
My procedure is as follows:
- Start with 220 Celsius.
- Set extrusion speed at 40 mm/s (that's what I'm using for most prints).
- Extrude 20 mm.
- Check quality of the extruded wire: it should be smooth, glossy and strong.
- If not, increase temperature in 5 Celsius steps, repeat 3 and 4 above and check here.
- If extrusion tend to be rougher, it means temperature is too high, go back to 215 Celsius, repeat step 3 and 4, downwards in 5 Celsius steps and check here.
- After a few iterations, I get a range of good extrusion temperature, where the lower limit is signaled by rough extrusion (usually well before the extruder starts skipping steps).
- Thea actual temperature depends mainly on the printing speed, but if the filament does not ooze excessively, I can easily print with one temperature from 20 to 80 mm/s.
Well, please note that I'm not using an UM, but I thing the procedure is machine independent.
- Start with 220 Celsius.
-
I think you tried to level the base after printing. Due to warping, the angle between the base and the lateral walls is greater that 90 degrees. When you brought the base flat, the walls went to outside, and perhaps a bit more bending caused the shape indicated with the green arrow above.
-
I think the issue is not having a gap of .2 mm width for a length of, say, .5 mm, but rather having to fill a triangle with the smal base of .2mm and a height of, say, 1 mm.
I can agree that one can print certain pieces on cold Kapton, but I don't think is a good practice in general, thpugh it could save some energy. For the given context, it's a warp-certain case, since the walls at the corners of the case would definitely create enough tension while cooling to warp the base.
That's just what I've learned and experienced
-
PLA does not stick to Kapton without heating the bed. The "standard" temperature is 70 Celsius, but some people print at lower temperature (down to 45 Celsius).
About the gaps in the print, I don't think will work, even with a smaller nozzle diameter. The reason is that the width of the lines that compose the model varies continuously. Even at 100% infill, any slicer could not plan a path that would fill any tiny gap. Some improvement in this respect would be to increase the flow, but this will generate other issues (e.g. blobs).
-
... Pronterface is the perfect utility for tests like this and a very easy to use GUI. It gives you much more control over your printer (although you have to heat the nozzle to 170C for the extruder to run)...
Or you may manually issue a M302 command thru the terminal, to allow for cold extrusion.
-
... i get horizontal lines on my prints...
What could be causing this?
... because pulleys are excentric ...
Well, I think you answered by yourself
-
Oops, I found something missing: a wildcard * after the SOBI prefix.
So the correct statement should be:
serial:
additionalPorts:
- /dev/SOBI*Without it, when enumerating the devices, it will look for an exact match, which will not return anything. Subsequently, when trying to open the port as named further, it could not find it and the server will not start.
Hope this will fix it.
Here is a full copy of one of my config files:
appearance:
color: blue
name: Huxy 1
controls:
- children:
- command: M106 S%(speed)s
input:
- default: 200
name: Speed (0-255)
parameter: speed
name: Start Fan
type: parametric_command
- command: M106 S0
name: Stop Fan
type: command
name: Fan Control
type: section
feature:
swallowOkAfterResend: false
printerParameters:
movementSpeed:
e: 600
x: 12000
y: 12000
z: 150
serial:
additionalPorts:
- /dev/Huxy*
autoconnect: true
baudrate: 115200
port: /dev/Huxy1
timeout: {}
server:
firstRun: true
host: 0.0.0.0
port: 5001
system:
actions:
- action: shutdown
command: sudo shutdown -h now
confirm: You are about to shutdown the system.
name: Shutdown Huxy Pi
temperature:
profiles:
- bed: 110
extruder: 220
name: ABS
- bed: 70
extruder: 200
name: PLA
webcam: {} -
Don't know what to say. I tried to download the octopi, but the speed was awful and after 4 hours I closed the lid of the laptop by mistake and guess what... I'll let it download overnight.
Might be something pre-configured that conflicts.
Is there the - (dash) in front of /dev/SOBI prefix definition in the yaml file? I just ask because I initially though is a mistake, eliminated it and... didn't work
-
I understand you already have a start up file that is called octoprint-start that came in the octopi package.
You have to comment out the line that starts the http daemon. Normally should be only one (don't know how octopi is configured by default).
The new lines have to be added after the original ones (normally at the end of the file).
Access via the browser is as you pointed out, with each printer in one tab or another browser instance if you prefer. The fact you cannot yet, means that the corresponding daemons are not started, but the one corresponding to the default configuration is working, which in turn means that the configuration is not completely correct.
Check if the names were defined correctly and they are actually attached to a physical USB port, for example enter:
ls -l /dev/Huxy1and should get something like this - of course, you should use the names you gave
lrwxrwxrwx 1 root root 7 Jan 1 1970 /dev/Huxy1 -> ttyUSB2If you get an error, there is an issue with names association, and while you can connect to the printers as /tty/USBx ports, you have no control of which printer is defined at that port.
If the name is available but you cannot access the web interface, then the daemon is not started, which means the start up configuration is not correct.
There should be no router specific configuration, assuming you just use the router as an Ethernet hub/switch on the LAN side (proved as you can connect one printer with default config).
I'll try to have a look inside the octopi pack.
-
Priming before painting has also the advantage of smoothing the surface a bit. It stands also true that it might also blur some detail. But does not consume too much time since it's easy to apply and cures very fast.
-
Oops, was not here for a while
Having more printers, we have to create and setup a configuration file (config.yaml like) for each
So, you need ONE config file for EACH printer. You name them, e.g. config1.yaml, config2.yaml, etc. Then, when you write the daemon start command line, you just mention the corresponding config file, as indicated in the main instructions of the github repository
You can also specify the configfile or the base directory (for basing off the uploads, timelapse and logs folders), e.g.:
./run --config /path/to/another/config.yaml --basedir /path/to/my/basedirSee run --help for further information.
Actually, you may want to write one such a line for each printer in the file used for automatic start up, as indicated in the wiki. Config (yaml) files are by default in ~/.octoprint, but you may place them elsewhere.
To recap:
- one config file for each printer
- one daemon started for each printer, the command line to start it having to specify the corresponding config file
As an example, here is my octoprint-start file that the RPi is executed automatically at the end of the boot sequence:
OCTOPRINT_HOME=/home/pi/OctoPrint
$OCTOPRINT_HOME/run -c /home/pi/.octoprint/huxy1.yaml --daemon start --pid /tmp/huxy1.pid
$OCTOPRINT_HOME/run -c /home/pi/.octoprint/huxy2.yaml --daemon start --pid /tmp/huxy2.pid
$OCTOPRINT_HOME/run -c /home/pi/.octoprint/huxy3.yaml --daemon start --pid /tmp/huxy3.pid
$OCTOPRINT_HOME/run -c /home/pi/.octoprint/huxy5.yaml --daemon start --pid /tmp/huxy5.pid - one config file for each printer
print is wider at the bottom.
in Improve your 3D prints
Posted
I think that any "tool chain" should always be thought to also include a "design tool". Not the one used to primarily design a model, but a functional component (of the chain) that would be able to provide the "adaptation" to the specific production technology.
Now, if adaptation is automatic, e.g. embedded into the host and/or firmware, or manual, e.g. you have to open the model in a 3D modelling program and alter it, this is another story. It would be nice to have it automatic
On the other hand there is the "finishing" component of the process, as well. For example, even injection molding requires post-processing. Once again, if this is achieved automatic or manual, is another story. So to say, you can file the bottom edges of a 3D printed part and admit that filing is a manual form of post-processing.
It might be that one day a particular production technology will be fully agnostic of the model design (or the other way around). But for the time being is not. Not even for the industry-standard injection molding (just for a quick lecture) which we always like to compare the printed parts to.