GregValiant
-
Posts
5,182 -
Joined
-
Last visited
-
Days Won
211
Content Type
Forums
Events
3D Prints
Posts posted by GregValiant
-
-
You've probably been looking at all the settings but is the prime tower enabled? Or a wipe or nozzle clean or a plug-in?
-
1
-
-
How about the post-processor plug-in "Time Lapse"?
-
Until someone shows up who knows what they're talking about - look under "Cooling" at Minimum Layer Time and Lift Head. You have a small cross-section there and Cura may be letting each layer cool sufficiently before continuing to the next layer.
-
It isn't a problem for all printers. The Creality machines don't like file names over 16 characters. Of course Creality doesn't explain that anywhere. Each of us pretty much has to find out by ourselves.
-
You would need to use a Brim for the model but then you could use "Enable support brim" to give the supports a wider base.
-
1
-
-
M420 S1 Z2 ;Enable ABL using saved Mesh and Fade Height
I don't have a CR10 but from reading the other threads, your printer doesn't like M420.
In the Cura forum, there is a thread about "How do I get rid of this triangle". There was a lot of talk and then a member named "civit" posted this:
"I was having this problem with a new cr10s pro v2. It was quite annoying. The fix suggested above works, however, and now my gcode from cura works perfectly.
"Change this
G28 ;Home
M420 S1 Z2 ;Enable ABL using saved Mesh and Fade Height
To this:
G28 ;Home
G1 Z5 F5000 ; lift nozzle
G29 ;Enable ABL using saved Mesh and Fade Height"That change seems to have made a lot of people happy.
There might be an updated "machine definition" file for Cura as well.
-
On 9/22/2020 at 4:22 AM, paddymcgowan said:
We are a research group looking at the energy demand for 3d printing, all methods and its carbon footprint.
100% of all the PLA produced on the planet will end up in landfills. - Me
-
Ok. Semantics is it. Well I don't like "modifier meshes". How about something exciting like "Variable Density Augmenters" or "Internal Structure Adjusters" or the ever popular "Inside Topography Transformer Thingies"?
-
1
-
-
I have had some success (depending on part geometry) using support blockers. Set "Per Model Settings | Modify settings for Overlaps" differently for the blocker. You can add walls or change the infill density to 100%.
I suppose you could bring in a 2nd custom shaped model and use it as the blocker. I haven't had to do that yet but off the top of my head it sounds like it should work.
-
1
-
-
After a while (the length of time is flexible) you need to re-format the SD card.
Large capacity SD cards are not handled well by the small processor in the printer. Stick with 8 or 16GB cards.
Long file names aren't a good idea either since the printer has to truncate them to DOS 8.3 format internally.
Finally, try saving the gcode file to the hard drive and then manually copying it over to the SD card. See if that works.
-
The teeth aren't quite connected to the pulley body. I tried the mesh fixer in Cura and it seemed to make a bit of a difference but it still wasn't forming the radius between the teeth. I think the problem is with the model as well. Instead of being part of the pulley, the teeth are just really close to the pulley.
-
Try putting an M220 S100 at the beginning of the Gcode file. If that works then put one in the Start-Up gcode in Cura so it will get added to every file. You'll still want to chase down what is really going on with the high feed rates.
-
Thanks a lot eldrick. Mari figures it almost can't be done. I spend 20 minutes coming up with a complicated incomprehensible solution that has a .0000001% chance of working, and you stick in a 2 line fix.
I'll have you know that if Xscale <> Yscale by more than about 50% that Horizontal Expansion won't work on a reindeer.
So there.😜
-
Feed Rate % is set by M220 and there isn't one in either file. The problem has to be in the printer/firmware.
In the second file the skirt is printing at a sedate 20mm/sec so at 800% it would be 160mm/sec which is really flying. Is that what it's doing? If it's moving relatively slowly then the display is wrong but the feed rate setting is correct at 100%.
-
But Mari - there is that other .0000001%...It's possible there is a window there but without knowing the exact model???
Now I'm not sure if this will actually work but it seems to have potential.
If we bring in a cookie cutter shape that has a wedge on the bottom, and scale it up say to 200% in X and Y and leave it at 100% in the Z (so it doesn't get any taller), the wedge scales up and the walls scale up. Playing with the Cura number of walls and tops and bottoms doesn't give us much of anything.
We know Cura always brings things in at their center-of-geometry so we know that point of the cookie cutter is at build plate center (0,0)
We bring in the same cookie cutter file again but this time scale it to say 195.2% X and Y and leave Z at 100% (that 195.2% number I calculated would leave walls 1.2mm thick) and then move Model 2 to 0,0 so it lines up perfectly with Model 1.
With the second (smaller) model selected we choose Per Model Settings and the second tool from the right Modify Settings for Overlaps. Then select Cutting Mesh (instead of Infill Mesh) and then Select Settings. From the selection of settings pick Wall Thickness, Top/Bottom Thickness, and Infill Density and set all three to "0".
In the main settings window set Wall Thickness to 2, Top/Bottom Thickness to 0, and infill to 100%.
Slice and dice and see how it looks. I think the key might be to getting the scale right on the smaller "Cookie Cutter Cutter" model so the wall thickness comes out acceptable.
I like oatmeal cookies with raisins and nuts so If the cookie cutter thing doesn't work out, I'm good with the old fashioned "Push the dough off the spoon" technique.
-
The sensors would need to be present and the firmware would need to be able to translate each sensor and likely be sophisticated enough to make a decision about what to do. The printer would need to be networked and have an address book of destinations to send the notification to, as well as internet access or a phone line of some sort. The destination hardware and software would need to be constantly monitoring for messages from the printer.
Once the printer decided that a particular malfunction at a sensor (let's say the "spaghetti sensor") needed to go out, it could send a text message to the users' phone where a special tone would notify the user of a problem. If this was to occur at 3am then the user's significant other would punch the user in the side of the head and say something constructive like "WHAT THE %#$@ IS WRONG WITH YOU. SHUT THAT %@%# THING OFF" At that point the user would go to the printer, look at the problem, and say to him/herself "Yep, that ain't right." and then shut it off till morning when the problem could be dealt with.
The technology is certainly available for all of that. (I myself have designed a spaghetti detector that would work.) Unfortunately, it's probable that none of it could be installed as an upgrade to an existing printer, and that on top of the additional $6k cost of a new printer would be another $2k for the privilege of getting whacked in the side of the head in the middle of the night.
-
1
-
-
After all that talk what I was saying is that any problem that Gallen99 was having is not in the "W1 underextrude.gcode" file so probably not in HIS installation of Cura. I printed 40 layers of that file and it came out very well.
Figuring out why we aren't getting good prints is part of doing this stuff. Sometimes it's software, sometimes it's firmware, sometimes it's operatorware. Getting to the root of the failure is important or you end up chasing your tail. You may be having software issues. Gallen99 may be having hardware issues. I may be dealing with a bad model. The Gloppy prints look the same so where do we start to fix things? That's why I like to chase stuff down. I learn as I stumble along.
-
On 9/13/2020 at 10:43 AM, grNadpa said:
... So I tried re-printing an object that printed properly before and, sure enough, it "glopped".
Did you re-slice the model and try to print the new Gcode file - or did you run the previous Gcode that worked?
Within the printer under Control / Filament make sure "E in mm³" is "Off".
Looking at W1_underextrude.gcode in line 4 it says it uses 20.868m of filament. Dumping the file into AutoCad I come up with 505720.78mm of extrusion and a total E of 20.89mm. So that's whats going on in the Gcode file.
With an initial layer height of .2, a layer height of .25, and a line width of .4 and doing some math (sorry grNadpa) then layer 1 is 16020.78mm extrusion and 534.5mm of filament. The rest of the file comes to 489650.5mm extrusion and 20357.30mm filament. Total filament on a piece of paper comes to 20.891meters of filament. That is a difference of .0238mm of filament which equates to about .5mm of extrusion out of 505720mm total extrusion.
I would say that in the case of the W1_underextrude file that the problem is not in the Gcode.
I did notice that in line 24 of the gcode file there is M92 E94.8. That number does represent your extruder E-steps right?
Sincerely,
Dr. Bunsen Honeydew
Muppet Labs
PS I printer about 10 layers and they looked good.
-
2
-
-
Back in the day people used to do a "burn in" on computer stuff. If the component lasted 24 hours it was good to put in a machine. A percentage did not make it but they didn't get put into machines either. So a company could warranty something with confidence that their Mean Time Between Failures was accurate.
Very few do it anymore. It's expensive and time consuming and they just don't care. Generally if a component is going to fail in the first 6 months it fails in the first few hours. If it lasts through the initial burn-in (which used to take place in the factory instead of in my house) it's good to go for a long time before it wears out.
The new business model includes
"Screw'em, we've got their money".
"90 day warranty" which is often a bald faced lie.
"24hr tech support" which is almost always staffed by people who don't know as much as you, or are just there to put salve on irate customers.
Getting into your pockets for a "Monthly Fee" rather than just selling you something at a one time price.
-
1
-
-
It's in the "Start Gcode" so Cura can't really show it in preview because Cura doesn't develop it (and I wouldn't want Cura playing with my personal settings).
After you save the file to Gcode you can open that Gcode file in Cura and it will show up. If it is interfering with the print, you could then go into the Gcode file and comment out (;) the G1 lines in the startup section and get rid of it for just that print.
-
It was the brother's last name (Valiant & Valiant). but it's also my middle name. Mom always loved the comic strip character Prince Valiant.
-
You're going to have to figure out those annoying noises. Where they're coming from and why.
Time to get the wrenches out. My advice is to take LOTS of pictures and use ziplock bags to keep the loose parts in.
And keep your fingers crossed in the faint hope that the noises are the in-expensive kind.
-
That does look good. I tried PETG over PLA and it didn't bond very well at all. It was an experiment so no loss.
Since you showed me yours I'll show you mine...(I went a different way. I'm a big hit at Walmart though).
-
The string lines are the shortest distance between the points and the arrowhead so Cura may have decided that it's the most efficient path.
Have you tried Combing mode? It makes for a lot of movement on a part like that since it would want to stay over the walls at all times. Moving the Z seam and the Layer Start points can have an effect too. That's a tough part though with those inside corners. Have you tried rotating the part on the bed to put one of the short points right at 6 o'clock? Then put the Z seam in "front".
-
1
-
Cura problems
in UltiMaker Cura
Posted
I scaled them to 10% so they would fit on the Ender build plate (they came in huge and didn't "fit within the build volume"). The BASE file sliced OK. The Clip file is "Not Watertight" but it did slice.