3D printing

[Darth Geek]

Well crap. The Ender 5 at work had a layer issue overnight and so that part failed. But ironically it was a completely different type of layer failure, one that I've never seen before. It wasn't a layer shift. More like not extruding for a layer or two, like if the roll got temporarily bound up or the hotend was losing heat for a little bit. Weird.

[MartyS (Gromit)]

Well crap. The Ender 5 at work had a layer issue overnight and so that part failed. But ironically it was a completely different type of layer failure, one that I've never seen before. It wasn't a layer shift. More like not extruding for a layer or two, like if the roll got temporarily bound up or the hotend was losing heat for a little bit. Weird.

My guess would be the filament bound on the roll.

I had the filament bind on me once long ago, now I'm super careful not to ever let go of the filament end when installing or removing a roll.  If I ever do let go and it springs back onto the roll I pull a good 5 feet off and roll it back on just to make sure it didn't get crossed.  Basically the end can go under another loop and it will seem normal for a while but eventually it will form a knot and get hard to pull off the roll.  So pull a bunch off and see if it's binding, the times I've let go and then took 5 or 6 feet off it can take several feet to show up, and might catch for a moment and let go and seem fine again but still be crossed and not bind up for several more feet.

If the roll seems fine now then maybe it was a defect in a sort section, I've seen under or over extrusion for parts of a roll due to a change in diameter but never a size change so drastic that it can't form a layer.

[Darth Geek]

Well crap. The Ender 5 at work had a layer issue overnight and so that part failed. But ironically it was a completely different type of layer failure, one that I've never seen before. It wasn't a layer shift. More like not extruding for a layer or two, like if the roll got temporarily bound up or the hotend was losing heat for a little bit. Weird.

My guess would be the filament bound on the roll.

I had the filament bind on me once long ago, now I'm super careful not to ever let go of the filament end when installing or removing a roll.  If I ever do let go and it springs back onto the roll I pull a good 5 feet off and roll it back on just to make sure it didn't get crossed.  Basically the end can go under another loop and it will seem normal for a while but eventually it will form a knot and get hard to pull off the roll.  So pull a bunch off and see if it's binding, the times I've let go and then took 5 or 6 feet off it can take several feet to show up, and might catch for a moment and let go and seem fine again but still be crossed and not bind up for several more feet.

If the roll seems fine now then maybe it was a defect in a sort section, I've seen under or over extrusion for parts of a roll due to a change in diameter but never a size change so drastic that it can't form a layer.

The weird part was that after doing a few layers badly that it eventually got back to normal and continued printing okay for many hours. So if it was a jam then it must have worked it's way out. Or maybe the temperature of the hot end had a problem for a while.

[MartyS (Gromit)]

 That is a strange one.  Nozzle clogs normally don't fix themselves so hard to believe that was the problem.  Unless there were a few pellets of a different plastic in the mixer when they made the filament, a little plastic mixed in with a higher melting point would cause a temporary slowdown.  Or it was a few inches of filament with a smaller diameter.

Since you have Prusa slicer have you switched to layer view and scrolled through all the layers of that model to see if there are any missing?  I've found issues with 3D models that way so I always look at all the layers before exporting the g-code even for models that look solid.  Sometimes when people take a CAD model and export it to a mesh it leaves behind odd internal stuff that can mess up the slicer.

[Darth Geek]

That is a strange one.  Nozzle clogs normally don't fix themselves so hard to believe that was the problem.  Unless there were a few pellets of a different plastic in the mixer when they made the filament, a little plastic mixed in with a higher melting point would cause a temporary slowdown.  Or it was a few inches of filament with a smaller diameter.

Since you have Prusa slicer have you switched to layer view and scrolled through all the layers of that model to see if there are any missing?  I've found issues with 3D models that way so I always look at all the layers before exporting the g-code even for models that look solid.  Sometimes when people take a CAD model and export it to a mesh it leaves behind odd internal stuff that can mess up the slicer.

I don't think so. This error and the layer shifting errors I have seen all happen at different heights on the model. If it were a problem with the CAD or slicer profile then it should be more consistent.

[MartyS (Gromit)]

Do the issues always occur in the top, larger part of the model?  Looking at the shape of it I see all kinds of ways it could expand or contract on one side more than the other with temperature changes.

It's a shot in the dark but maybe putting a cardboard box around the printer to try and keep the temperature constant.  It's amazing how much pressure the nozzle puts on big areas like that when printing the infill, and with a small base it's just asking for layer issues.  For a top heavy model like that I would probably print it in 2 parts and glue them together.    The Prusa slicer has the option to save what's on the build plate as an stl, so for stuff I print in parts I'll move it down a specific number of mm, save the top half, reload and move it down again the same amount and save the bottom half, then just flip the bottom half over for slicing. 

Would the top of it print without infill?  If it can zero percent infill would be another option if you don't need it to be strong.  I've found that zero infill but adding a few extra perimeters still makes parts that are pretty strong, but it's not always an option depending on the shape of the top of the model.

[Darth Geek]

Do the issues always occur in the top, larger part of the model?  Looking at the shape of it I see all kinds of ways it could expand or contract on one side more than the other with temperature changes.

It's a shot in the dark but maybe putting a cardboard box around the printer to try and keep the temperature constant.  It's amazing how much pressure the nozzle puts on big areas like that when printing the infill, and with a small base it's just asking for layer issues.  For a top heavy model like that I would probably print it in 2 parts and glue them together.    The Prusa slicer has the option to save what's on the build plate as an stl, so for stuff I print in parts I'll move it down a specific number of mm, save the top half, reload and move it down again the same amount and save the bottom half, then just flip the bottom half over for slicing. 

Would the top of it print without infill?  If it can zero percent infill would be another option if you don't need it to be strong.  I've found that zero infill but adding a few extra perimeters still makes parts that are pretty strong, but it's not always an option depending on the shape of the top of the model.

It is supporting the weight of the rest of the statue and upper lamp weight, so yes it does need pretty good infill (I have it at 20%)

It is always the top of the model that it has problems. I had not thought of putting something around it for temperature consistency. I'd have to find a decent sized box, but I could give it a shot.

[Darth Geek]

I finally got the linear rails on the machine. I made sure I lubed them properly and the X and Y axis are running smooth and square. AND SOMEHOW THE RIPPLLING IS EVEN WORSE! What the hell is going on?!

[MartyS (Gromit)]

Ouch, classic case of making the machine more accurate in one way and it shows off a deficiency in another area...

That is so regular on both axis, and it doesn't decay like normal ringing does.  Normal ringing and ghosting has big close waves near the direction changes and they get lower and farther apart farther away from the corner that starts it.

Is the distance between the "waves" consistent no matter how large the part is?  Also is it the same with different filaments?

It looks like something is happening in both axies at regular intervals, so look for something common to both.  I guess both stepper motors could be having the same issue but that seems unlikely...

That wave pattern isn't the same size as the teeth on the belts is it?  If it is check to see if the inside of the belts are rubbing against something.

It could also be slight repeditive changes in how the filament is being fed.  I would take the extruder apart and check the "gear(s)" that grabs the filament, clean the teeth out real well, look for anything that might cause the filament to bind and slip as the gear turns.  Cleaning out the filament path is something that has to be done periodically, so if you have been printing a lot a cleaning might be due.

[MartyS (Gromit)]

 Might be good to print one of those cubes but cancel the print half way, then look at the top layer with a magnifier.  Might need a 10x magnifier or maybe a close up picture.  You may be able to tell if the waves are caused by the extruded plastic getting fatter and thinnier or if they are caused by side to side movement.  Fatter/thinner it's an extrusion issue, side to side is an axis issue.

[eegah]

I finally got the linear rails on the machine. I made sure I lubed them properly and the X and Y axis are running smooth and square. AND SOMEHOW THE RIPPLLING IS EVEN WORSE! What the hell is going on?!

Are those googgly-eyes on your socks?

[Darth Geek]

I finally got the linear rails on the machine. I made sure I lubed them properly and the X and Y axis are running smooth and square. AND SOMEHOW THE RIPPLLING IS EVEN WORSE! What the hell is going on?!

Are those googgly-eyes on your socks?

Ha ha, no it's the out of focus Under Armor logo.

[Darth Geek]

So I came across an article with an interesting suggestion of stronger bed springs. Looking at my machine, the springs are the stock ones that came with it, but they were also almost at their full extension. So they were applying very little load on the bed. I tightened them down and readjusted the stop sensor to accommodate. That seems to have helped. Interestingly, I also increased the wall speed to 40mm/s instead of 25 and it looks better. Maybe 25 just happened to coincide with my machine's natural harmonics? I also turned off acceleration and jerk control in Cura and used the controls in the firmware of the machine, and I lowered that from 500 mm/s2 to 300 mm/s2.

[Darth Geek]

Well I just had a very bad situation. My printer started continuously blaring a very LOUD alarm, and the screen was telling me there was a thermal runaway error and I had to reset. But turning or pressing the knob didn't do anything! Turning it off was the only thing I could do. Is that what it meant by reset, or was there something else I was supposed to do?

After waiting a little, I took the cover off the hot end and it looks like one of the thick red wires wasn't attached to anything.



I'm guessing there was a wire connected to the white disk.


EDIT: So I restarted it and resumed the print just to see what it would do. The bed heated up okay but once it started heating the hotend it gave me the same error:
Heating Failed: E1

[MartyS (Gromit)]

You'll need a new heater cartridge, $10 to $20 usually. 

The white disk is the ceramic inside a metal cylinder that contains the heating element, those wires are usually hard to break but if not cable managed so they don't move the constant flexing can brake any wire.  Someone on the heater block there will be a screw that when you loosen it the heater cartridge should slide out.