Almost always missing layers / underextruding

[nallath 1,120]

The current one is glass filled nylon. We will swich to glass filled POM in the near future. Bas van Deursen (also active on the forums) is currently on the feeder improvement project. Some improvements have already been made to prevent the little pieces problem that you describe.

 

[Nicolinux 288]

Hi Nico!

Van is my middle name, so, try again

 

Ok Steve, I'll call you Mr. X then

 

Well, we've seen that a well adjusted UM2 with a new nozzle and teflon part should be able to extrude PLA at 10 mm/s without problems. That said, in older printers differences between PLA's are highlighted. Some PLA's still extrude well, while others don't. I hope to be able to publish a report here soon where you can see the differences that I'm talking about.

 

Oh, can't wait. This will surely bring some light in this underextrusion matter.

 

My main task at Ultimaker is materials research. This encompasses improving printing and understanding of new materials, as well as developing new materials!

So if you have any serious questions about materials, drop me a message!

 

Very interesting. I think right now bigger leaps in 3D printing will be made with new materials than with other printing technologies or improvements to existing ones. Because as interesting as say metal printing would be, the time it takes for this new technique to stabilize might be quite long. In the meantime and with the existing tech, newer interesting materials will produce far more interesting results.

I think people here on the forums would be able to give great insights in demand for new material. Also there is a lot of experience. You could start a discussion thread about it.

 

[jbeale 0]

Not sure if this is the right place to post this, but anyway: my 10-day old UM2 works OK on very small pieces, say 1 cm across. Right now I decided to try a larger print, this one https://www.youmagine.com/designs/fanduct-another-idea

and I notice what should be a solid surface is not, it's thin crisscrossing lines. I can make it solid by going in the "tune" menu while it is printing and adjusting the flow rate up to 230% where it started at 100%. That seems like a very large adjustment, when I didn't need to do any adjustment on the smaller pieces. Is this normal and should the adjustment be so large?

I am using Ultimaker blue PLA with 0.1mm layer height, 25% fill, 25 mm/s print speed, Head 205 C, Bed 60 C, fans 100%. Using Cura 14.07 to generate the gcode.

 

[nallath 1,120]

Ultimaker blue is probably the worst fillament that we sell. The only reason we sell it because it somewhat resembles our company color.

 

[skint 61]

Then you should sell it to us for £5 a roll

 

[jbeale 0]

Ultimaker blue is probably the worst fillament that we sell. The only reason we sell it because it somewhat resembles our company color.

 

Seriously? so it's just bad filament? The odd thing was, though that later during the model print it gradually extruded more and I eventually dialled back to 105% extrusion. But the bottom few mm of the print definitely needed more than 200%.

So what is the recommended best PLA filament that I should be using, if the part that came included with the printer is the worst?

 

[Nicolinux 288]

I don't think that the blue filament can be that bad so you'd have to double the flow rate...

Maybe you could post a picture of the failed print (at 100% flow rate).

 

[cyan 0]

I managed to print the cylinder up to about 14mm3/sec (test cylinder at 200% speed) without problems using ultimaker blue and without touching flow rate (which isn't valid for that test). 230% flow rate seems crazy. You sure you're not feeding it 1.75mm filament? (A joke, but the math almost works out)

[nallath 1,120]

The blue fillament isn't that bad, but it pretty much is the worst we have. I've had the best results with the grey/sliverish fillament.

 

[jbeale 0]

Here is another part I printed with the same problem, although not quite as bad. On this one I had it around 105% extrusion and on the front face of the part (which was against the build plate) you can see in some places there are just separate threads laid down, instead of a solid surface as intended.

There are a few white patches, where a little excess glue stick on the plate stuck to the part.

https://www.youmagine.com/designs/pololu-37d-motor-bracket

 

[gr5 2,172]

Sounds like a first layer only issue. That's usually related to leveling. There are a bunch of bugs related to leveling such that if you don't have the absolute latest firmware (from July 2014) then you should level without using the leveling procedure and instead just turn the screws.

If the bed is .1mm low and your first layer is .1mm thick then you will need about 200% flow to do the first layer.

Also 205C is a bit cold for this PLA. I could see how this might cause some underextrusion. I print with this color at 220C often and it works fine for me.

I recommend you do first layer at .3mm (which is the default in cura but maybe you changed it?) as this is more tolerant to bed leveling errors.

 

[illuminarti 18]

Also, there shouldn't be enough glue on the bed to stick to the print like that. Apply it thinly, and spread it evenly with a wet paper towel, so you cover the print area in glue solution. As the bed heats, it will dry to a thin film.

 

[m3dical 0]

 

If you get under extrusion it means your resistance for feeding through the filament is greater than de push tha steppektor can deliver. The resistance is caused by your extrusion capacity (speed and temperature setting) and the friction of the bowden tube configuration (curvature of the bowden versus cirvature of the filament).

Is you want to fix underextrusion you can decrease the resistance of the filament by straightning it or increase extrusion capacity by increasing the temperature setting or lowering the printing speed.

It is all about balancing the energy you need for extrusion and the energy delivered by the heaterblock and the filament steppermotor. Get the balance wrong and you get underextrusion. Or strings if you go too far the other way.

 

[jbeale 0]

Sure enough, holding the filament in my fingers between the reel and extruder I could feel it was slipping (not advancing) in one specific area on each rotation of the extruder stepper motor, causing a regular pattern of under-extrusion on larger fill areas. I increased my head temp from 205 to 210 C, and also slowed down the mm/sec print speed. After that, I did not notice any slipping, and the extrusion appeared nearly perfect with a setting of 105% flow rate. Lesson learned.

 

[donmilne 2]

Thanks to Illuminarti for his extrusion rate test print - very intuitive and definitive. I wanted to try it out after installing IRobertl's feeder, and the results I got with it are quite counter-intuitive to me. This was ColorFabb 2.85mm Leaf Green, almost new reel.

I've always understood that the slower the speed, the less stressed the extruder will be. Notice however that my results are quite crappy in the 3 to 4mm3/s band, then it gets better at higher speeds. Optimum seems to be around 7mm3/s.

Below is the third of three test prints (sorry about picture quality) - all three showed a similar defect at low speed. The first print also showed minor underextrusion at 9mm3 and 10mm3, but that was much less severe on the two later prints (you have to hold it up to the light to make them very obvious).

Can someone suggest why the printer will have problems at low speed, but not higher?

p.s. I had a theory that the nozzle was too close to the bed and was being blocked for the early layers. So just before the third print I levelled the bed very carefully. I think it was well nigh perfect: e.g. the skirt was pefectly defined.

 

[Nicolinux 288]

You are right, this makes no sense... Maybe look at the filament spool while printing. Maybe it hangs somewhere and snaps out of it when a certain threshold has been reached. Then it is fine until the next rotation.

I know, it is a bit far fetched, but that's the first thing that came to my mind.

Another idea - maybe the temperature sensor or heater cartridge are a bit flacky and the temperature takes a bit to stabilize. Go into the tune menu while printing and watch if/how the value changes.

 

[illuminarti 18]

Perhaps this is the result of a temp swing as the fan kicks on?

[donmilne 2]

Perhaps this is the result of a temp swing as the fan kicks on?

 

I'm using the gcode you posted (I'm not sure if anything else was ever available), so presumably my fan kicks in at the same time as yours?

@nicolinux: I do occasionally hear a loud clunk which I associate with the reel binding and then jumping. Not too often though, and I haven't yet seen what it could possibly be binding on. The winding of the filament on the spool looks nice and uniform, no catches there. Still, I'll mark this mentally as a possibility. Strange that it's been so repeatable though.

 

[illuminarti 18]

Yes, but the height of the nozzle, angle of the ducting, exact response curve of the temp sensor etc etc could all be contributing to your machine responding differently when the fan starts running. Monitor the temperature, and see if it dips around that height on the print.

 

[donmilne 2]

I did check the head temperature after Nicolinux mentioned it, but I saw nothing wrong at that moment in the unrelated print I was doing. Temp sensor was showing within +/-1 of the setpoint (210C). When I have time I'll repeat the test and see if there's anything odd going on in the early layers - though nothing seems to happen until around z=5mm. I would have thought that was well past the fan turn-on event etc.

 

[gr5 2,172]

around z=5mm. I would have thought that was well past the fan turn-on event etc.

 

But this is a VERY fast print so even if the fan comes on around z=3 to 5mm the maximum cooling might be a 10 layers later.

 

[donmilne 2]

I'm beginning to think that perhaps I didn't construct IRobertL's feeder correctly. In particular, the idler bearing rotates around a pin (axle). The description doesn't mention this pin so I've assumed it's an M3 bolt again, except there is no nut on the other side of the yoke, and no room for one either. What am I missing?

 

[donmilne 2]

Well good heavens. Apparantly, having a wobbly idler bearing is counterproductive! I don't have a cure, I just reset the pin, and backed off the pressure a bit. I also edited the extrusion test gcode and set the feeder current limit to 1500mA. Below is a picture of the result. You can't see the fine detail, but I can tell you that even holding it up to the light, there is absolutely no underextrusion - it's uniform from bottom to top.

 

[donmilne 2]

Ok, I've just gone back over the feeder notes. I still don't see a discussion of how that idler bear hub pin is supposed to be held in place. Surely it isn't just supposed to screw into the plastic of the yoke - or float loose? Help!

 

[cyan 0]

Ok, I've just gone back over the feeder notes. I still don't see a discussion of how that idler bear hub pin is supposed to be held in place. Surely it isn't just supposed to screw into the plastic of the yoke - or float loose? Help!

 

If I remember correctly, one of the holes is slightly smaller than the other. When I printed it (v2 I think), an M3 screw fit perfectly, so that it could just slide through the one and bite into the other. The smaller hole is meant to be toward the body of the machine, so the screw can just be pushed through from the front, through the bushes that hold the bearing, and then be screwed into the smaller diameter hole on the back. I had to cut the screw down to length (I could only find longer ones at my local hardware store). Mine was completely solid when assembled, not loose or floating at all.

I suspect that generally only people using printers that are giving some sort of trouble bother with printing an alternative feeder, so print quality issues might be to blame for some of the problems people have had (broken latches, etc). I'm still using the first |Robert|'s feeder I printed, and I honestly cannot imagine it physically breaking. Perhaps we need a bootstrap feeder design that is optimized to be as easy to print and as tolerant of underextrusion as possible, sacrificing beauty, openness, quick filament change, etc. Then when the machine is printing better parts, the final feeder can be printed. I can't immediately think of how that could be achieved with those trade-offs, but there are many designers on this thread smarter than me