Help with 0.2mm nozzle print quality  

Re: Help with 0.2mm nozzle print quality

[...] I pull it out the side when this happens, and cut it off. Too make it exit the top hole, becaus the blob on the end wont fit the hole.Not a perfect solution, but there are printable covers on Thingiverse that unscrew or swivel to allow removing the blobs without cutting.

What causes this blob?The short answer is heat. The longer answer gets into the causes of heat and other contributing factors.

I have experienced similar problems, but have always been able to mitigate them using the volumetric speed settings and/or backing off on retractions for detail parts. My ambient printer temp is 76F/24.4C. PJR is a lot more experienced than I am, so I'll defer to his suggestions on those settings, but turning back max volumetric speeds below 11.5mm^3/s has consistently worked well for me. YMMV of course, and printing conditions make a big difference.

If you're not aware, Prusa has released STL files for a modified extruder design expressly to help with these temp issues that you can try. I'm personally holding off because of concerns about increased noise (I need mine quiet in my office) and questions about airflow, and because I'm able to work around the problems.

Re: Help with 0.2mm nozzle print quality

The 0.25mm plated copper E3D nozzle has worked very well. If I'm understanding correctly, the plating helps with sticky materials like PETG as well, so that's an added bonus. Truth be told, I only went plated because they were out of plain brass.

I also ran plated for a while. The "Placebo/Cost" effect caused me to believe they were better. But I still changed them just as often. Went back to plain brass and, honestly, I don't notice any difference.

Peter

Re: Help with 0.2mm nozzle print quality

With regards to speeds and layer heights.

The original Mk1 and Mk2 had brass filament drive pulleys with about 161 micro E-steps / mm. Slic3r came with defaults that included 50 micron layer heights.

The the MMU came out and was supplied with larger Bondtech drive pulleys with 140 micro E-steps / mm. Slic3r minimum layer height increased to 70 microns.

Now we have the Mk3, with the same Bondtech drive pulley and 1/32 stepping to give us 280 micro E-steps / mm, so theoretically a 35 micron (or maybe even 25 micron? is possible)

In order to get the best printing experience you really do need to be able to do the math and (as previously explained) work out what speeds work for your print. However, there is just so much to take into account with all this. And that includes the imponderables such as "how quickly does gravity pull the Z axis down after a Z-Lift, and taking into account the backlash in the lead screw?"

Always better to be well within the Min and Max boundaries, hence why I usually use 10mm^3/s as the maximum extrude volume and 40 micro-step/sec minimum!

Peter

Re: Help with 0.2mm nozzle print quality

The short answer is heat. The longer answer gets into the causes of heat and other contributing factors.

I have experienced similar problems, but have always been able to mitigate them using the volumetric speed settings and/or backing off on retractions for detail parts.

"backing off on retractions for detail parts", please tell more about this, I get a feeling the retracts are some problem of mine, becaus I get the feeling that the more retracts there are the faster this blob will come. And yes as I understand the volumetric speed will set a top speed for the X/Y feeds based on how many cubic mm of plastic you say the printer can melt, correct? Because I've been told on Reddit not too change it from 11.5, but that doesn't make any sens to me.

If you're not aware, Prusa has released STL files for a modified extruder design expressly to help with these temp issues that you can try. I'm personally holding off because of concerns about increased noise (I need mine quiet in my office) and questions about airflow, and because I'm able to work around the problems.

I will try too find those parts!
EDIT: https://www.prusa3d.com/prusa-i3-printable-parts/ 😛

Re: Help with 0.2mm nozzle print quality

[...] Ah yes this sounds like what I was thinking, some kind of heat creep past the heat-break?
Would it help adding some other or one more fan to the heatsink maybe? I got my printer in the garage so a little more fan noise is no problem.

There are at least a couple of potential problems that can result in extruder clicks, jams and extruder motor heat:

[...] "backing off on retractions for detail parts", please tell more about this, I get a feeling the retracts are some problem of mine, becaus I get the feeling that the more retracts there are the faster this blob will come. Exactly right. Retractions require the extruder motor to pull the filament up and push it back rapidly. It also pulls molten filament up towards (though not into at 0.8mm) the cold end which is where the plugs happen if things get too warm. The more retractions occur in a short time, the more motor movement, resulting in yet more heat to an already precariously warm extruder.

There's a bit of an ugly cycle I think:

The key is to break this cycle. Try increasing Printer Settings->Extruder 1->Retraction->Minimum travel after retraction so it doesn't do retraction for short moves. Slic3r defaults to 2. Prusa changes this to 1. Try a higher setting, maybe 5. It may string more, but if it completes without jamming, you can work on the stringing.

I've been reading a lot on the KISS slicer forums, and one of the prominent advocates there (sublimelayers.com) recommends reducing temps for PLA in almost all cases. He suggests manufacturers promote artificially high temps to make sure their stuff "works" in most printers, but that such high temps will contribute to stringing (see the cycle above). Reducing the feed rate (volumetric speed) and keeping temps lower might help break the cycle.

And yes as I understand the volumetric speed will set a top speed for the X/Y feeds based on how many cubic mm of plastic you say the printer can melt, correct? Because I've been told on Reddit not too change it from 11.5, but that doesn't make any sens to me.Well, 11.5mm^3/s is how much the E3D V6 extruder can handle at max. That should be your upper end.

You can get into issues printing too slowly if molten filament accumulates, again contributing to the overall heat creep issue. However, the extruder is only going to process up to that 11.5 limit. Some materials simply can't be pushed through that quickly. I had massive extruder skips, bucks and jams in winter which I finally fixed by adjusting max volumetric speeds. If you look at the Slic3rPE presets for PLA, max volumetric speed is set to 15 ... too fast for the E3D V6. Notice how many people in the heat discussions can print materials besides PLA just fine? Slic3rPE presets for PETG, NGEN and XT it's 8. For Colorfabb XT-CF20 it's 1. I was unable to print TPU until I throttled it to 6. [I think PJR may disagree on some of this, but I've definitely seen improvements by adjusting this setting.]

I think it's a balancing act. You don't want to print too slowly lest molten filament contribute to problems. Yet you still have to respect the extruder limits. I generally recommend trying to slow down as a troubleshooting step, and not as a fix. Slowing down using the front knob is an easy way to test reducing retraction rate (artificially admittedly) and max extrusion without having to re-slice. If you can print slower, you know it's a slicer setting issue. If not, it may be a hardware issue.

[...] I will try too find those parts!
EDIT: https://www.prusa3d.com/prusa-i3-printable-parts/ 😛
There are some lengthy discussions going on here on the Prusa forums on those parts. They do seem to help with the cooling issue. I'm personally holding off as I'm reading a lot of comments on increased noise (very important to me) and concerns about fan nozzle efficiency. I'll likely upgrade at R4 in the future, but for now, I'm able to mitigate these problems with slicer settings. The solutions are going to vary by every printer and every owner's specific circumstances. Unfortunately, some of the threads are turning into ripping out extruder motors and other heavy-handed fixes. I'm staying away from those discussions and focusing on working with the out-of-the-box Mk3 printer. I'm just trying to share what has worked for me in the hopes some of it will help others with similar problems. PJR and others here know a lot more about the intricate details than I do, so take all this into consideration and try what works for you.

Re: Help with 0.2mm nozzle print quality

Retractions. Why? Simple explanation. Retractions are necessary to reduce the pressure in the nozzle to zero so that stringing does not occur during non-printing moves.

Linear Advance (or PreloadVE or any other similar method) is designed to reduce nozzle pressure to zero at the end of printing moves and conversely to rapidly increase nozzle pressure at the start of printing moves.

So if you are using Linear Advance (or PreloadVE or...), and that is set up and calibrated fcorrectly for the material you are using, you don't need to use retractions as another method of destringing. Well not quite, because there are other factors at play. What you actually need to do is to empty the nozzle bore (the bore being the short length of nozzle that is drilled at the tip and is the nozzle diameter) of filament and that should suffice.

So how much is that? Well, an E3D 0.4mm nozzle has a bore length of 0.6mm, so you need to retract about 0.075 mm^3 or 0.0314 linear mm of filament. Most users (myself included) over-retract when using these "black magic" methods of pressure correction.

But then there is another consideration and that is the nozzle is not perfectly sealed; the seal is dependant on the inside diameter of the PTFE tubing and the filament diameter.

None of this is an exact science and what we need to achieve is the lowest retraction value possible to suit our printer, filament and environment.

Moving on, the "Max Melt Volume" is not only dependant on the extruder, but also the filament in use. Although the figure of 11.5 mm^3 was provided to me by E3D a very long time ago, that figure was in relation to ABS at 230 degrees (from memory; it may not be correct, but it's here in the forums somewhere...). PLA will be pretty close to the 11.5 mm^3 at 210 degrees.

The hardness (or stiffness) of the filament also comes into play here, for example, I would set a "Max Melt Volume" for a soft TPU at maybe as low as 2.5 mm^3, simply because the complete extruder cannot work any faster than that; I don't have to go messing around with the print speeds, because they will be limited by that filament (material) value (sorry, but this is a material parameter within KISS, my slicer of choice; I have no idea about Slic3r).

Peter

Re: Help with 0.2mm nozzle print quality

[...] Linear Advance (or PreloadVE or any other similar method) is designed to reduce nozzle pressure to zero at the end of printing moves and conversely to rapidly increase nozzle pressure at the start of printing moves.Have you found that either method, or a combination of both, is sufficient to eliminate the need for retraction? I notice Slic3rPE doesn't have the coasting and some other features other slicers (e.g. Cura) seem to be expanding on. That would make sense with the Prusa-centric focus of Slic3rPE since they know LA will be present. I'm wondering how close LA tuning can get to eliminating the need for retraction, wipe, coast, z-hop and myriad of other extruder-rattling techniques. I'm puzzled with LA as I know it's a feature in Marlin, but I assume it would have to be tuned heavily for specific hardware. I've never read about any such work being done by Prusa or other manufacturers. Of is it just a calculation done by algorithm that users are expected to tweak?

[...] So how much is that? Well, an E3D 0.4mm nozzle has a bore length of 0.6mm, so you need to retract about 0.075 mm^3 or 0.0314 linear mm of filament. Most users (myself included) over-retract when using these "black magic" methods of pressure correction.I am working to reduce retraction distances (0.8mm in Slic3rPE by default) and increasing the minimum move required to trigger a retraction (1mm in Slic3rPE by default).

[...] The hardness (or stiffness) of the filament also comes into play here, for example, I would set a "Max Melt Volume" for a soft TPU at maybe as low as 2.5 mm^3, simply because the complete extruder cannot work any faster than that; I don't have to go messing around with the print speeds, because they will be limited by that filament (material) value (sorry, but this is a material parameter within KISS, my slicer of choice; I have no idea about Slic3r).The capability to apply throttles based on volumetric capacity are what got me to shift over to Slic3rPE. I'm also intrigued by KISS' approach -- so much so that I sprung for the premium license -- because of what sounds like tremendous potential (as well as an amazing slicing engine). I've got a raft of questions on KISS, but I think those are more slicer-specific than Prusa-specific, so will pester you with questions on those forums.

Re: Help with 0.2mm nozzle print quality

Have you found that either method, or a combination of both

Unfortunately, it is not possible to combine the two; my thoughts as to why are in the other place 😉 And again, unfortunately, they both have their own advantages; if it were possible to use both, I think it would present a much better situation.

I am working to reduce retraction distances (0.8mm in Slic3rPE by default)

0.8mm retraction is the value to use without LA or PVE. As for "coast", I see that as a lazy person's "passive method" to reduce nozzle pressure, whereas LA and PVE are both active methods, whereby the extrustion values required to do the job are calculated and properly applied.

so will pester you with questions on those forums.

Not a problem 🙂

Peter

RE: Help with 0.2mm nozzle print quality

Posted by: PJR

Check out my comments regarding calculating minimum speeds. 280 E-steps / mm. That's just less than 0.2 linear mm/sec or about 0.4 mm^3/sec (quick calc!) If you are using 50 micron layer height and 0.2 mm width, then 0.4 / 0.05 x 0.2 = 40 mm/s is the minimum speed I would use (in fact I would most likely double that with those settings).

I believe this reasoning is incorrect on the Mk3. The Mk3 has TMC2130 motor drivers, which internally interpolate to their full resolution of 256 μsteps. So although the printer sees 280 E-steps/mm, the motor is stepped at 2240 μsteps/mm. This means the printer can go 8x slower than what you predict.

This gets even more extreme with a Bondtech geared setup and a 0.9° pancake stepper. The microstepping is typically set at 8 or 16 in Marlin for such a setup, resulting in 450 or 830 firmware E-steps/mm, but the extruder sees the interpolated μsteps, which comes down to super-smooth 13280 μsteps/mm.

RE: Help with 0.2mm nozzle print quality

Posted by: bobstro

I'm wondering how close LA tuning can get to eliminating the need for retraction, wipe, coast, z-hop and myriad of other extruder-rattling techniques. I'm puzzled with LA as I know it's a feature in Marlin, but I assume it would have to be tuned heavily for specific hardware. I've never read about any such work being done by Prusa or other manufacturers. Of is it just a calculation done by algorithm that users are expected to tweak?

I assume that Prusa did some tuning of LA for the Mk3, because Slic3r-PE has different LA values set for each filament preset. They're hidden in Filament Settings -> Custom G-code -> Start G-code. There is a code snipped like this:

M900 K{if printer_notes=~/.*PRINTER_HAS_BOWDEN.*/}200{elsif nozzle_diameter[0]==0.6}15{else}30{endif} ; Filament gcode

which boils down to

M900 K30

setting the LA factor to 30 for this particular filament. In the end, it seems that the LA factor is dependent more on the filament than the extruder, since the gears-nozzle distance is minimized on all printers and thus similar.

And now I'll go print some LA calibration towers ...