Slic3r and J-head hotends

ttsalo Wrote:
-------------------------------------------------------
> But... bridging totally sucks. The filament droops
> down into pretty much a hemicircle for any bridge.
> I have bridging speed set at 25 mm/s, half the
> speed of everything else, and I thought I could
> fix this by decreasing the bridging feed rate
> multiplier, but using 80% had no effect. Is it
> possible to get decent bridging by just adjusting
> the feed rate and speed multipliers, I wonder? I
> don't want to change the layer height or W/T,
> since those are such a good tradeoff between
> precision and speed.

Have you adjusted the bridge flow rate in the advanced tab? If the filament is drooping I would decrease the flow until it stretches the filament.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Triffid_Hunter Wrote:
-------------------------------------------------------
> this is exactly what I'd expect to see if slic3r
> is trying to make the 3mm holes 3.4mm in diameter
> and you're pushing too much plastic

Just a quick report on my problems above:

Triffid_Hunter was right, I was pushing too much plastic. Playing with Skeinforge helped understand better what's going on with the plastic. By using the stretch plugin I finally managed to get all holes printed to size.

I still have problems getting Slic3r to do the job.

Sublime Wrote:
-------------------------------------------------------
> Have you adjusted the bridge flow rate in the
> advanced tab? If the filament is drooping I would
> decrease the flow until it stretches the filament.

I went as low as 0.1 bridge feed ratio and 10 mm/s bridging speed and no matter what, the filament was flowing out of the extruder about twice as fast as the extruder was moving when bridging, making just a pile of filament instead of a bridge. When I switched back to ABS, it worked much better even at 1.0 bridge feed ratio. Maybe bridging just does not work with polycarbonate. That would be a shame, since it's such a great material in most other respects.

ttsalo Wrote:
-------------------------------------------------------
> I went as low as 0.1 bridge feed ratio and 10 mm/s
> bridging speed and no matter what, the filament
> was flowing out of the extruder about twice as
> fast as the extruder was moving when bridging,
> making just a pile of filament instead of a
> bridge. When I switched back to ABS, it worked
> much better even at 1.0 bridge feed ratio. Maybe
> bridging just does not work with polycarbonate.
> That would be a shame, since it's such a great
> material in most other respects.

Sounds like it is really liquid when being printed. Have you tried lowering the temperature to the lowest you can while lowering the printing speed to something like 20mm/s? Then with the temperature really low and the print speed really low try increasing the bridge speed to say 60mm/s so it stretches the semi cool filament.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Sublime Wrote:
-------------------------------------------------------
> Sounds like it is really liquid when being
> printed. Have you tried lowering the temperature
> to the lowest you can while lowering the printing
> speed to something like 20mm/s? Then with the
> temperature really low and the print speed really
> low try increasing the bridge speed to say 60mm/s
> so it stretches the semi cool filament.

The 300°C I was using is really the minimum for getting a good print. I would have to lower the temperature just for the bridge... anyway, bridging was so bad that I suspected a bug in Slic3r, but since it's working for ABS more or less like before, it's probably just because of the material. I'll ask about that elsewhere.

ttsalo Wrote:
> --------------------------------------------------
> The 300°C I was using is really the minimum for
> getting a good print. I would have to lower the
> temperature just for the bridge... anyway,
> bridging was so bad that I suspected a bug in
> Slic3r, but since it's working for ABS more or
> less like before, it's probably just because of
> the material. I'll ask about that elsewhere.

What make PC are you using?
What hot-end are you using?

It's possible to print PC at 265 Deg C at upto 40mm/sec - if I remember correctly NYC designer managed 255 deg C on a MK7 printing slower.

---

[richrap.blogspot.com]

richrap Wrote:
-------------------------------------------------------
> What make PC are you using?
> What hot-end are you using?
>
> It's possible to print PC at 265 Deg C at upto
> 40mm/sec - if I remember correctly NYC designer
> managed 255 deg C on a MK7 printing slower.

I was printing Orbi-tech's 3mm PC from Arcol V4 at 50 mm/s. I did some prints at 280C, but got skipped E steps, somewhat uneven extrusion and worst off all, weak layer fusion. I didn't try slower speeds, though.

I've tried all the suggestion in the forum here and have no love for slic3r with my 0.5 J-head.

No matter what I do, I have filiment gaps....

[www.majhost.com]

Any suggestions? Please?

Otherwise I guess I'll have to try skeinforge and give up on slic3r.

-Rob A>

@RobA

Can you provide layer size, speed and temperature?

Also you may want to look at your extruders idler. If it is not tight enough it will slip during solid infill because you run out of melted plastic and the force required to extrude increases so it slips.

I have also found that increasing the temperature can help with the above problem.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Have you tried doing the infill slower? It shouldn't affect the density but it seems to with Slic3r and Marlin.

---

[www.hydraraptor.blogspot.com]

I've taken all the suggestion here and seem to be having more luck, now that I have also set the extrusion multiplier to 0.9

Not sure why a lower multiplier would give me better top surfaces, but it does. Not even sure why it needs to be dialed back by 10% since it was calibrated to the actual intake within 0.5% I'm speculating that artificially setting the nozzle diameter low also needed a corresponding dial back of the extrusion, even though I think that would have been taken care of by the maths.

Thanks for the suggestions!

-Rob A>

RobA Wrote:
-------------------------------------------------------
> Not sure why a lower multiplier would give me
> better top surfaces, but it does. Not even sure
> why it needs to be dialed back by 10% since it was
> calibrated to the actual intake within 0.5%

this happens because we can't measure the filament diameter with enough precision. remember that extruded plastic is proportional to square of diameter, so even a tiny inaccuracy in diameter can have visible effects on output volume

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

The flow rate maths in Slic3r seem broken to me. When I went to lower layers I had to change the extrusion multiplier to 0.85. When I went back to higher layer height I had to put it back to 1.

I can measure the filament diameter to 0.01 and that is plenty accurate enough with my own flow rate maths. However changing the hardness of the plastic or the spring tension requires the E steps to be recalibrated and I think large changes in filament diameter probably do as well.

---

[www.hydraraptor.blogspot.com]

Triffid_Hunter Wrote:
> we can't measure the filament
> diameter with enough precision. remember that
> extruded plastic is proportional to square of
> diameter, so even a tiny inaccuracy in diameter
> can have visible effects on output volume

Fair enough.

I've also attached a photo of my current print quality:


-Rob A>

RobA Wrote:
-------------------------------------------------------
> I've also attached a photo of my current print
> quality:

looks good to me, a touch of lash perhaps and occasional flow inconsistency

---

-----------------------------------------------
Wooden Mendel
Teacup Firmware

nophead Wrote:
-------------------------------------------------------
> The flow rate maths in Slic3r seem broken to me.
> When I went to lower layers I had to change the
> extrusion multiplier to 0.85. When I went back to
> higher layer height I had to put it back to 1.

I have found almost the exact same thing. I am using the flow rate modifier in Marlin which reduces the steps per mm on E and it is really obvious when you change layer heights.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Sublime Wrote:

> I have found almost the exact same thing. I am
> using the flow rate modifier in Marlin which
> reduces the steps per mm on E and it is really
> obvious when you change layer heights.

With slic3r 0.8.4, I tried the obvious experiment, and sliced a 40mm cube with no skirt, and 100% infill, at .1 layer height, and .25 layer height.

Both reported a filament requirement of 9536.0mm, and 65.2cm3.

65.2cm3 does seem a bit high for a 40mm cube - shouldn't that be 64cm3?

raldrich Wrote:
-------------------------------------------------------
> Sublime Wrote:
>
> > I have found almost the exact same thing. I am
> > using the flow rate modifier in Marlin which
> > reduces the steps per mm on E and it is really
> > obvious when you change layer heights.
>
> With slic3r 0.8.4, I tried the obvious experiment,
> and sliced a 40mm cube with no skirt, and 100%
> infill, at .1 layer height, and .25 layer height.
>
> Both reported a filament requirement of 9536.0mm,
> and 65.2cm3.
>
> 65.2cm3 does seem a bit high for a 40mm cube -
> shouldn't that be 64cm3?

Thats what my maths would say too.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Sublime Wrote:
-------------------------------------------------------
> nophead Wrote:
> --------------------------------------------------
> -----
> > The flow rate maths in Slic3r seem broken to
> me.
> > When I went to lower layers I had to change the
> > extrusion multiplier to 0.85. When I went back
> to
> > higher layer height I had to put it back to 1.
>
> I have found almost the exact same thing. I am
> using the flow rate modifier in Marlin which
> reduces the steps per mm on E and it is really
> obvious when you change layer heights.

I think this is a function of the way Slic3r picks what model to use for output volume based on the relationship of layer height to nozzle diameter.
There is a fixed point where it transitions from one calculation to the other, and that's likely why you need 2 different extrusion multipliers.

Seems like it switches from a model that is correct to one that isn't. I don't see why there should be two models. Its just a rectangle with semi circular ends. Sure, when it gets very thing the semi circles might be some other curve but they are such a small part the difference would be negligible.

---

[www.hydraraptor.blogspot.com]

I would tend to agree, having multiple models almost guarantees an issue since there is always going to be at least a discontinuity in the differential, if not in the actual values.

Hello, I finally had a chance to read the full thread. I've been working on some changes to the flow math lately, thanks to the feedback I received and the tests I did myself since when I got a 0.35 extruder too.

@Sublime, could you please test the new-flow branch on GitHub? I've been getting very good results with it, and I believe it should also work correctly with your J-head. To switch to that branch, just type git checkout new-flow. When you want to go back to master, type git checkout master. I recommend to turn your extrusion multiplier back to 1 if you had it lower before.
I'm collecting feedback before merging the new-flow branch into master. So far, I got very positive feedback so I hope it is the final solution.

@raldrich, the answer to your doubts about the total plastic volume exceeding the expected volume of the cube lies in those tiny segments connecting rectilinear infill lines in solid layers. Without them, the total plastic is equal to the volume of the cube as you would expect. Geometrically speaking, those segments shouldn't be there, as their space is already filled by the main infill lines, so plastic will exceed there. So we could omit them when printing solid layers. However, without them (as it was in the first Slic3r versions) there are usually little gaps left because plastic tends to follow the new extrudate, and also because the sudden lack of pressure makes the extrudates less regular. Usually, nobody prints a 100% dense cube like your experiment: solid layers are kept to a limited number. So the excess plastic is not a problem and it is absorbed by the non-solid neighbor layers.

Edited 1 time(s). Last edit at 07/06/2012 01:19PM by Sound.

I will try it over the weekend.

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

Printed a Raspberry Pi case this evening with new flow branch. Appeared to print well. Will print my perimeter calibration piece tomorrow and let you know how it went, sound.

---

Using ABSPrusa Mendel Zaphod with Pronterface and slic3r 1.3.0. Printing well with 3mm PLA and ABS through 2 x J Head Mk IV b and Wade Geared Exruders. Controlled using RAMPS1.4 board running Marlin_v1.1.9

I found no improvement using the new-flow branch over the previous releases concerning this problem. When I try slicing something at 0.15mm layers and have the nozzle diameter set to 0.5mm it tries to produce a path 0.7mm wide which is way to wide for it to print and causes skipped steps on the extruder, paths not touching etc.

My finding over the past two years of printing low layers (never printed above 0.25mm) is that the lower you go the closer to the nozzle diameter you need your path. Once you reach 0.1 you need the path equal to the nozzle diameter. At 0.15 you need the path about 0.02mm wider then the nozzle. At 0.2 you need the path about 0.05mm wider then the nozzle. At 0.25 you need the path about 0.075mm wider then the nozzle diameter. Above that is a total mystery to me as I have never printed any thicker then 0.25mm. It turns out the reason I learnt to print at low layers is because my first machine was a bowden cable machine and low layers are required when printing with a bowden cable (netfabb for Ultimaker default is 75micron layers for a reason).

---

FFF Settings Calculator	Gcode post processors	Geometric Object Deposition Tool Blog
Tantillus.org	Mini Printable Lathe	How NOT to install a Pololu driver

I print widths from 0.525 to 0.72 through a 0.5mm nozzle but I always use W/T 1.5 for optimum build speed and horizontal resolution.

Regardless it should not lead to filaments not touching, that means the flow rate is wrong. The effects I get are it snaps when bridging if too thin and sags when too wide and secondary effects like holes shrinking more when too thin.

Interesting you go closer to the nozzle aperture when using thinner layers. That implies die swell gets less and that would seem to make sense because the plastic has less distance to swell over. That could lead to a formula for width with die swell, nozzle diameter and layer height.

I must admit I haven't printed much with lower layer heights because I don't get good results, perhaps this is why. I tend to keep W constant and just reduce T, whereas I should reduce them both.

---

[www.hydraraptor.blogspot.com]

Hi, I'm not sure if you guys have seen this, I read the thread yesterday and can't recall anyone mentioning it, apologies if so.
It seems that reifsnyderb actually drills the J-heads he makes slightly undersized (0.4572mm for a nominal 0,5mm nozzle) to account for wear. This explains the "lack" of die swell at least.

Pondering the Nozzle Orifice Drill Bits

I just stumbled across this when I looked up some details for my marlin config, had this thread fresh in mind and figured you would like to know...