Replacement Hotend (E3D)

Hi,

[quote=“bour12@casema.nl”]I was looking in Thingiverse fo the number 221997 but can not find it?[/quote]probably because the search function is a bit sketchy. The advanced search would have found it, but then again, the link for the advanced search itself is hidden at first …
You can simply append the “thing:…” to the base URL, like this: http://www.thingiverse.com/thing:221997.

Cheers,
kuraasu

Maybe you can use teflon, it only melts on 317°C
I use teflon now around the head (original extruder) for mounting the thermistor instead of the washer with screw

kuraasu, would you mind sharing your Slic3r ABS settings for this hotend?

I managed to provisionally install this hotend yesterday with an aluminium bracket (Ill soon print your holder for a more permanent fix).

So far, I did only one print using the default Velleman ABS profile (including incorrect nozzle width setting, I only realized that later). It came out better than expected for a first print, but with several large blobs that seemed to happen upon restarting after a retract and/or layer change. Clearly Ill need to tweak something, so Im hoping your settings will give me a headstart. Other than the blobs, it looks promising. The flow is very smooth, it heats up very fast. I hasnt clogged yet :).

For the record, I used the original hotend heating element so I can maintain 15V for now, Im powering the fan from a separate 12V PSU. and fixed the supplied thermistor with only kapton tape and some alu foil to fill the hole. During the “burn-in” at 300C, it seems some overheat safety feature of the controller board kicked in. I havent looked yet, but its probably something you have to change in the firmware. I did get it up to 295C or so before that happened, so that was close enough to tighten the nozle (which isnt leaking, hurray :slight_smile: ).

One more thing; I havent yet upgraded the heat bed, but by doping the kapton with acetone/ABS mix, the model sticks VERY well to the plate at 50-60C. So well I actually broke my little frog test print trying to peel it off. Ill need to experiment with how much dope to apply and probably raft settings in slicer so I dont ruin my kapton tape on every print.

Played some more with slic3r settings and Im getting close. Here is my third print, using 0.15mm layer height and glow in the dark ABS:

Not perfect yet, not enough plastic in some places, but far better than I had hoped for!

Settings Ive changed so far:

extrusion multiplier: 1 (I probably need to increase that a bit)
retraction length: 1
extra lenght on restart: 0
nozzle diamter: 0.4

The rest is as in the velleman default abs profile.

Off to print kuraasu’s mount now, though Id like to change it accommodate some LEDs.

Hi P4man,

[quote=“P4man”]Settings Ive changed so far:

extrusion multiplier: 1 (I probably need to increase that a bit)
retraction length: 1
extra lenght on restart: 0
nozzle diamter: 0.4[/quote]I don’t have the profile available atm to look into it, but you pretty much got the main points that I adusted for the E3D by yourself as I can see. I tried layer heights between 0.09 and 0.18 mm (0.03 multiples due to 1.5 mm lead on Z …), however, the surface quality when going down from 0.15 / 0.12 to 0.09 did not improve as expected, so I stuck to the “mid-range” values from then on.

Extrusion widths between 0.4 and 0.5 mm all worked well with the 0.4 nozzle; for the top infill I went down to 0.35 or 0.3 mm.

For the overhang present in the mount, cooling is a must. At least having the fan at 40 to 50% was enough in my case, limiting the cooling effect on the heatbed as far as possible.

Cheers,
kuraasu

[quote=“kuraasu”]Hi P4man,
]I don’t have the profile available atm to look into it, but you pretty much got the main points that I adusted for the E3D by yourself as I can see. I tried layer heights between 0.09 and 0.18 mm (0.03 multiples due to 1.5 mm lead on Z …),
[/quote]

I should have thought of that :slight_smile:

The effect of cooling keeps confusing me. I thought for ABS you wanted as little as possible, only for bridging?

I accidentally turned off all cooling while printing your mount. The overhang printed ok, although it curled up pretty badly. I lowered the temperature from 240C to 230C midprint and that seems to have helped some. I should probably lower it further, but Ill experiment with that later. I got halfway through the print and then the gear on the extruder motor started slipping. Cant tighten the little bolt anymore, it turns “mad” so I need to tap a new screwhole for it… dont have that equipment right now, so further experimenting will have to wait. BTW, what the hell is with the spare gears not having a screwhole?

edit: fixed it. Turning on the fans clearly helps in fighting curl with ABS. Not a hint of curling this time.

second edit: Fail! It looked good, and no curl whatsoever, but I guess I printed too slow or with too much cooling, the print basically fell apart. Very little inter layer adhesion, I can crack it open at any layer with very little force. Experimenting continues :slight_smile:

Did you use the automatic cooling routine? In my case, it often adjusted the fan speed to a level too high (for ABS), so that the heatbed temperature dropped significantly. Luckily, the prints did not come loose, but still I didn’t want to take the risk. So, I tried to set the cooling to a level which has only little to no effect on the bed temperature, but still produces good print quality. The speed reduction integrated into the auto cooling routine has been quite benificial for delicate objects. Therefore, depending on the object to print, I switch between manual control of the fan and automatic control with adjusted parameters, e.g. max. fan output (to keep the auto routine from driving the fan at high rpms).

[quote=“kuraasu”]
Did you use the automatic cooling routine? [/quote]

Yes, both for the fan and for slowing down the print. I did reduce maximum fan to 35% as I have 2 fans and one is fairly powerful and well funneled. Still even with low to no fan Im seeing poor inter layer adhesion. Any other tips to solve that? The prints look good, until you pick them up, they fall apart with almost no effort. My PLA prints where much stronger.

What’s your extruder temperature, layer height, extrusion width, and extrusion multiplier? Did you yet calibrate the latter value? Any special material or “normal” ABS (the one from Velleman)? You mounted the original thermistor that came with the E3D, right?

What’s your extruder temperature, layer height, extrusion width, and extrusion multiplier? Did you yet calibrate the latter value? Any special material or “normal” ABS (the one from Velleman)? You mounted the original thermistor that came with the E3D, right?[/quote]

IM printing at 245C (tried 230-250, little to no difference). I am not using the velleman thermistor (I cut it too short and soldered it, so I couldnt reuse it), so I used the supplied thermistor; however I did test the K3D one in ambient and boiling water, and it gave almost exactly the same readings as the original one. Also I can manually extrude ABS from ~210C onwards, so I dont think the measurement is way off.

Layer height is now 0.21mm.
Extrusion width settings are the default for the velleman ABS profile, ie 100% default (120%, 175%, 180%, 110%, 110%, 0)
extrusion multiplier is 1. If by calibrating you mean changing that setting and seeing what happens, then yes I guess I “calibrated” it. If you mean something more objective, then no. Is there another way? FWIW, I tried printing some herringbone gearset. With multiplier 1.1 they were fused together, with multiplier 1 they broke free with a little effort and they rotated as they were supposed to, so this seemed okay to me.

Ive tried 2 different ABS filaments, one “glow in the dark”, and one ordinary blue. They are not from velleman, but recommended temperature range stated on the roll is 220-250C.

BTW, I just completed your mount in one go, the closing of the fan enclosure at the top was a little… tense, but it looked good. However, it was again brittle enough that I actually managed to break it clean while trying to remove it from the heatbed. It seemed to snap particularly in places where I put the fan on, but elsewhere the adhesion was also too weak IMO.

Hi P4man,

[quote=“P4man”]Layer height is now 0.21mm.
Extrusion width settings are the default for the velleman ABS profile, ie 100% default (120%, 175%, 180%, 110%, 110%, 0)[/quote]Just to be sure, I can’t find a post where you mentioned that already - you’re using a TR8x1.5 on the Z axis?
Extrusion widths are a bit low for my taste, even though the 0.4 E3D nozzle of course has an advantage over the stock one with 0.5 mm. But those values are all far below the nozzle diameter. Copy your settings to another profile and change the widths to the following (examplary) values: 0.45 0.4 0.45 0.45 0.45 0.35 0.45.
You can also start another 0.05 mm higher at 0.5 0.45 0.5 etc. if you want, that should work, too.

[quote=“P4man”]extrusion multiplier is 1. If by calibrating you mean changing that setting and seeing what happens, then yes I guess I “calibrated” it. If you mean something more objective, then no. Is there another way? FWIW, I tried printing some herringbone gearset. With multiplier 1.1 they were fused together, with multiplier 1 they broke free with a little effort and they rotated as they were supposed to, so this seemed okay to me.[/quote]By calibrating I mean printing an object with a single wall and checking the wall width against the set extrusion width as described here: https://github.com/alexrj/Slic3r/wiki/Calibration for example. Instead of any special single wall stl file, do this with a solid object (cube or so) that’s sliced with 1 perimeter, 0 infill, and 0 top solid layers, which is the best way to produce the right G-code.
Note that ten percent is a much too coarse step for setting the extrusion multiplier, you should adjust that down to two digits of precision.

Cheers,
kuraasu

Care to explain? What does “100%” actually mean and how does it related to extrusion widths you quote? And how does it relate to the multiplier?

[quote]Copy your settings to another profile and change the widths to the following (examplary) values: 0.45 0.4 0.45 0.45 0.45 0.35 0.45.
You can also start another 0.05 mm higher at 0.5 0.45 0.5 etc. if you want, that should work, too.[/quote]

Ill give that a go, but would be nice if I understood what I was doing. I do see that when you reduce those numbers, the slicer produces thinner threads (but also more of them), but Im not sure I see the big picture.

[quote]By calibrating I mean printing an object with a single wall and checking the wall width against the set extrusion width as described here:
[/quote]

Aha! Thats already one thing I misunderstood. I thought you have to adjust extrusion width to make that work and already I noticed precious little difference, but when I printed that test cube, I was within a hair of the nominal width, so left it at that. Ill redo that to start with.

BTW, thank you very much for your time, its greatly appreciated. Im taking so much of your time, Im almost ashamed to bring this up, but maybe someone else can help with this; Id like to combine your mount with a fan shroud. Here is a mockup of the idea:

I used this shroud:
thingiverse.com/thing:120249

I would also make it for a bigger fan, like an 80mm. However, I can only work in sketchup and Im not even that good at it, no way I can produce those kinds of shapes. Maybe if I could start with some boolean operations on the existing models, but both STL models when imported in sketchup have like a million mesh errors (sketchup isnt very good at importing STL), so I guess Im stuck.

Does someone feel challenged to make a printable model like that?

Another useful potential starting point:
thingiverse.com/thing:147115

only the mount on the top would need to be changed for the K8200

To see if I correctly understood this calibration; I printed this single wall object:
thingiverse.com/thing:32471

with the extrusion width settings you suggested (0.45mm for perimeter). I measured the print and thickness varied between 0.4 and 0.5mm. Does that mean I got the extrusion multiplier right? For the record, its at 1.05.

I’ll split this up a bit …

[quote=“P4man”]IM using the same rod as the supplied one, M8, which I believe has 1.25mm pitch. I guess I need a
different multiplier for the layer height?[/quote]you do, see “Optimal layer height” calculator.josefprusa.cz.

Care to explain? What does “100%” actually mean and how does it related to extrusion widths you quote? And how does it relate to the multiplier?[/quote]Extrusion width in percent calculates in terms of the set layer height. So, with 0.2 mm layers, 100% e.w. will give you 0.2 mm wide extrusions - which is quite low for a 0.5 mm nozzle.
The percent setting is useful, when you want to ensure that the nozzle always creates the same cross-section. E.g. with a setting of 200%, the extruded line should always be a (flat pressed) ellipse with a width twice as large as its height.
For small layer heights, that will obviously produce problems, as the extrusion width will be small, too. The nozzle works best when the extrusion width is slightly larger than the nozzle diameter. It still works well if the width is equal to the nozzle diameter or a bit smaller.

For low layer heights, you either have to increase the extrusion width percentage in order to stay in the range of feasible values, or enter the extrusion width as a fixed value (in mm).

[quote=“P4man”]I do see that when you reduce those numbers, the slicer produces thinner threads (but also more of them), but Im not sure I see the big picture.[/quote]That’s true, Slic3r will automatically space the tool paths that it’s generating by the extrusion width. Thus, every layer will have the correct amount of infill. E.g. for partial (linear) infill of 0.25, there distance between two lines will be four times the extrusion width, so the material will fill 25% of the total volume. For a layer that shall be solid, the lines are printed next to each other with just the extrusion width as distance.

[quote=“P4man”]To see if I correctly understood this calibration; I printed this single wall object:
thingiverse.com/thing:32471

with the extrusion width settings you suggested (0.45mm for perimeter).[/quote]
That combination may have been a bad choice, since there’s no specification on how thick the designed wall is. If it’s 0.5 mm, and you slice it with 0.45 mm perimeter width, you won’t know what Slic3r actually prints - 0.45 or 0.5 mm (yes, it can “widen” such a line by producing a very fine zig-zag movement).
Use e.g. the second (solid!) box in here http://www.thingiverse.com/thing:5573, or this http://www.thingiverse.com/thing:152672 one, and slice it the above mentioned parameters to generate your “own” single wall object (actually, it’s single-walled only in the G-code, but that doesn’t matter).

It probably won’t be exactly 0.45 on all four sides of the object, but the variance should be lower than +/- 0.05 mm.
Measure at multiple points around the object, and try to “grab” only one or two layers with the caliper. If one value is far off from the others, leave that one out (perhaps there was a small error in the wall at that point?). Average over the other values, and adjust the extrusion multiplier accordingly. In the end, you should have a print with all the walls within, say, 0.02 or even 0.01 mm of the set value.

Received my E3D today - it’s printing right now using the provided adapter. Works like a charm!
Thank you very much for the adapter, guys. Really great work!

~ Tectu

I had some problems with my E3D, regarding filament jam in the hotend. Even good quality filament could get stocked.
The problem have been solved with the “canola oil trick” and I designed a small “thing” that fits the K8200 to pre-oil (a very small amount of oil) and clean the PLA just before it hits the extruder. Not one single jam since!! Just for letting you know if you get the same problem.
BTW in am using peanut oil instead with success.

The pre-oil:

Yesterday i printed the “tree frog” from thingsgivers. Layers 0.15mm in PLA:

[quote=“kuraasu”]
Use e.g. the second (solid!) box in here http://www.thingiverse.com/thing:5573, or this http://www.thingiverse.com/thing:152672 one, and slice it the above mentioned parameters to generate your “own” single wall object [/quote]
Okay did that, did both actually. Tried a few times and settled on 1.07 extrusion multiplier. That seems to give consistent 0.5mm walls and the cube is exactly 20mm. However, that hasnt solved the problem. I see infill layers are not getting quite enough plastic, there is some minor spacing between the threads.

ANd more importantly, layer adhesion is still insufficient. I can peel off layers with a little effort. Larger prints with thin walls even seem to split while being printed, looks like due to strong adhesion to the bed (actone/abs mix on kapton tape) and then shrinking as it cools pulls on the layers and sometimes a small crack will occur. My heatbed is as high as it will go (~65-70C on start, ~60C if the fans kick in after the 10th layer).

Layer height: 0.2mm

Im not sure what else I can do. Ive ordered a new heated bed, but I doubt that will solve the layer cohesion, since that problem also occurs further up.

Here is my full config:
pastebin.com/6TEduEVY

Hi,

jangermann: nice solution for the tricky filament!

P4man: just to be sure - the height of your prints is correct, right?

Cheers,
kuraasu

[quote=“kuraasu”]P4man: just to be sure - the height of your prints is correct, right?
[/quote]

Id say so. The 10mm high cube I measure 10-10.05mm.
To be clear: objects like that small cube are indestructible. Its larger, thin walled objects that I can peel apart. Its not dramatic, but bad enough that I cant easily remove those prints from my heated bed without destroying them (or peeling off the kapton tape every print). Last object I tried was that round fan shroud:
thingiverse.com/thing:120249

walls are thin, its never going to be very strong, but Im sure if I printed that in PLA with the old nozzle and config, it would work fine.