E3D V4 All metal hotend

[mhackney]

The root of the problem I describe above is, indeed, friction. Friction from the increased diameter of the filament AND especially from the roughed up surface of the filament. I've now printed 20 sets of parts in the lemon drop yellow with out fail, all perfect. I'm sold on this resolution, at least for PLA. I'll be testing ABS and other materials soon. As long as this drive gear can push these other filaments it will be a winner in my book.

[ceramichammer]

I switched current to 220 and uploaded that... and did a couple of other things.

- Put an aluminium foil shield between the lower part of the fan and the heat block in case there was some air washing over the thermistor and giving me bad readings.
- Put slic3r back on it's default extrusion width settings (which I don't like)
- Switched the PLA to a brand new unopened pack.
- Set hot end temp to 200

I actually managed to finish the print! It was a little wobbly from heat (I think it's definitely too hot for this pla) but it actually finished! I was ready to say that my other PLA must have been to humid, but thought I should do another test first...
Exact same print, but back to my custom extrusion width settings and 190 degrees on the hot end... Couldn't even get it to extrude again... Ok back to 200, wait for the heat up... Nothing. No extrusion again.
Cried a little...

Manually reversed the filament and could feel that same rough resistance that I've felt before. Played with it a little, trying to figure out what it is that I'm feeling. Took the bowden tube from the hot end and manually moved the filament forward and backwards again (I've done this before of course, just doing it again for old times sake), and did feel a different, yet definite resistance. I wouldn't say that it's probably enough to stop an extruder, but it's something that shouldn't be there, and perhaps once added to the resistance of the hot-end it's compounded.

I think I've isolated it to the PTC connector on the extruder. It's a 3rd party one since the stock one blew out on a failed print a while back. Perhaps it's opening is smaller.
I have some more of the e3d PTC connectors, So I'm going to print an ABS version, tap and fit the e3d PTC to that and reinstall.

Isn't it wonderful to be able to print the parts we need?

I'll update later.

I have the E3D-v5 3mm hotend on an ultimaker clone that only prints PLA. When I first installed it, I had great success. All of a sudden, I'm having major issues with jamming. It could be a lot of things. I'm having a hard time organizing all this info so I'm just dumping it.

Even when the machine was rock solid reliable, I occasionally had problems loading filament. It felt like the filament would get hung up after it entered the heatsink and before it came out the nozzle. In trying to diagnose my recent print failures, I discovered that thru hole of my heatsink is not concentric with the thru hole in the heat break. This creates a lip that the filament gets hung up on. I'm also suspicious that the track marks created by the hobbed bolt get caught on the lip. (As I'm typing this, I realize that after every failure, I pull the filament out, cut off all plastic with track marks, and reinsert it. Maybe the failure occurs when the machine has extruded enough plastic that the track marks get to the lip.)

Here is a picture of the lip that I'm talking about.

[img]http://i.imgur.com/a45G6a5h.jpg[/img]

Possible problems:
"sticky" PLA
oblong PLA
E3D manufacturing error
switching to smaller nozzle



First few weeks of successful printing: 220C, 0.6mm nozzle, retraction enabled:
5x successful 10 hour prints with "smooth" ultimachine filament
plenty of successful small prints with other filaments laying around the shop

10x print failure within 30 minutes of clearing the jam: 220C, 0.4mm nozzle, retraction enabled:
Filament from JustPLA.com surface finish has a texture. It is not glossy smooth.

[mhackney]

Have you attempted to remove that lip? Seems like a good thing to do and report back.

[626Pilot]

At precisely that moment I looked at the box on my desk overflowing with extruders, hot ends and electronic components and caught glimpse of a hobbed gear on a direct drive extruder that I experimented with last year. It was magnificent, the teeth were at least twice as fine as the stock EZStruder teeth and not nearly as deep or sharp. "Hmm, can I substitute it for a quick experiment?" I thought. Turns out it is a direct replacement for the stock! I installed and aligned it, pulled the gnarled up filament from my Bowden.

This is why I come here. This is the stuff! I'm going to order one of these and see what happens. It makes sense. It will require extruder recalibration, won't it?

[McSlappy]

My God, Mhackney, have you solved my problem?!?!?
It just so happens that the PLA filament I'm using sits in the range of 1.78 to 1.8mm diameter, a fat PLA by your reckoning. You are also showing failed prints that look just like mine - the filament starved variety.

I almost cannot bear to hope that this might be the case.

I also have not received my new PLA (it's taking a looong time)

I also cannot bear to wait for a new hobbed gear...

Gah suspense will be my lot I suppose.

I'll report when I know more.

[McSlappy]

My money is still on friction somewhere between the extruder and the hotend. Did you resolve that catching you felt where the bowden meets the heat sink? I also put a very slight bevel on the entrance like I believe you did. Do you feel a catch where the aluminum heat sink meets the stainless heat break? How about heat break to nozzle?

I'm not sure it's exactly where the bowden meets the heatsink. I tried removing the bowden, heating up the hot-end and just hand feeding filament in, it was smooth as butter! But then again I think this is a problem which doesn't always present itself reliably.
When I did find a jam, I removed the extruder end of the bowden and tried pulling and pushing the filament, which felt very 'catchy' and took a lot of force to reverse and then to forward. Once I did get it to forward, I felt it melt quickly and flowed normally without much friction.
As to where this might be occurring within the hot-end.. No I have not figured that out. My multiple inspections of the hot-end have turned up nothing that looks like a catching spot for the filament.

One more thing to consider if you haven't is your bowden tube itself; I have one that turned out to be just a tad narrow, and when a slightly larger or more eccentric filament would get deformed by the drive gear it resulted in enough friction getting the filament through the tube to cause problems.

Good luck!

My e3d bowden tube is narrower than the stock one, which I was actually pleased with because I believe it cuts down on the filament flex within the tube... Perhaps I should swap it back and trial it with the stock tube.

At this point, I thought I'd wait and try the new PLA and see what results I got from that before continuing - could be I need a thinner filament if Mhackney's results say anything. Or a finer hobbed drive gear.... The filament seems the faster thing for me to check, and it gives me a mental break for a while which I think I'm beginning to need here

I'll report back.

[mhackney]

In looking at the stock hobbed gear closer, I think it can be "detuned" by hand. Of course, if you have a lathe then that would be best. Chucking in an electric hand drill would come in a close second. Finally, this could be done by hand:

lathe or drill - spin the part and touch a triple folded piece of 400 grit emery aka wet/dry sandpaper in the groove. Try to fold the paper to make a radius similar to the hobbed part. Don't take off to much, do several light passes and look at the part with a magnifier to see how far you've gone. I think if you just take the sharp edge off, things will be much better.

by hand - simply sand in the groove with the folded sandpaper. It shouldn't take too long (maybe a few minutes) to break the sharp edge.

Also, yes, with the new finer tooth hobbed gear, recalibration is a good idea.

[McSlappy]

Do you think that it's the deformation of the filament that's the issue?
Wouldn't either; replacing the hobbed gear as you have done or reducing the spring tension achieve similar results?
I'm a little afraid of milling the current hobbed gear down in case I reduce it's ability to grip filament while keeping it's 'crushing power' the same. Thoughts?

[626Pilot]

I'm a little afraid of milling the current hobbed gear down in case I reduce it's ability to grip filament while keeping it's 'crushing power' the same. Thoughts?

I just paid $12.94 for the other drive gear so that I wouldn't have to think! Thinking takes too long and it's dumb and it sucks.

[mhackney]

In looking at my other filaments, they also are very marked up. I think it is just a statistics thing, that these may get hung up at some point too, just less likely to do so because the effect is not so pronounced.

BTW, I'm suggesting to SAND not machine and to simply dull the sharp edges of the knurling on the stock gear. It should be a safe operation. Don't go gang-busters on it and take all of the knurling off! Less is more.

[McSlappy]

I'm a little afraid of milling the current hobbed gear down in case I reduce it's ability to grip filament while keeping it's 'crushing power' the same. Thoughts?

I just paid $12.94 for the other drive gear so that I wouldn't have to think! Thinking takes too long and it's dumb and it sucks.

Ha ha ha.

In looking at my other filaments, they also are very marked up. I think it is just a statistics thing, that these may get hung up at some point too, just less likely to do so because the effect is not so pronounced.

BTW, I'm suggesting to SAND not machine and to simply dull the sharp edges of the knurling on the stock gear. It should be a safe operation. Don't go gang-busters on it and take all of the knurling off! Less is more.

Who knows what micro differences exist in differing filaments...
I think I'll order a MK7 drive gear and leave my stock one alone, at least then I can return to full stock if I need to.
Now I've got to wait ages for shipping :/

[McSlappy]

Well it's been a few days.
I was hoping that my new filament would show up and I'd be able to rule filament out of the equation, alas it did not. I was also planning on buying a drive gear, but opted out of it... Not sure if it was the thought of spending $30 on a gamble (when I don't have $30) or having to wait a week for shipping.
I did end up lightly sanding the drive gear as Mhackney suggested. It was a terrible idea, the 80 grit paper just tore up the teeth and now it don't work!

I am of course joking, I used 300 grit and lightly tickled the teeth while the gear was on a drill bit.
After cleaning the teeth and the rest of the extruder, I put it all back together and ran some tests. All seemed good. I also turned on 'wipe' in slic3r (why have I never tried that before ??)
I was thinking that perhaps the problem was solved and so I ran a large print... It ran very well for 3 hours and was printing the nicest PLA I had ever run. I was super stoked but kept some skepticism, checking on it every 20 mins while I watched a movie. I ended up going to bed confident that it was going to run perfectly based on how beautiful it had run so far.

I eagerly checked in the morning and discovered that it had failed in the night, not long after I went to bed

I left it for a few hours.

I had been mentally running over what possible causes there might be, mentally questioning them and mentally checking them off...
I came back to a cold machine, and sliced a simple print with the filament temp at 220. Almost immediate plugging. I checked the heatsink temperature with my finger and all seemed ok. Normally with a jam like this I can reverse the filament by hand and push it back through, feeling the melt and having it clear out. This time I could not physically push the filament through. Whatever was stuck in there, was stuck very well!
I pulled the filament out (which also required some strength) and tore down the head once more. I knew the filament was sticking somewhere in the head, so I thought I'd check again... Maybe I missed something.

Pulling it apart this time I noticed that the heat break tube (I think that's it's name) was a good 50% full of solidified filament, while the nozzle was clear enough for me to see light through.
This seemed to me like a heat problem, which has been pointed out by many of you.
I wondered.. If it's heat, then why does the heatsink feel barely warm? Also the fan is on 100% of the time powered directly from the board's 12v... How could the heat creep like this?
I've considered this many many times. I've tested new fans in case the current one is too slow, but they all behave just like the stock e3d one - same speed. I've pentuple-checked the airflow to make sure it's traveling THROUGH the hot end and not away. I've felt the fins many times to make sure they aren't too warm. I had lightbulb moment and thought that perhaps for some reason the fan wasn't reaching 100% speed, which would make all my other fan tests also fail since they were on the same power source. Switching the source to FAN 0 and setting it to 100% produced the same results.
Somehow, for some reason it seems I have heat creep in my head.
I did just run a print now at 180 temp and it completed, where others had not... I will now run a larger, multi-hour print at 180. Perhaps I can control the heat creep by just printing at a lower temp. I say "lower" but it's what I used to print PLA at. I've had my PLA fail at this temp, and it's the temp that I started at, but maybe I missed something along the way.

I still don't have a definite solution, but I thought an update would be in order. Avagoodweegend!

PS. I'll have some better troubleshooting done when I get my temp sensory thingy and my new filment.

[Eaglezsoar]

You have been through hell with this thing, I do hope that you get it resolved. Once you get your thermocouple meter I think you may discover
the source of your problems.

[mhackney]

You made my heart stop with that "80 grit" bit! Nice joke.

So, that is actually very telling. By removing the sharp edge it sounds like you were having somewhat better results. Maybe smooth the teeth a little more (use 40 grit this time) and try again.

On the plugging, 220° for PLA is high. If you have molten PLA creep up the heat break and then a few blasts of cool air happen at that little bottleneck area (even from rapid hot end movement) there is the possibility it could solidify momentarily and lead to a jam. It is likely a statistical thing. I'm wondering if a little Kapton tape wrapped around that area might not eliminate this by insulating that area. And turn down the PLA temp to something "reasonable" like 190-195°. Do you get get "air extrusion" flow at that temp, if so, print at that temp.

EDIT: forgot to add, the reason the heat sink doesn't feel warm is 1) the stainless steal heat break is much less thermally conductive and 2) the heat sink has a lot of mass and surface area to radiate heat. The insulating properties of the SS break are a blessing and a curse. Blessing because it does effectively insulate the hot zone from the cold end. A curse because it can insulate too effectively and if the melt zone is moving up and down a bit, could "freeze it" in place. This is what I think is happening to you.

Cheers,

I'm off now for an 8-9 hour drive to visit EaglezSoar!

Cheers,
Michael

[McSlappy]

I like jokes.

I hope your trip was good. Perhaps I'll come to the worldwide Rostock meetup when it's next on.

Yes 220 is way high, way too high I think. I have been theorizing that my setting higher temps has actually accelerating the jamming by causing faster heat transfer up the heat break... This makes no sense if the fan is doing it's job...

I get air extrusion starting at 165, I feel that this filament runs nicely (not too stringy, not too thick) at 180 degree range.

Do yo mean Kapton tape around the thin part of the heat break?

Eagle, you're right, I'll know a lot more once I get that thermocouple meter All this waiting is killing me... I should take up another hobby.

[Eaglezsoar]

You chose the right hobby, everything will work out fine.
Patience is a virtue they say.

[McSlappy]

Yes, this IS the right hobby - I just mean something to do while waiting for parts :/

In the meantime I am printing some spiral vase prints without issue, so that's kind of fun.

[cyrus104]

I have the E3D V5 but I wanted to check and see if anyone could provide their slic3r settings for this hotend running a .4mm nozzle. I am looking for the layer heights and the Advanced / Extrusion Widths. I am running into a few issues where it sometimes feels like I'm over extruding and the nozzle is dragging through the ABS and then other times when it clearly under extrudes and there are clear gaps between lines on a flat solid surface.

Thanks

[McSlappy]

For my .4mm nozzle on ABS I'm using
.2mm layer height
.2mm first layer

2 perimeters

2 base layers, 3 top layers

.2 fill density

30mms speed on everything, but 50% on external perimeters and small perimeters
10mms on gap fill
50% first layer speed

150mms move speed

4mm skirt (to keep that sucker held down)

.5mm extrusion width on all widths (slic3r's defaults are not good)
.55mm on the first layer because reasons.
.4mm on perimeters (I'm only doing this right now because I had some thin walls that I wanted to fill without the crazy patchy zig zags that slic3r does)

4mm retraction
.3mm z lift
50mms retract speed

Let me know if this helps

Tell me if this helps.

[McSlappy]

UPDATE.

Got my new PLA filament today. Measured it at between 1.78 and 1.8mm - seemed like a very consistent roll.

Printed a spiral vase with no problems.

My next print failed on layer 1 after completing it's 2 outside loops and starting the actual object

I have decided that I must have on of the rare possessed heads that Sanjay mentioned. I will send it back and get a replacement, yay for customer service!

In a few weeks I guess I'll get back with you about the results

[cyrus104]

That helps a bunch, we are pretty close on everything but the advanced settings. Before you posted I changed a few of them to 150-200% and this made a good difference but I'm still looking to get better. I'll plug your numbers in as soon as this current print finishes.

Is there any reason you go with .3 lift over .1 or .2?

Thanks

[McSlappy]

You'd want your z lift to be at least your layer height. I could have gone .2 since my layer height is .2mm, but I'm doing .3 just to be safe

[cyrus104]

The settings worked good, I needed to change a few things. I tried it at .1mm layer height but that's didn't work out so well. It wasn't add smooth as the .2mm.

The retract lift was good at .3.

[mhackney]

I've noticed that both the Kraken and E3D - which are essential the same hot end/heat break - may be sensitive to too much retract. I had run with 4mm with my SeeMeCNC and JHeads and had forgotten that I tweaked it down to 1-2mm for the E3D. I just realized this over the weekend so I tweak it back up to 4mm to see what would happen. I definitely got skipped extruder steps. This is very analogous to the "hydraulic plugging" phenomenon I wrote about a looong tome ago with the Steve's extruder on an H-1 printer. When I turned it back down to 1.5mm, all was well again. The effect this has depends a lot on geometry, fill, etc. So it is possible you may bump into it on some parts and not at all on others.

My theory is that the melt zone in these hot ends is so small that 4mm represents a large % of the melt volume. My estimate is 33% as derived:

1.75mm diameter
hot end melt zone about 12mm
retract 4mm

melt zone volume = Pi((1.75/2)^2)*12 = 28.86mm^2
retract volume = Pi((1.75/2)^2)*4 = 5.78mm^2

So 5.78/28.86 ~= 33%

That probably results in a lot of molten plastic "pumped"up into the heat break where it can solidify as it cools down, causing a micro plug.

That's my theory anyway!

[Eaglezsoar]

A mathematician too? Is there anything you can't do?