Cooling the @E3D-Online v6 hotend - this is something i’m struggling with for a while and only do see it on my Eusthatios (Bowden Setup) so far and even a bit worse with my new Hotend holder/carriage system. I’m attaching some CAD screenshots to clear things up.
As you can see I designed (and tested) two differen versions. One with a 40mm fan and one with the original 30mm fan. The original version seems to be better but still I do see some problems where I guess they are related to insufficient cooling: When extruding really slow (on small prints and also manually tested with 10mm per minute feedrate) i get underextrusion and then the filament that was missing suddenly comes out in big blobs.
I wonder if my design is so horribly wrong that this happens? Is the E3D really this sensitive when it comes to cooling? Are the back exhaust channels probably not big enough to get the air out?
Some data:
24V setup with 24V Fans
Fans connected to a mosfet output with 255PWM softwarewise wired to Repetier Firmwares extruder cooling feature (triple checked that it runs with 255).
Any idea how I could get closer to this? If anyhow possible I would like to avoid changing the carriage as this would mean quite some work but if necessary i’d do it of course.
In my experience underextrusion is usually not caused insufficient cooling. Normally this is caused by bad PID tunings and/or weak power on the hotend. But the last you’ll normally only see on fast prints of big parts.
Insufficient cooling normally leads more likely to jams. And in case of ABS i.e. it can sometimes be an advantage to heat the material up in the cold end a little bit. For PLA this normally leads directly to jams. But in general its a very dangerous way.
By the way, I made some tests with a heatsink based on the E3D concept but with 50% reduced fins and height and couldn’t see a big difference in the heating inside the filament channel (measured with a probe).
But which material do you print?
And by the way, 10mm per minute is not a lot. Or do we talk about 3mm filament?
Three things I have found about printing slow on a long bowden tube in materials that are more sensative to heat creep like PLA.
1.) Try and get the heat sink fan down as close to the heat break as possible. That transition zone is the key part. make sure air flows well across the last heatsink fin.
2.) Long bowden tubes are fighting far more friction on the way up to the print carriage than shorter lengths or direct drive systems. I have to run 5C+ hotter on my bowden tube systems than a similar setup in direct drive. This lowers the backpressure and as an added bonus makes our long bowden retractions more effective (in my experience).
3.) Wrap your heater block. Or use one of the new E3D silicone socks. I use a high temp cotton strips with Kapton. It can be tricky to cut out a pattern to work well around all the wires on the hot end. But I find getting the top and sides of the block insulated well has two main benefits. It keeps heat from the top of the block from radiating to underside of the fins and the heat break. And insulating it allows me to run my part cooling fan higher without dropping the temperature of the hotend. So my printed plastic solidifies faster without forcing my hotend to crank out additional heat trying to maintain temperature.
Just one additional comment regarding your design :
It generates a lot of back pressure for the fan. And axial fans are not very efficient in such a case. So I would try to reduce the pressure drop for the air flow in your design.
@Sven_Eric_Nielsen My main problem is PLA here and it’s not really underextrusion that is the main problem but that blob of material afterwards. I should have made some pics of prints or a vid showing this. Maybe my description was just problematic.
I think @Eclsnowman nailed the thing and also got me to an idea. The lower fins might indeed be a main problem and that is something I should be able to fix. Maybe also at first by modifing the part that is currently in the printer with a Dremel before completely reprinting it. Regarding the silicon shoes for the E3D I have tried them and have mixed feelings about them. I’ve also seen them failing horribly twice. The little hole was blocked out by some filament/burn and the whole sock filled up and material swallowed out on the sides. Not a good thing. Basically I would like them a lot and maybe I’ll try to build my own wrap or silicon shoe. Maybe I’ll try modifying the original ones and cut out a bigger hole for the nozzle. This should still work great.
But at first I’ll try to optimize my cooling. I guess @Sven_Eric_Nielsen is right about back pressure. My cooling channels aren’t good and I should try to add some room for air at least in the lower part. Problem is that the rods are crossing behind and the LM8LUU are taking up lots of space there. Have to find a better way to do that.
your link doesn’t work for me but I’ve looked through marcs posts and think I saw what you were talking about. Looks nice and is basically the same i’m using. My biggest problem is that the way I’ve setup bearings and everything is too compact for the E3Ds cooling needs. I’m already playing around with things now and there’s not much chance to imrpove cooling here. Guess I have to kick it away and redraw from the beginning.
I tried 4cm fan with genuine E3DV6 before, but the cooling is so good that when I fine-tuning the retraction upto 5.5mm, molten PLA was drawn up to the heat break forming a hard plug. Had to remove the E3D from the carraige, remove the fan, heat up to soften the PLA to clear the block. So I guess 3cm fan is good enough already. My room temp is around 25C.
For silicon sock, tried a self molded one. I was hoping to print ABS without parts cooling fan, but without air cooling, sharp corner still curl up a bit. There are improvement but not as much as I hope.
I used to get opposite to what you observed. Over-extrusion when print slow, and under extrusion when print fast. The top solid fill will have tiny gap between lines, but around the edge (when hotend change direction) material builds up. I originally thought that was perimeter overlap, but it can’t be tune away. Printing higher temp (lessen back pressure) solve that problem, but overheating is worsen (curled corner, fine feature etc)
Sometimes I wonder if all the excellent print photo are all PLA only.
Unfortunately i don’t see the effect with 40mm fan at all that you had but probably my 40mm fans were all too bad. Mine were barely better than the 30mm and I really find it hard to get cfm values they promise to have.
But nice idea to rotate the extruder fan by 45 deg. Will think about that.