Yet another all-metal hotend design

After having read countless posts on all-metal hotends, and problems assosiated with them, i thought it might be usefull for someone if I posted my design.

It hardly gets any simpler, only drawback is that you properly need a lathe to make it work, and the extruder base plate needs to be made of aluminium.

This is version 2 of it, made with 6mm threaded rod:


The feeder tube is made from stainless steel, and is ~40mm in length with a 3.2mm bore, on version 2 I polished it on the inside with Brasso, though it properly isn't needed.
The threads is deliberately not removed from the feeder tube for 2 reasons, they are usefull when assembling the hotend, and they allmost doubble surface area for added cooling.
To be honest I don't know if the surface area from the threads helps on the cooling, but on a 6mm rod with a 3.2mm hole through, you really don't have much material to remove, before it starts to have an impact on structual integrity.

Heater block is copper on version 2, and brass on version 1, both seems to work eqally well, but copper conduct heat about twice as fast, which in theory gives more presice temperature control. Problem is that silicone is not recomended for copper, as it is acidic, but so far it seems to work fine. But if I need to remake it someday, I'll properly use brass.
For 6mm rod 10mm is a fine height for the heater block.

The tip is made from threaded brass rod, and can be changed relatively easy. I found that an oriffice length of 3-4 times the nozzle diameter works well.

Tip and feeder tube is faced of in the lathe where they meet inside the heater block, so far this have worked exellent to prevent leaks, nor have they some loose.

So far I have printed about 5kg of ABS and 1 kg of pla with this design, and I have yet to see it jamming.

I estimate the melt zone to be in the order of 10mm when printing, but I haven't tried really hard to measure it.. It could properly be shortened by adding some active cooling, or a heatsink to the feeder tube, on the expense of a bigger power budget.

I use a 6.8R resistor in version 2, and @220C with ~11.5V supply voltage duty cycle is ~50%

Version 1, made with 8mm rod.



Edited 1 time(s). Last edit at 05/14/2013 10:54AM by Ralf.

Amazing, I just did the same thing yesterday! With one slight difference that I drilled the nozzle directly into the threaded rod: one less part and one less chance for a leak (but no way to change the nozzle). I had to do it a couple of times though, because I drilled it too much the first time (I have no lathe and used a vertical drill press...).

Now I did not expect it to work without a heatsink on the threaded rod! Don't you use active cooling? How hot did you go so far?
cheers

---

Blog: [tridimake.com]
Designs: [www.thingiverse.com]
Profile: [www.linkedin.com]

Excellent idea to simply keep the thread! You get a cooling surface size comparable to those hotends with grooves turned in.

Not sure wether it's neccessary to put in a distinct rod for the nozzle. It's true, stainless steel isn't exactly a first class heat conductor, but you have to conduct the heat over just a 3 mm distance.

---

Generation 7 Electronics	Teacup Firmware	RepRap DIY

Apart from the possibility to exchange nozzles, there are a coupple of reasons why I have made it in two pieces.

I think that Stainless steel for the nozzle would work fine also.

But machining nozzle and tube in one piece reduces the margin for errors quite a bit. And my lathe is too big and clumsy to drill a 0.35mm hole without breaking the drill bits, leaving them embedded in the workpiece.

I also drill the input hole in the nozzle with a 90 degree center drill, this should in theory give better flow compared to a standard 116 degree bottom, and it also leaves more material to conduct heat all the way down the nozzle. And my center drills is only about 30mm long, so they wouldn't be able to reach all the way to the bottom of the tube.

And no.. no active cooling, I normally print at 230c, but I only think that this is possible bacause of the aluminium base plate I use for the extruder.
I have tried to catch the temperature gradient with a thermal camera, but the uneven surface of the threaded rod together with the low thermal emissivity of stainless steel makes it dificult to get any usefull readings.

But assuming a relative linear temperature gradient down the 30mm of exposed tube, 40c in the top and 230c in the bottom, this would mean that the glass transision zone would be approximately half way down the tube at idle, but properly much closer to the heater block when filament is pushed down the tube.

This is strikingly similar to my first ever working hotend. Way back before I thought about going into business selling hotends!

I built a Mendel90, but instead of a printed X-Carriage I milled one from 3-4mm aluminium plate. My barrier was just a piece of stainless tube like yours, coupled to a brass nozzle made from a brass bolt with an aluminium heater block.

The one difference between mine and yours is that I did turn down my stainless tube to have a 3-4mm section of stainless tubing with only 0.5mm or so of wall thickness. I was using 1.75mm filament with a 2mm internal bore. This meant that I had a very short overall profile instead of the long stainless tube like yours, but I did have to put a computer fan and heatsink onto the carriage to cool it.

It still works, and I run the machine often. The super low profile makes for a very short distance between hob and melt zone which I think helps quite a lot. I get very precise prints from this hotend.

I'll try and get some pics up at some point.

---

I Design/Sell all-metal hotends. My company is called e3d-online - you can buy at [www.e3d-online.com]

So I finally got some PLA to test with, and it ended up in pure fail..

It simply doesn't work reliably without cooling of the stainless tube, for a start it all looked ok, but when things heated up, the hotend started jamming. When not printing, I could extrude all day long both very slow and very fast, but as soon as I started printing it jammed.

So for PLA I have added a heatsink starting 5mm above the heater block, and this seems to have solved the jamming issue, but I guess that's pretty much the standard way of fixing PLA jams in all-metal hotends..

And yes Sanjay it looks pretty much the same principle, simple and relatively easy to manufacture, given that you have a lathe.