When I set out to design this hotend in January I had no idea, just how many ways there are to fail with a hotend design. By now I'm confident that I have encountered at least more than half of them. Really. I hope....
So my buddy machined the parts and I put them together, full of confidence that it would just work out of the box. Such a simple design, how can it go wrong- right?
Without even having the hotend hooked up to my Reprap yet, the first problems immediately became obvious. Let's take a look at the design:
Fail 1.
The melting zone, where the filament is heated by the aluminum block, is 5mm long. This only works for VERY slow extrusions. otherwise even the 1.75mm filament does not have sufficient time to reach temperature and melt while it travels down the block.
Fail 2.
I have no recollection how this happened, but somehow I specified a 7.0mm diameter hole to fit a 4mm diameter resistor as a heater. There was lots of slop that needed to be filled with carefully wrapped aluminum foil around the resistor.
Fail 3.
And there's something else wrong with the design. It shows the thread in the block bottoming out. You can't machine a hole like that with a perfectly cut thread all the way to the bottom. The effect was that when you screw the barrel into the heater-block, there is a gap between the end of the barrel and the beginning of the melting zone. In my mind this was just going to fill up with molten filament, but have no real effect on the extrusion.
That appears to not be the case. It caused a certain springiness of the filament feed. Pushing the filament in by hand, I could actually feel it springing back, when I let go.
The barrel.
As you can see in the drawing, the one end is threaded and that's all the machining I had planned. But in order to avoid the gap between the barrel and the melting zone, we need to turn down the last few millimeters of the barrel to a diameter smaller than the diameter of the core drill for the internal thread on the heater block. Easy enough, just put it in a drill, spin it and apply a file to the end.
And there's more fail to come!
The Reprappers' fever has finally got to me and I just HAVE to design my own hotend. Because. You know. All commercially available hotends are awful and/or too expensive. I am sure they can be improved upon. Any kindergartner could come up with a better design!
If it works, I would like to sell it. A friend of mine has a machine shop with a few CNC machines, so we can make small production runs.
Design considerations:
- Must be able to handle high-temperature extrusions (>=300ºC)
- 1.75mm filament (I think 1.75mm is the future)
- Use with standard type or with bowden-style extruder
- Low cost
- Raw materials can be easily obtained at a good price
- Easy to take apart
- Easy to clean
- Simple construction
- Few custom machined parts
- Keep machining times short
- Minimize tool-changes
- Precise
- Deliver high-quality extrusions
1. Barrel
For high-temp extrusions the conventional wisdom is to use stainless steel for the barrel. Why? AFAIK there are several reasons:
- It can handle high temperatures (PTFE and PEEK can only handle up to 250ºC)
- It won't corrode when used in a hotend
- It's pretty strong
- It has reasonably low thermal conductivity ( around 14-16 W/mºC for ASTM A269)
McMaster offers an interesting looking seamless stainless steel tube with OD 4mm and ID 2mm. Perfect. 1.75mm filament will fit nicely and we can just connect it at the cold side to a 4mm OD PTFE tube with a compression or push-to-connect fitting for a bowden extruder.
Unfortunately Stainless is not the easiest metal to machine, so we should try and keep the machining of it to a minimum.
Then we need a heater block, a nozzle and a heatsink.
2. Heater block and nozzle
In order to keep things simple, we'll keep the nozzle as part of the heater block for now. Similar to the J-head extruder. This makes it harder to change nozzle size, but will keep machining, and thus cost, down.
3. Heatsink
The heatsink has the job of keeping the transition zone as short as possible. It is made from aluminum and held at the bottom by a nut and at the top by a clamp.
The whole thing should look something like this
