Geared extruder nozzle
Release status: working
This is a very reliable design. I have used it to reprap all the parts for five complete Mendels and counting, and it has required no attention. It has four main parts: a PTFE insulator that separates the hot end from the RepRap machine, a brass nozzle, a brass heater block, and a PEEK support that holds the nozzle in place.
All the forces in the design are taken by the PEEK and two threaded rods. This means that the PTFE is not under tension. The PTFE is also inside the brass nozzle, which means that internal extrude pressure tends to seal the join between the two, rather than opening it up.
All the files for this design are in mendel/mechanics/solid-models/extruders/geared-pinch-wheel, or see here to browse them online.
Please also read the Variations section at the bottom of this page. Variations tend to be improvements...
Here is the drawing for the four machined parts:
Though it says brass, you can make the nozzle from brass or aluminium. Both should work well.
And here are the parts themselves together with the two reprapped extruder components to which they attach:
The brass nozzle has been screwed onto the PTFE, and lock nuts have been assembled onto the ends of the threaded rods (see below).
Bill of materials
|1||brass heater block|
|92 mm||M3 threaded rod|
|58 mm||M3 threaded rod|
|2||10mm square scrap al. sheet|
|~200 mm||Kapton tape|
|~100 mm||PTFE tape|
|1||3W, 5%, 6R8 vitreous wire-wound resistor, 200oC|
|1||glass-bead thermistor (100K, NTC?)|
|a few ml||fire cement|
|50 mm||2mm heat shrink sleeving|
The PEEK support block (being small) is not at all expensive. But suppliers won't sell small quantities of blank 10mm plate. This is a very good reason to crowd-source: buy a 10mm plate (try here or here); cut off what you need; cut the rest into blanks and sell them for a reasonable price on Ebay to other reprappers... Another solution is to use round PEEK Rod as described here
Machine the PTFE, the brass and the PEEK.
Making the brass nozzle is simplest if you make it from a longer length of 10 mm brass bar and do it in this order:
- Face off the end
- Mark the centre shallowly with a very small centre drill
- Drill the 0.5mm nozzle hole 4 or 5 mm deep in the end
- Turn down the cone, making sure to obliterate the centre you drilled
- Turn down the 8mm diameter
- Turn down the 6mm diameter for the outer M6 thread
- Cut the outer M6 thread with a die in the lathe tailstock, rotating the chuck by hand (turn the power off for safety); use cutting compound
- Part off behind the flange
- Offer up a 3.5 mm drill to the side of the nozzle and use a felt pen to mark on it where you need to drill to to just leave 1 mm or so for the 0.5mm nozzle; beware parallax...
- Reverse the nozzle in the chuck
- Face off
- Mark the centre
- Drill the tapping hole for the M7 thread
- Tap the M7 thread using a tap in the tailstock rotating the chuck by hand (turn the power off for safety); use cutting compound
- Drill the 3.5 mm hole to your mark; again beware parallax...
- Blow through with compressed air from the nozzle end pointing the other end downwards to remove any swarf
Cut the M3 rods to length.
Cut two pieces of aluminium sheet about 10 mm square and drill two 3 mm holes in the middle of them. These will spread the load from the tension in the threaded rods over the plastic parts of the reprapped extruder.
The precise diameter of the hole in the heater block will depend on the heating resistor you use, so get that first and measure it. You need a 6.8 ohm wire-wound resistor. Ideally get a vitreous one like this, though I have used silicone-coated ones like this quite satisfactorily.
At one end of the block, drill and tap and M6 through hole. At the other, drill a hole slightly larger than the diameter of the resistor. Finally drill a small hole to put the thermistor in. I used this thermistor, which needed a 1.5 mm diameter hole. Put the thermistor at the extruder end of the block away from the resistor, and at the bottom nozzle end. Make the hole for it about five mm deep.
Thanks to Nophead...
This device uses a resistor embedded in a metal block to heat the extruder nozzle. The resistor is the at the left of the picture. The block has a thermistor embedded in it to measure its temperature - the thermistor is the small device with the fine wires extending from it.
Both of the resistors listed above have a power rating of about 3 watts. You are going to run them at about 20 watts, but don't worry - the power rating given is for when they run in air and you are going to run them in a metal block which will take the heat away. They also have a top temperature of 200oC and you are going to run them a few tens of degrees hotter. Again, don't worry - the temperature limit is that beyond which you can no longer rely on the accuracy of the resistance; if you run them a bit hotter it won't damage them.
Fit the resistor into the block. I used fire cement diluted with a little water to hold it in place. You could also use a high-temperature glue, such as JB Weld (according to nophead's post JB weld isn't suitable) . If you use the fire cement, place the block on a cooker plate on its lowest setting for about 15 minutes to set; alternatively put it in an oven at 200o C. Glue will probably take longer to set.
Crimp leads onto the ends of the resistor using bootlace ferrules or similar. Don't solder them - they will get too hot for solder. Insulate the join with heat-shrink sleeving.
I insulated the thermistor leads with sleeving taken from some PTFE-insulated instrumentation wire. Alternatively, insulate the thermistor leads with PTFE plumber's tape - fiddly, but doable.
Place the thermistor in its hole and secure it with Kapton tape.
Screw the nozzle onto its PTFE insulator.
Important step: Screwing the brass and PTFE together will slightly compress the 3.5 mm hole down the middle of the PTFE. Keeping the pointed end of the nozzle pointing upwards, twist a 3.5 mm drill bit by hand up the hole to enlarge it back to the correct size. Pull out any PTFE swarf created and, if you have access to an air line, blow through from the nozzle end. You are keeping the nozzle pointing upwards so that gravity is tending to clear out the swarf. It is important not to leave any lose pieces of PTFE in there as these will block the nozzle.
This shows the device in place on the extruder.
The base and the driven-gear holder of the extruder have to be assembled together before this nozzle can be attached. Put them together with two M4 bolts, two washers and two nuts.
Wrap Kapton tape round the head of the PTFE insulator until it is fat enough to fit snugly in the hole in the extruder body. This hole is deliberately over-sized to allow alternative designs to be epoxied in here.
Wrap several winds of PTFE tape round the M7 threads on the PTFE, then screw the brass nozzle tightly onto those threads and that tape.
Put two nuts on the ends of the M3 rods and lock them together. This effectively makes two very long M3 screws.
Put a washer and a 10mm aluminium plate on each one and put it through the two holes in the extruder body.
Push the Kapton-expanded end of the PTFE into the hole in the extruder.
Put the PEEK block on the nozzle and the two M3 rods. The smaller 8mm end of the central hole should be uppermost and should bear against the flange on the brass nozzle. The larger 10 mm hole in the PEEK is to reduce thermal conductivity between the brass and the PEEK, to keep everything cool.
Screw the heater block onto the nozzle, remembering that the thermistor goes at the pointy end. Make sure you don't damage the wires on the M3 threaded rods.
Put M3 washers on the M3 rods and tighten two M3 nuts onto the PEEK to hold the whole device together. Don't over tighten these - things need to be firm; that's all. Finally, put two more M3 nuts on and lock them hard against the nuts you just tightened (without tightening them any further). You need lock nuts to hold the whole thing together and to stop it vibrating lose. Nylock nuts won't do - the bottom end gets too hot for the nylon.
If you are used to running a nozzle with the thermistor attached to its surface, you will have to set temperatures higher for this design. For example, with a surface-thermistor nozzle I extrude PLA at 205oC, but with this device I use 220oC. That is because the buried thermistor measures the core temperature, which is hotter than the surface temperature. The block and nozzle combined have a bigger thermal inertia than the nozzle on its own, so this design will take a little longer to heat up. It also means that it can be heated using simple bang-bang thermostat control; there's no need for PID control.
To disable extruder heater PID control in the RepRap firmware, simply set its proportional band to 0. That is to say, in your configuration.h file, set E_TEMP_PID_BAND to 0.
Note: * For a drill chart useful for making your own nozzle visit http://whatisacnc.com/index.php?cID=241
Variations on this design
An even better alternative is to replace the PTFE with PEEK and to put a PTFE sleeve down the middle. I based this design on the one by Alain Mouette here. It needs no thread cutting. Everything is push-fit, and so it is easier to make.
As you will see from the drawing, all the holes except the 3.5mm hole down the brass nozzle are counter-bored, not drilled (that is, they have flat rather than conical bottoms). I don't know if this is strictly needed; it may work with ordinary drilled holes; if you make one like that and it works, do edit this page.
But I didn't buy expensive slot drills to make the flat holes. I drilled them with an ordinary drill, then squared off their bottoms with a drill of the same diameter with a ground-off tip:
Check your grinding with a square held up to the light - you want it at right angles. This trick only works if there is a smaller-diameter hole continuing the counter-bore which you drill first (which all on this design have). The ground drills (unlike proper slot drills) do not cut at their centres.
The PTFE sleeve came from Adtech. It is their tube PT1/8x1/4. It has an internal diameter of 3.2mm (perfect for 3mm filament) and an outside diameter of 6.4mm.
The 16 mm PEEK rod I used is this one from RS.
The brass rod I used for the nozzle was, in fact, 3/8" imperial size. But that is only 20 microns different in diameter from 9.5 mm, and is a very common size of rod.
You want an interference fit between the 9.5mm end of the nozzle and the PEEK insulator. Fortunately here the way that PEEK machines comes to your aid: it gets very hot with sliding friction against the side of the drill drilling it, and so expands. So when it cools, the hole tends to be a little under-sized. This is perfect, as it will give a good tight fit with the brass. File a slight chamfer on the end of the brass, and twist a big drill by hand in the end of the PEEK, so you are not trying to push two perfect right-angles past each other.
You also want an interference fit between the brass nozzle and the heater block. Make the block first, then, when you are making the nozzle, reduce its diameter in very fine increments towards 8mm, trying the block against the nozzle at each increment. Stop when the nozzle diameter is still just a shade bigger than the block hole. Again file a chamfer on the nozzle and twist a drill in the block to get them to go together.
Embed the heating resistor in the heating block and put the thermistor in before you assemble the device. To hold the thermistor in, it's best to run Kapton tape right round the block flat on the faces with the 8mm hole through. Then cut it out of those holes with a scalpel.
When you assemble the whole thing line everything up (not forgetting the PEEK block above the heater block) and then drift (as engineer's say) the assembly together with a soft hammer. Or an ordinary hammer with a block of wood between. Make sure you put no pressure at all on the nozzle tip. If you rest the heater block on a part-open vice with the nozzle in the gap, that works well.
PTFE sleeve nozzle
Finally the PTFE Sleeve Nozzle is currently both my simplest and also most reliable design. It has a 4mm thin-walled PTFE tube (Adtech tube STW08) as a liner.
The secret is that this liner is under tension - it is anchored at the top, and doesn't reach right to the bottom of the 4mm hole down the device. This means that the friction force exerted on it by the filament being forced down does not tend to crush and concertina the liner. It therefore lasts a long time and stays straight and smooth. If it does fail, it is also cheap and simple to replace.
The liner is longer than the hole down the device. Push it down until it hits the bottom of the 4mm hole, then back off by a few mm. Mark it with a felt-tipped pen so you can reposition it. Then split the free end sticking out the top into three with a scalpel down to the mark so the three parts can spread out like petals. Put a short length of filament down the device to hold the tube open. When you push the PEEK into the extruder body those three petals should be trapped between the outer sides of the PEEK and the inner wall of the extruder drive. This is what anchors the PTFE at the top.
You end up with PTFE almost all the way to the tip, so friction is very low. The PTFE stops about 10 mm from the tip, so lots of heat can get into the filament just before it extrudes.
See also: Extruder Nozzle Variations