Lyman v5 Filament Extruder
Contributors: Brian Esquivel, James Mitchell, Kevin Takalo, Jesse Jenkins
This is a build and process blog about the Lyman Filament Extruder v.5. Information about the original project is documented by the inventors Hugh Lyman and Filip Mulier here on thingiverse
We need to make a machine that quickly melts and extrudes crude filament of ~1mm diameter. This crude filament will be chopped up using a system on Thingiverse.
We plan to build a “re-extruder” machine using a large 2″ steel pipe as a crucible with a piston plunger and a nozzle made from a pipe cap with a hole drilled in it. This design comes from the “Re-Extruder” by Imaginos1892
Pellet Extruder Design guidelines
to avoid problems typically encountered with rotating/auger screw machines, i.e. seen in the “GranuleExtruder” by Adrian.
-Make the lead screw able to be sped up
-Make it long enough that the pellets don’t melt too far up the channel
-Use a piston rather than a screw
-Interesting guidelines for auger-type pellet extruders may be found in this paper by Tokaya et al.
The Lyman v5 Filament Extruder
This extruder works with uniform feedstocks, like pellets, and can extrude filament with tight enough tolerances to use in a RepRap 3D printer.
Initial tests (8/2015)
We tried to extrude some powder/pulverized plastic. The plastic source was from failed 3D prints and the method of plastic pre-processing was a blender. The pulverized plastic mixture (pictured below) was introduced to the preheated and spinning extruder/auger (~12 RPM) at 185 – 215 degrees Celsius. It was clear that the extruder was spinning in the right direction because plastic was building up near the entrance to the pipe, but at no point did plastic come out of the hot end. The auger stopped spinning when the temperature was down at 185 degrees Celsius, but would spin without seizure at 210 degrees celsius.
We decided to use a more uniform feedstock (chopped up 1.75mm PLA filament) for the next test. This seemed to do the trick, within a couple minutes we had filament coming out of the end at a steady pace. We do not have the puller box assembly finished. The following needs to be done:
- Design a motor mount for the 14:1 NEMA 14 stepper motor to couple it to the puller box assembly.
- Plug in the electric cooling fan
- Find a way to easily make pellets from any plastic source
- Troubleshoot the filament width sensor (it seems to have died).
The electronics enclosure was dremeled out to fit RAMPS 1.4, a solid-state relay, a 12V – 6A power brick, LCD screen, switches, fused 120V socket etc. (Pictured below)
Marlin firmware was modified to fit this extruder according to the Lyman Filament Extruder v5 instruction manual published on thingiverse.
Filament Width Sensor
Contributors: Jesse Banks, Brian Esquivel, Jesse Jenkins
The filament width sensor for the Lyman v5 filament extruder needs to be soldered and programmed using the freescale semiconductor USBDM.
Soldering surface mount components
I am going to use the software found here for the USBDM. This is a $20 item I acquired from here. Last week Jesse Banks of Lumina Labs taught me how to solder on surface-mount components. I worked my way around the board one component at a time. The trick is to use very little solder on just one joint of each component, even multi-pin components. This method’s beginning is to tack the component in place.
Then after it is tacked into place you can add a bit of solder to each leg.
The solder just flows right into place. If you add too much solder you can remove it with some solder wic
Summary: A lot of things are not going together how we expected. This is a quasi-organized monologue about the trials and tribulations of building and assembling the Lyman v5 filament extruder to-date.
We printed most of the parts for the filament extrusion system. For a complete list please see Lyman’s v5 build manual. Beware, there are new versions of some items, such as the level-winding assembly and the filament width sensor, which you should consider printing instead.
The Lyman v5 filament extruder is broken down into subsystems according to Hugh Lyman’s manual. Every one of these items mounts onto a wooden base.
Here is what we have made so far.
Extruder: We printed the box out of ABS, getting it to print without warping or curling was difficult, but we succeeded. We did not find an auger bit that had both the hex-shaped shank Lyman specified on his motor coupler AND the proper OD and length (I will specify our exact auger dimensions later). We made the auger we needed by cutting and welding the shank of one auger onto another one of the proper OD. We used the lathe to ensure the pieces remained co-linear and TIG welded the two pieces together and then ground the weld smooth. We had to boar the seam on the inside of the iron pipe to fit the auger bit inside. We then got all the parts bolted together. Currently, the auger bit does not sit perfectly parallel with the iron pipe it is supposed to turn inside of and so it rubs against the sides and is no longer rotating by hand.
Things we need to do on the extruder:
- Resolve the auger binding and rubbing. It might be lessened either by using a flexible motor coupler (as is typically used in couplers motors to shafts) or by shimming the motors.
- Hook the extruder motor and heater up to a controller and relayed power source to see if it will extrude plastic.
Puller box: We printed and purchased the parts for this assembly. We ordered the NEMA 17 14:1 geared stepper motor that is specified in the Lyman v5 manual from Kysan electronics as suggested by Lyman. Kysan has a minimum order requirement which we emailed them about and they waived it for us. Shipping from China for this motor cost just as much as the motor itself (~$35). After receiving the motor, the bolt holes in the transmission portion do not line up with the bolt holes in the puller box.
Things we need to do on the puller box:
- Redesign the motor-mount portion of the puller box to fit the geared stepper OR design a conversion plate OR use a NEMA 17 without a transmission (which would bolt right up to the current holes). The need for a geared stepper motor for this assembly is definitely questionable.
- Find a spring to fit in the top of this assembly
- Find bearings that will fit onto this assembly
- Make shafts to mount the urethane rollers onto
- Mount and assemble the puller box to the wood base
Level winder: Lyman says this needs a lot of “dressing” in the manual to make it work. The worm gear in this assembly is the most difficult to print because it has such drastic overhangs. We have not yet got this to work.
- “Dress” the worm gear
- Cast the Lead weight
- Machine the PEEK weight-holder piece
- Mount and assemble the level winder to the wood bas
Spool winder: This is a great overall design. I don’t like that the spool required is printed. We have a lot of leftover spools from the filament we purchase. It would be great to reuse them! To use an assortment of spool sizes with this assembly would be ideal. It should be possible with minimal redesign of the assembly.
- Mount and assemble the spool-winder
Electronics box: Lyman uses a large 3D-printed box (too large for our 3D printers) to house the power supply, RAMPS electronics, solid-state relay, chocolate box, and switches. We plan to make a similar box out of folded Aluminum sheet metal.
LCD mount: We have not yet printed the parts we need for this assembly. Mostly because this item is more of a luxury item that will be taken care of after the basic systems are working.
Filament width sensor: We printed the parts for this assembly and we’re pretty confused how it is supposed to go together. We ordered and received the 3 printed circuit boards (PCBs) from OSHpark, located in Portland, OR, and they look fantastic. I will upload pictures of them in a later post. We ordered all of our surface mount components from Mouser electronics. The Mouser bill was less than $20. I plan to use our friend’s circuit-specific heat gun tool to mount the components onto the board. Jesse Banks of Lumina Labs will be helping us with this section.
We are in the process of ordering the PCBs for the filament width sensor (prototype #2). We are checking out OSHpark, a local PCB manufacturer, for prices on these prototype boards. Brian Esquivel from Art Force Studios ordered the necessary surface mount components from Mouser in December of last year. Thanks to Brian and his brother for printing all of the parts for the level-winder assembly and some parts on the spool winder assembly. We ordered the 14:1 geared NEMA 17 for the puller box assembly. We received the geared 12V DC motor for the spool winder assembly. We plan to start extruding PLA to begin with since that is what we consume the most.