Monday 30 March 2009

Extruder thoughts

I've been extruding all weekend so I thought I'd share my thoughts.
NB I'm not quite as organised as Nophead, so no pretty graphs and figures yet.
Friction and barrel size seems to have a big effect on the force. I built two heater barrels to the same design:

One had 3.2mm PEEK hole and 3mm welding nozzle tip. Filament *just* fits into the 3mm heater section (requires a little force).

Heating to 240C allows good extrusion with hand powered force.

One to the same design had 3.5mm PEEK hole, with a short 3.2mm section near the barrel. Nozzle was drilled to 3.2mm and filament fits easily.

Heating to 220C allows good fast extrusion with less force.

I suspect that the half-melted filament expands before melting fully and presses against the sides of the barrel, causing extra friction. I suspect that's why nophead's reaming made so much difference - a little extra space allows the filament to expand as it heats.

I also think my BfB extruder probably does work if the screw thread is adjusted correctly. Nophead has reached >12Kg with the screw extruder, so mine slipping is probably due to the spring/acrylic flexing and allowing the ABS to slide. I'll see if I can tighten up the pressure (maybe use some steel reinforcement) to drive the thread properly. Then I'll have to rebuild the heater barrel.

Three heater barrels and a new drive

I made some extruder barrels over the weekend:
(click on the links for more pictures, design and test results)

From renoir1_extruder_quick_v0.01

quick and dirty extruder

From renoir2_extruder_v0.01

Version 2 - Welding tip

From renoir3_extruder

version 3 - wider feed channel (3.5mm through PEEK) for the filament

And I mucked around with a minimalist drive (Keep it Simple)

From Renoir1_extruder_drive

Simple - shove a screwdriver in a hole
This worked OK - as long as it was driven at a slower speed than the screwdriver default (2.4v instead of 3.6v)
I also had problems driving the screwdriver motor from the RepRap electronics

From Renoir2_extruder_drive

Version 2
used the BfB motor. Plenty of torque, but I can't drive it slow enough!
Below about 50/255 the motor stalls: above, the speed and torque is good but it stalls the filament once it hits the end and chews up the filament.
It works well (filament speed is aprox 5mm/sec at the 3mm diameter end) and as much drive as the filament can stand. The star screwdriver bits work well and seem to drive well. Perhaps I'll have to get a stepper motor and driver circuit for the extruder to get the speed down?

I'm going to try and reconnect the original BfB screw thread drive to my version3 heater barrel, and see if that drive will screw down at a slower rate.
*embarassed* I can't exactly use the original BfB heater barrel anymore- I was drilling some out stuck filament when the drill snached and ripped off the ends of the nichrome wire and threw the thermistor circuit across the garage. I still haven't found it...

Sunday 29 March 2009

Extruder progress

I spent the weekend building extruder nozzles and drives and finally managed to build one that might work.

I'll write up the details of the working/not working designs over the next few days.

Wednesday 25 March 2009

Chocolate Extruder

Although my RepRap isn't quite there yet ( the ABS extruder isn't quite working), I have experimented by building a chocolate extruder from copper pipe fittings and a drill:
From choc_extruder_v0.01

A copper tube and bowl hold melted chocolate which is fed to a drilled-out welding nozzle via an auger made from a drill. The heater is a standard nichrome wire and fire cement, along with a 100k thermistor, connected to the standard RepRap arduino electronics.

Drive for the auger is provided from a Tesco Value 3.6v rechargeable screwdriver. Currently, this is manually operated as I'm having a problem driving the motor using the PWM DC driver.

From choc_extruder_v0.01

The heater took a while to heat up, but the entire apparatus reached target at 60 degrees C. Shoving a few chunks of chocolate in, it melted very quickly.

Extrusion worked much better than I expected: when the motor was off, there was a very slow drip of chocolate through the 0.6mm welding tip (approx 1 drip every 5 seconds).
With the motor on, I was able to extrude a steady stream at several mm/sec and drew some lines on the plate. With the motor in reverse, chocolate was actually sucked back up the nozzle!

From choc_extruder_v0.01

All in all, I'm pretty pleased with the experiment so far. I've written up some construction details at chocolate extruder. A couple of photos are on my picasa album.

Still to do :

  • Drive the motor from the RepRap electronics DC motor driver board

  • Design/build a mount (probably when I get my reprap working)

  • Print a chocolate object!

As soon as I've printed something, I'll post some pics. Might be a little while as I'm trying to get an ABS extruder working first...

Monday 23 March 2009

New parts arrived

I ordered some bits'n'pieces for extruder experimentation and modification. I grabbed some 8mm PTFE, 10mmPTFE and 8mmPEEK rod from DirectPlasticsOnline , and I ordered some wirewound resistors from Farnell, along with some thermistors and a set of 10-way pins and headers so that I can unplug and change extruders easily.

I spent some time in the garage cutting and drilling to make some simple insulators - and found that it's hard to drill vertical holes without a lathe or pillar drill. I also collected together some scavenged parts - some large washers, plumbing bits and drilled out a few 0.6mm welding tips to 3mm.

Sunday afternoon I spent soldering on the 10-way headers, and wiring up a couple of extruders. I'll test them out soon.

Sunday 22 March 2009

BitsFromBytes slow extruder

Detroit RepRap Builders


I have a similar issue with the BfB extruder, very slow extrusion. Checking out the alternatives, the heater and nozzle seem to be similar to other designs, and with additional force it seems to extrude OK. The obvious difference seems to be the drive mechanism.

There's two main possibilites : the screw drive and the heater resistance.
My guess is the screw thread can't generate enough force to push the extrusion through and slips. Once it slips, it chews up the filament and can't get going again.

Give that Nophead's produced some useful figures for the extruder nozzles and forces produced by the pinch wheel design, I'll see if I can measure the output and identify the problem side.

Monday 16 March 2009

Quick and dirty extruder design

Combining several recent ideas from the forums, I've put together a quick and dirty extruder design. It uses many of the labour-saving tips (welding tips, the BfB 'big washer' design, Nophead's excellent resistor work) provided by others.

The PTFE runs right into a copper pipe end stop, which has an M5 drilled and tapped hole to allow a welding tip to be screwed in. The welding tip is *carefully* drilled out to provide minimal resistance to the extrusion.

Heat is provided by 2x 12Ohm resistors (wirewound, 6W, max 350degreesC) wired in parallel (which I reckon should give me a 6ohm load). Two resistors - one for either side of the nozzle - and they're squashed between the copper pipe end, nozzle, and a big washer on the bottom.

The big washer provides support to the PTFE barrel and is attached to the motor through long bolts.

I'm now soliciting any comments or advice before having a go at building this.

Possible problems:
  • ABS melt leaking along the side of the PTFE join?
  • Not enough thermal inertia to keep the melt hot?
Possible improvements:
  • Pack the gap around the resistors with FireCement to improve heat conductivity?
  • Use silicone 'washers' to help seal the joint between PTFE and copper pipe?
Any comments or suggestions welcome.

Monday 9 March 2009

My Reprap

Added some static pages describing My RepRap and my current setup. I'll post updates and built objects on the blog too, but it's sometimes easier to have a static list. I've added some pages for planned/built objects and ones found on the web.

Sunday 8 March 2009

My 'Workshop'

On the RepRap forums, there are several users who have access to some pretty good workshops, either using college or school facilities, even those with lathes and CNC machines at home.
I'm not one of those users.
Here's a picture of my workshop:

It's my dining table, covered with an old curtain. I have a box of tools containing the following items:
  • several small screwdrivers
  • allen keys
  • some small pliers, needle nose and flat
  • two small wirecutters
  • soldering iron, stand, helping hands
  • some small needle files
I've also got a garage full of junk, but it's also got a black'n'decker workmate, and some standard DIY tools,including a small hacksaw and electric drill.

That's about all you need to build a RepRap from a BitsFromBytes kit. Obviously, if you have a full-size workshop, things are much easier - but my RepRap sits behind the TV most of the time. Building from scratch or from a RepStrap would probably be a bit messier and needs a bit more time and equipment, but building from a kit can be done in your living room.

It helps a little if you can dedicate some space for a permanent setup (a boxroom or office would do nicely) but my RepRap is sturdy enough to get it out when I'm experimenting and put it away between times.

Saturday 7 March 2009

Choi Hawarang Do

I've just finished a session of Choi Hawarang Do at the Martial Arts Leisure UK. It's a martial art, based on effective self-defence and personal development - a really positive direction.
Inital impressions:
Everyone was really positive and friendly. Right from when I first walked in, Jenny , the assistant instructor, was welcoming and helpful, and everyone was willing to help along a new person, explaining what was going on and making sure I was OK. I was surprised and pleased at the mix of ages and people there - there was a good selection of older and younger students, right down to age 10 or so, and several ladies. There was none of the 'macho' atmosphere that is sometimes associated with martial arts and it all felt warm and friendly.

After the standard line-up and introduction, we did some guided stretching as a warm-up, and then onto the lesson. I was quickly paired off with one of the senior grades to show me the basics of the stance, and a basic block, punch and kick, and to run through the first block/punch/kick sequence. Once I'd shown that I was no ninja by wobbling through the simple kicks and punches, we rejoined the main class for a 2-on-1 exercise - where you simulate two 'attackers' (holding pads as targets) against one defender. After a few minutes, there was a quick break for a drink of water and then we ran through several self-defence techniques, including getting out of a headlock, a choke hold and a bear hug. A final line-up ended the lesson.

It made me breathe pretty hard at a couple of points - perhaps I shouldn't have spent the last couple of years on the sofa watching TV. That's good, because I needed a bit of exercise.

I'll definitely be back next week.

Friday 6 March 2009

Configuration of XYZ axes

I spent some time writing a long blog post about some testing and configuration, and then realised it was almost a set of instructions. I re-wrote a few lines and posted it on the RepRap Builder's Wiki, in case it was useful for anyone else. You can find it here:
and a couple of pages here:

Basically, I've configured the X, Y and Z axis, so that if you tell them to move 10mm they move 10mm.

I did find that the x-axis stepper was slipping, as the x-axis wasn't consistent. I disassembled the stepper and took off the grub screw. I filed a larger flat on the shaft of the stepper, and re-assembled, making sure the grub was tight on the flat. It moves much better now.

I have just started to test the extruder... more soon.

Tuesday 3 March 2009

It's alive! It's alive! The beast is alive!

I finally got round to applying the electrodes and waking the beast from it's slumber (I plugged in my RepRap).
It's been stuck behind the TV for a while until I had a free evening to start configuring and testing everything. I spent an hour sorting out an Arduino communication issue (I'd foolishly uploaded the firmware compiled in test mode, and it kept saying 'start' down the comms channel).

Using ReplicatorG, I was able to jog the x-axis a little back and then forth - which worked!
I then wiggled the y-axis a bit - and that worked too! Then Z - both up and down worked!

With my hand on the PSU power switch, I gingerly pressed the 'home' button for the x-axis. The x-axis moved back to the home position, triggering the endstop and stopping on it's own! Then the same for Y - and Z - Things are going really well!

I spend a few more minutes winding the x, y and z back and forth. All was well until a z-home caused the wobbly z-flag to hit the side of the endstop, meaning the bed tried to ascend to the attic, winding itself into the print head on the way. The z-axis on the motor slipped, meaning that the bed was badly skewed before I got to the power switch. The z-belt also fell off.

I put it back together again - I don't think it's damaged but it gave me a scare.

Note to self: Hand on panic button ( the PSU power switch) until I'm really confident it's working.

Monday 2 March 2009

Java! Reprap Host software!

I spent most of the weekend getting the official host software working. The 'host' software is written in java, and uses several external libraries. After lots of fiddling, I got it working. The biggest problem was getting the RXTX libraries - I had set the LD_LIBRARY_PATH but I'd misspelled the directory name. The java software compiled OK, but didn't run as the native .so files were missing.

Once I'd got that running, I was getting preference file errors - all sorts of missing settings. Eventually tracked that down to a reprap.preferences file in ~/.reprap that came from a previous attempt early last year - well out of date. I replaced that with the latest reprap.preferences.dist and all was well.

The latest software looks much better - a better UI, and easier to understand controls. Skeinforge looks great for complete end-to-end prints, for dumping 3d models to a configured system, but the host software has easier options for configuration and testing. I've still got a bit of tweaking the options to match my setup.