Tag Archives: Replicator

Cleaning a Clogged Replicator Nozzle

Last week I came down with what must have been the flu, and all my creativity and motivation ebbed away like so much mucus.

So I haven’t updated the blog in a while. Thingiverse redesigned its website, which broke my /dev/random scraper script, and I’ve been so busy digging out from under a pile of flu-delayed work I haven’t had the opportunity to repair it yet.

I did manage to restore my dualstrusion Replicator to full operating condition, though. Here’s what it looks like when your nozzle gets clogged.


A little bit of filament will still emerge from the nozzle, but it’s anemic and thin and bunches up around the tip. You won’t get any prints out of this.

MakerBot’s replacement nozzles have been out of stock for as long as I’ve been looking for them (about six weeks), so it’s up to you to solve this problem yourself.

They do have some maintenance tips that will help you get the nozzle off.

I skipped the “unbolt the stepstruder” step. The nozzle unscrews easily with a wrench or pair of pliers, but be careful, that little guy is hot if you’re following the instructions.

Here’s what eight months of melted PLA/ABS gunk on the inside of a nozzle looks like, as the late, great, Toshiro Mifune intimidates the nozzle into functionality.


(I have a Yojimbo poster in my office. I think I’ve seen every Kurosawa movie at least once.)

The instructions say to leave the nozzle in acetone overnight, but I think that’s overkill and, in the case of a really clogged nozzle, probably ineffective. A combination of acetone baths and physical goop removal got mine working again.

So first, an hour in acetone. Find a container that the acetone won’t dissolve. A shot glass will work. Do not drink the acetone.


I tried a few little tools before I found one that really scraped the gunk out well. Needles and bits of wire didn’t do the job, but this small drill bit did. I just stuck the bit in and twisted back and forth; no electricity involved.

Then back in the acetone bath for an hour, and then the drill bit again. Lather, rinse repeat two or three times.


Replace the nozzle, load your filament, and before you can say Bob’s your uncle, you’re back in show-booty-bidness. I didn’t even have to recalibrate the build platform.


I’m working on a couple of dualstrusion models now. Watch this space.

Raftless Printing with PLA

I’m no expert at 3D printing, but through a lot of trial and error I’ve discovered a few nuggets of information that I wish I’d known a few months ago.

A few weeks ago I said I was chasing a raftless print. I’ve finally got it, at least with PLA and ABS, printing Seej bloxen.

This comparison isn’t entirely fair, since the bloxen are sourced from different models, but the difference between a hot, rafted print on the left and a cool, raftless print on the right should be pretty clear. Both of these are printed from the same roll of 1.75mm PLA.

Here’s what I’ve learned.

PLA likes a cool build platform, somewhere around 45° C: My MakerBot Replicator ships with a heated build platform, and ABS plastic seems to stick nicely around 115°.

When I first started printing with PLA I just assumed it was more or less like ABS. When prints started shifting off the platform I kept cranking up the temperature on the HBP, eventually resorting to elaborate raft structures with painters’ tape.

It never occurred to me that lowering the HBP platform temperature was the solution.

Changing the build temperature in gCode is easy: just find a line that looks like:

M109 S100 T0 (set HBP temperature)

And change S100 to S45. This assumes that you’re using ReplicatorG to generate your gCode.

There’s one other change that seems to be helping: print a nice, gooey first layer of PLA, and then back off the printing temperature to maintain the model’s integrity through the rest of the print. The gCode for this isn’t much more complicated, but you have to get it in the right place. First, set the extruder temperature at the beginning of the print, close to the top of the gCode:

M104 S240 T0 (set extruder temperature)

Change 240 (or whatever it is in your gCode) to 210.

The last step is to find the point in the gCode where the first layer stops and the second layer begins, and back off the temperature a bit.

In the other two instances there are existing M codes that tell the machine to wait for the extruder or bed to reach the specified temperature before proceeding. You’ll have to put those codes in yourself for this line.

Find the first example of:

M73 P1 (display progress)

and add these lines after the next <layer> tag:

(*custom gCode here*)
M104 S190 T1 (set extruder temperature)
M6 T1 (wait for toolhead, and HBP to reach temperature)
(*end custom gCode*)

That should drop the extruder temperature down to 190° and then resume printing. The delay can be a little unexpected the first time your Replicator just seems to stop printing abruptly, but then your forebrain will kick in and you’ll realize it’s just doing what you told it to do.

The usual warnings about bricking your Replicator or burning down your apartment building apply. Use this code at your own risk, and you should probably own couple of fire extinguishers anyways.

The Perfect is the Enemy of the Good

This happens to me all the time. I start looking for information on some technical subject and I end up wading through ancient forum posts and forgotten wikis in a fruitless search for clues. A lot of this stuff seems to be written by and for people smarter than I am. It can be very frustrating.

I usually just want to do X, where X is something straightforward like “pause the print, move the Replicator’s extruder head out of the way, wait for user input, and then resume the print.”

After a few hours of of data forensics and a fair amount of trial and error I have a gCode solution for X. Keep reading.

This is the beginning of a Seej Tournament Bloxen print.

I found the need for this script when I started printing with PLA a week or so ago. I haven’t got the temperatures quite down yet, so my rafts curl up a bit at the edges and sooner or later end up taking the whole print for a joyride around the build platform.

I’m still looking for that perfect raftless print, but along the way I’ve got to actually produce some printed models. So I’ve taken to printing a raft, pausing, and then using painters’ tape to bind the raft down to the build platform.

I’m a huge fan of having a bag of popsicle sticks around any hobby project. They’re cheap, disposable, and can be quickly modified into a variety of simple tools (gaffs, hooks, spreaders, etc) with a pocketknife. They’re also great for evenly applying painters’ tape to rafts.

It’s kludgey, but it works. The only problem I’ve had so far is getting around the print heads to get the tape down on the raft, but with this new script the Replicator moves the nozzles out of the way before pausing the print.

Fair warning: this worked on my Replicator, but there are no guarantees it will work on yours. If this code crashes your extruder head, slags your controller board, or burns your house down and torches all you hold dear and dry-humps the ashes, it’s not my fault.

That said, it’s pretty vanilla gCode and I don’t expect you’ll have many problems with it, assuming your gCode is using millimeters and absolute positioning, which I think is the default output from ReplicatorG.

The first lines you’ll need to look for are:

(<raftLayerEnd> </raftLayerEnd>)
M73 P6 (display progress)

This indicates the end of your raft’s print. The next significant line you’ll see is one that begins with G1:

It should look something like G1 X-4.8 Y-4.68 Z12.6 F3300.0

This is telling your Replicator to start extruding the first layer of your model. In between these two significant lines, you’ll have to add this code:

G91 (*set to relative positioning*)
G1 X-60 (*move the print heads -60mm in X, assuming the print is using mm*)
M71 P60 (Press button to resume print)
G1 X60 (*move the print heads back 60mm in X*)
G90 (*return to absolute positioning*)
M73 P6 (display progress)

Be careful with your X move values. I don’t see anything keeping you from accidentally ramming your extruder heads into the side of your Replicator with too large a value.

I’m sure there’s a more elegant way to write this code, but I’m never one to let perfection be the enemy of the good. It’s done, it works, and I can improve upon it later.

This code also seems to disable the Pause button once the user has pressed it, which is irritating but I can’t bring myself to spend the energy tracking that particular bugaboo down. I’d also like to find a way to get the nozzle’s current position, move the heads, and then return to that position. I’ve found tantalizing hints on just how to do that, but that’s a hack for a different day.

Socket to me.

I can’t think of many designs that aren’t improved by making them glow. LED’s are the tinker’s equivalent of Photoshop Layer Effects.

The LED socket is my attempt to play industrial designer; I wanted to make an object that was intuitive, elegant, and easy to use. So I started small.

It’s designed to hold a 5mm LED and battery without trimming the leads.

The recessed cradle for the LED is straight on one side so the user is guaranteed to get the orientation of the cathode correct. The grooves on the sides should just fit 5mm LED leads if they’re bent with two 90° angles around the bottom of the socket.

The bottom of the socket is also grooved so that the socket can stand on its base like a candle.

Instructions: Feed the leads through the holes at the top of the socket. Insert the battery through the hole at the bottom of the socket. If the bulb doesn’t light, flip the battery around.

Once the bulb is lit, bend the LED leads around the bottom of the socket and press them into the side grooves to keep them out of the way.

This uses a DL1025 or equivalent battery.

Zheng3 Penny Ballista

This nefarious device can easily launch a U.S. penny across a room when printed at 1:1 scale. It’s one of the primary engines used in a game of Seej.

This is another design inspired by The Art of the Catapult.

The nock on this model is the very tip of the throwing arm from the Zheng3 Penny Catapult, turned on its side and modified a bit so that it fits smoothly into the firing groove.

I repurposed the winding keys from the Zheng3 Penny Catapult to hold the ballista’s rubber bands in place.

To assemble:

Construction is really simple! Needlenose pliers are recommended, but not strictly necessary.

You’ll need two elastic bands to complete this ballista. Tie a single band through the bore on one of the two D-shaped locking keys. Repeat the process with the other locking key.

Feed the end of one band through the rectangular bore on the ballista’s left post. Insert the tenons of the locking key into the square mortises on the post to lock the band in place. You’ll be tempted to repeat the process with the other side, but don’t give in. It will save you precious minutes of frustration.

Unlike the Zheng3 Penny Catapult, these locking keys should be a snug fit. If some settling of the plastic during printing has made the mortises too small, trim the tenons a bit with your blade of choice.

Center the nock on the knot that connects the two rubber bands. Make sure that the penny’s cradle is facing forward.

Feed the end of the band that’s been locked down through the rectangular bore on the nock. Tie the free end of the band to the second band. Now feed the free end of the second band through the outside of the rectangular bore on the right post and loop it through the second locking key.

The nock has a wee nub on its underside. Slide this nub into the ballista’s firing groove; it will improve accuracy.

Load a penny into the nock, pull it back, and fire away. Be safe! Don’t hit the cat.

Flagrant stagecraft alert: There’s a piece of hookup wire holding the nock in firing position for dramatic flair.

You can download this model for free here. You might also be interested in the Seej Starter Set.

Flattr this

The Penny Catapult

Download the STL here!

Flattr this

Here’s a basic torsion catapult. It uses one or more elastic bands to launch a U.S. penny if you print it at 1:1 scale. You probably won’t get enough torque to be impressive with less than 3 bands.

Construction is snap-together, but it wouldn’t hurt to put a drop of super glue on the joints for durability.

Feed an elastic band through each of the bores on the side braces. Tie it through the bore on each winding key. Slip the end of the throwing arm through the rubber bands. Repeat with as many rubber bands as will fit through the holes.

Twist each key 10 times and then lock it in place by inserting the locking tenons into the square holes on the side braces. The more twists, the more launching power you’ll have. Careful, pennies can sting, especially if you take an Abe to the forehead.

You can change the payload’s arc by raising the front of the catapult.

Inspiration for this catapult came from The Art of the Catapult. You won’t find a better introduction to medieval siegecraft than this book. It’s got plans to build all kinds of stuff from catapults to trebuchets. Very kid-friendly.

If you read The Art of the Catapult you’ll learn that “catapult” is more of a catch-all term for this breed of siege engine. This model would more properly be called a mangonel.

If you’re looking to build stuff with a PG rating, Whoosh Boom Splat: The Garage Warrior’s Guide to Building Projectile Shooters is a better bet.

Flagrant stagecraft alert: the clever-eyed among you will note the use of black electrical tape to keep the throwing arm cocked for the photo.

ReplicatorG Settings:

HBP: 120° C
1.8mm Black ABS
Using Raft
Layer Height: .25
Number of shells: 1
Feedrate: 25 mm/sec
Travel Feedrate: 55