From Noisebridge
Jump to navigation Jump to search

The Makerbot is Currently Not Working

I ran it without the cooling fan for a half hour and due to a design flaw the extruder self destructed. After reassembly it only extrudes for 10 seconds or so before jamming up. Serious re-engineering of the extruder head will commence after make faire. In the meanwhile you could try searching in the cardboard box for a functional extruder barrel...

How to Use the Makerbot

if it breaks, we can fix it. don't be afraid of the makerbot! I've crashed it hundreds of times and only permanently broke something once.

install some software first

  • get yourself a copy of replicatorg:

  • and skeinforge <ref> this is an old version, but it's what i'm using. if you use a later version my settings files will probably not work. </ref> :

  • optionally sudo apt-get install python-psyco
  • and my skeinforge settings:

File:Good.skeinforge-2010-04-27.tgz.jpg (if you can't get raftless to work)

File:Raftless red.skeinforge-04-27-10.tgz.jpg (rename and) untar the settings directory and move it to ~/.skeinforge

  • get raftless and install in your skeinforge:

  • apply this patch (it's actually a .tgz)


ummm.. ok we done yet?

using the software

run skeinforge with:


click on "Skeinforge" in the lower left, select an STL file, for example pulley.stl and get ready to wait. the larger the object, longer the wait. you can kinda see what it's doing on the terminal. that file took 2 minutes on my netbook, and larger objects 15 minutes or more. eventually a 2d layer diagram will pop up, and then you can open the .gcode file in replicatorg.

click on the dotted kidney bean (simulate) in replicatorg to make sure everything is working. it should tell you how long the print takes, something like

fire the makerbot

  • plug in the heated build platform. yes, it's just some wires from the transformer shoved in a monitor power cord.
  • turn on the makerbot PSU, it's a switch in the back
  • plug in the makerbot usb cable
  • run replicatorg (./replicatorg)
  • please don't upload any firmware
  • the first time you run it, machine->driver->cupcake CNC
  • click on the 4 arrows symbol (control panel)
    • lay a small piece of paper on the platform
    • remove any crud from the nozzle with bent nose pliers (hidden under the EFF hat)
    • move X, Y to the center of the platform
    • jog down in Z first in 10mm increments, then 1mm increments, until there is maximum drag force on the paper
    • jog up in 0.1mm increments until the drag lessens
    • click "set zero"
    • at the bottom of the control panel there are some settings, set them in this order (motor control actually sends the target temperature command for some reason)
      • Motor Speed (PWM): 255
      • Target Temperature: 222
      • Motor Control: Stop
      • Cooling Fan: enable
    • wait until the temperature is above 200 then try a test extrusion (click Motor Control: forward)
      • if nothing happens in ten seconds, stop before you strip out the teeth on the filament
      • if you accidentally start the motor before it's hot, you can fix it by manually pushing the filament into the extruder with some pliers
  • the build platform should be plenty hot now
    • there might be an IR thermometer in a little duct tape pouch thingy
    • it should be around 120-130C for best results
  • remove ooze from the nozzle with pliers
  • file->open the .gcode file you created earlier with skeinforge
  • click 'build' (the solid kidney bean)
  • if the first layer doesn't stick at all or peels loose immediately, adjust the z axis down by turning the pulley on top of the makerbot
  • if the first layer starts getting fugly during the infill, with crap building up on the nozzle, adjust the z axis up a few steps until it looks perfectly flat with no "fuzz"
      • note to self: get some pics and video of first layer too high, first layer too low, and verify zeroing procedure
  • now since it probably screwed up somehow, hit stop and remove the grossness with the plastic scraper and/or pliers
  • it should still be zeroed, and hopefully at the correct height adjustment now, so just click 'build' again
  • if you can't get it to stick at all, the platform is probably at the wrong temperature.
    • wait five minutes for it to warm up completely
    • or maybe you need to adjust the little cpu fan to blow on the platform to cool it down to 130C (150C is too hot)
two sample pieces, side view
your part should look like this or better. I still hadn't tweaked the feedrate perfectly for colored plastic, hence the blobbiness.
  • ok now presumably you're done printing and nothing went wrong.
    • jog Z+ 20mm to give yourself some room
    • rotate the makerbot counterclockwise from above about 30 degrees and turn on the big fan for a minute
    • now just peel it off with your hand
      • be careful, the part is still soft and can deform
      • it may need a little help with a disposable utility knife or the plastic scraper to break the seal underneath the part
      • it's easier if you remove the build platform - it's held on with magnets

tweaking skeinforge

two sample pieces, top view
note the incomplete infill on the right; this was due to improper perimeter width settings. note the square teeth on the left; this is because "stretch" is turned on. this is a good thing.

there are about 300 different "craft" settings in skeinforge, and they all interrelate in confusing ways. here's a quick overview of the important ones:

  • carve->layer height: should be about 0.8*nozzle hole diameter
  • speed->flow rate: 255
  • speed->feed rate: start around 25mm/s
  • speed->bridge feed rate multiplier: can be important on parts with bridges, but has a bug that messes up other parts
  • fill->infill solidity: 0.4 is adequate and not too slow. higher for stronger parts
  • fill->infill pattern: controls whether you get hashbrowns or honeycombs
  • comb keeps it from driving through empty spaces and making stringies
  • raft prints a waffle grid thingy to keep the corners held down
  • raftless slows down the first layer and makes a short test extrusion, set this to about 10mm
  • oozebane is useless, forget about it
  • jitter is good, set it to a large value like 100
  • stretch makes corners un-rounded, but can cause data overload stuttering if "stretch from distance" is too small

you can safely ignore the rest

care and feeding

  • don't leave the makerbot unattended for hours, as it can theoretically catch on fire (those heated bed wires are iffy)
  • don't leave the extruder hot for hours. periodically flush through plastic to prevent it from oxidizing.
  • if you are about to run out of filament, cut off both ends square with diagonal flush cutters (the blue cutters for electronics) and simply hold the filament together as it goes in
  • if the filament strips out during use (extruder motor is on, temperature high enough, nothing happening) try pushing it down manually
    • if this doesn't work, try reversing the motor until the filament comes out, cut it cleanly, and reinsert
    • if it keeps stripping out, you may have to completely disassemble the extruder and remove plastic gunk from the gear teeth.
      • use a long 5/64" or 2mm allen key to set the extruder idler wheel/feed gear spacing
    • try not to let it happen in the first place. stripping can be caused by setting the first layer too low, blocking the nozzle and letting pressure build up
  • don't upgrade the firmware unless you're willing to deal with the consequences
  • keep it away from bright light
  • don't get any water on it
  • never, never ever feed it after midnight.

Usage log

please write your name, what you tried to make, and what happened below (pics or it didnt happen)

mini mendel Z leadscrew base and motor mount. This bastard took 2 hours and almost the entire remaining roll of green. It started to peel up badly at the back corner. I rescued it by melting the corner back down with a soldering iron. Fenn 13:07, 1 May 2010 (UTC)
caption text

Construction log

This describes the current progress of the building of the Makerbot. Instructions are found here, and a forum for it is found here.

Noisebridge's Makerbot kit, thoughtfully purchased by Mitch, arrived at 83C on Monday evening, July 20. Leading the assembly team is Joachim, along with Christoph, Rachel, Steve (Mr Domino), and Jeffrey (ieatlint).

It took about an hour and a half to go through and check the parts against the parts list. We think we are missing a bag of "laser cut pulleys" but it's hard to be completely sure, as there are a LOT of parts and they are not all labeled exactly as in the parts list A bag of "laser cut pulleys" was replaced by MakerBot-made pulleys, but the parts lists weren't updated. We determined to start assembly anyway.

Cupcake Electronics Assembly

Instructions Christoph soldering one of the stepper motor drivers, one of the endstop switches, and the SMD parts on the motherboard with only occasional bitching to people to STOP MOVING THE TABLE.

Motherboard is tested and able to control stepper motors. Regular firmware needs to be flashed back onto it. This may help. Ieatlint 02:55, 25 July 2009 (PDT)

Cupcake Pulley Assembly


Cupcake CNC Body Assembly

We got the first part of the body together without difficulty, but had trouble with the Z stage as one of the parts had changed from when the docs were written. The Z stage rod holding guides changed from being O-shaped to being U-shaped, and are marked "Z guides". The forums told us to use them anyway.

The next problem was with threaded rods, which had been cut to size. The ends of some would not take a nut. We solved this with judicious use of the bench grinder. There were LOTS of nut/rod jokes, and we were in a state where they were actually funny! Rachel 10:08, 21 July 2009 (PDT)

Progressed to and including the step entitled "Z-stage stepper wiring" as detailed on the body assembly instructions. The body assembly is nearly completed. Ieatlint 03:52, 21 July 2009 (PDT)

Stages assembled and motherboard and stepper driver circuits in place. Driver boards are tested. Axis stops are installed, platforms still need craft stick beam breaks installed. -- JSharp 06:45, 2 August 2009 (PDT)

Cupcake Y Stage Assembly


Cupcake X Stage Assembly


Plastruder MK3 Assembly


  • Note: Ubuntu 9.04 specific: the and the RXTXcomm.jar that come in the replicatorg download need to be replaced with some binaries from this other place that I don't remember. -Skory
  • Progressed through to end of Heater Barrel Assembly. Miloh 04:24, 21 July 2009 (PDT)
  • Progressed through the Filament Drive Assembly, and Joachim mastered the extruder electronics. Miloh 02:03, 24 July 2009 (PDT)
  • Progressed through Attach Heater barrel through step 3. Miloh 13:17, 24 July 2009 (PDT)
  • Progressed through Attach Heater barrel to the end. Miloh 20:46, 26 July 2009 (PDT)
Continue at Plastruder burn in test.  The burn in test can't be completed until the Thermistor and extruder board are recognized by the control system.
  • Plastruder board burnt out a few components on power up test. Ordered new plastruder driver from makerbot, should arrive at 83C soon. -- JSharp 06:43, 2 August 2009 (PDT)
    • New plastruder control board received and installed. -- JSharp 10:52, 9 August 2009 (PDT)

Status Reports


Martin - The makerbot is working and printing objects well. You will need to download and install "ReplicatorG" and "skeinforge" in order to use the bot. The software runs on Mac, Linux, and PC. (skeinforge requires Python, ReplicatorG requires Java) Example objects are next to the bot.

The X/Y/Z stage endpoints are not yet installed. It requires a number of popsicle sticks to be cut to the correct length and the stage to be calibrated. You can operate the MakerBot with ***great*** care without them.

The extruder head has been upgraded to the Mark-4, with a new bearing and idler wheel. Also, the extruder head is now insulated with the same material they use to insulate the Space Shuttle heat tiles from the skin of the shuttle itself. *** IT IS EXPENSIVE. PLEASE DON'T TAKE APART! *** I can't replace the material, it was given as a gift from a friend at JPL.

Please read all the instructions on how to use the MakerBot on the [Wiki Site] before attempting to use it.


Thanks to excellent efforts and support from Martin, the MakerBot is now in a semi-functional state. There are three things that need to be worked on, however:

  1. The Y stage endpoint is installed incorrectly (the Y stage and physically crash into it)
  2. The nichrome wire needs to be better insulated so it can maintain higher temperatures and consistent temperatures.
  3. The feeding mechanism needs to be reviewed, as it appears that it fails to operate reliably.

Please do NOT attempt to disassemble the plastruder header. Also, do NOT attempt to operate the MakerBot unless you really know what you're doing.


Much has happened to the Makerbot since last noted, the broken idler wheel was replaced and the extruder head assembly was upgraded by me using parts from Bre when he was here. Thanks Bre! I also at that time upgraded the firmware and printed a bunch of skull throwies, yay! Time has passed between then and now and many people have been reporting the makerbot in sickly condition, after much testing I am glad to report this not the case. go go makerbot!


Bre was here on 20/2 and installed a heated build platform on the Makerbot. The heated platform is a new development, still in prototype stage. It is awesome because it prevents warping, and allows workpieces to stick to the build platform without having to build a raft.

He left a sheet of instructions, they say:

"... While here we set you up with a heated build platform. It's hooked up to the fans pot. So before you print, go into the control panel and click [x] fan and it will turn it on. Let it warm up for at least 5 minutes and then print on it. When the model is done, let it cool for 5 minutes and it will 'pop' right off the warm kapton tape.

We are also leaving you a spare parts kit for the extruder. Your nozzle is super thick and slow. This new nozzle has a smaller nozzle hole. Make it, use it, love it!

Bre and Zack, Makerbot"

When asked about the workings of the heated build platform, he said:

"Yup, ours is a prototype part, it's awesome but not very well documented.

You want to plug it in above the mosfet that usually runs the fan.

You can then go into the control panel and just turn the fan on, let it heat up for 8-10 minutes and it will get to about 110F.

Print away and remember to turn it off and unplug the bot when you're done!"

Please note that he's left us the parts for an improved extruder - up to us to build it - and some of the new colored ABS plastic.


After using the makerbot all day I (fenn) noticed the Z axis was consistently jamming up when traversing above 125mm/min. Several hours of head-scratching later I discovered that the heated platform would cause the large acrylic plate that holds the extruder to heat up and expand, pushing outwards on the Z-axis nuts and causing the mechanism to bind ever so slightly. The simple fix of filing the plate down was complicated by the lack of small files and the fact that the machine was assembled wrong, requiring removal of the four small u-shaped plywood "guide plates", which I put back in the correct orientation: gap outwards.

I'm not sure the heated build platform actually ever gets hot enough for ABS to stick to the kapton tape. It seems to max out at 95 celsius, as measured with an infrared thermometer. The RepRap wiki suggests 120-220 degrees for ABS. Perhaps connect the platform directly to a higher voltage power supply?

Lacking a hot enough bed for the first layer to stick to, blue masking tape sorta almost works (the first time you use it), and I got much better results when I taped a business card on top and printed on that instead. I printed a few gears and hope to add a "wade extruder" soonish.


I (fenn) came in today and it appeared that nobody had touched the makerbot since I last left it. But the nozzle wouldn't heat up. Voltage to the wires, it just didnt get hot. So, heeding the warning to not disassemble the nozzle, I removed it and swapped in another one, which works a lot better anyway. Then I attached a 16V transformer to the heated bed, and now it gets up to 165C which is much better than the old 90C running on 12V. Now the first layer actually sticks! Z axis seems to still be jamming up; I didn't file the notches deep enough I guess. Wish I had more time to play with it. Skeinforge settings that seem to work with ABS: layer height: 0.4mm, flow rate: 255, feed rate: 25mm/s.


added a fan to cool off the acrylic plate. make sure you enable the fan before printing and it shouldn't jam up anymore. seems to be working reliably now.

hint: instead of trying to adjust the Z zero in software, manually adjust the height as it prints the first layer.

caution: plugging in or unplugging the heater transformer may cause the power strip breaker to trip and reset the makerbot.


leveled bed by shaving down tabs that were sticking up, pushing magnets back in, and adding a small strip of kapton tape. adjusted toothed gear/idler wheel spacing with 5/64 allen key.

Some of the magnets on the Y stage are in sideways (poles on left & right), not upright (poles on top & bottom). I found this when building my Makerbot, since I wanted the build platform to be interoperable, I tried placing the magnets on the Y stage of the NB 'bot to check polarity, and found that they were sticking corner-to-corner, not face-to-face. - Gian Pablo