Thingomatic 3d

Revision as of 22:32, 30 April 2015 by Yair (talk | contribs) (Yair moved page 3d printers to Thingomatic 3d: its devoted to the thingomatic, we have the prusa_i3 now)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

We have a working Orca, and two Cupcakes and a Reprap Mendel that are not functional. This page will track the current status. Makerbot Thing O Matic MkVI that might be able to print small things at 400 microns, so some good calibration should do it good, and an iprint evolution which is in some unusable state.

Quick links

  • drop

Reprap: M92 X40 Y40 Z3200.00 E200 poweruser gcode


2x 7mm spanners

springs for the extruder as per wade extruder design.

  • gcode reference - TODO


Software links

looks below in log, recommended software for the original cupcake is replicatorg28 ( it includes skinforge35 in there as well.

  • Printrun (aka pronterface) (no s3g support) possibly can print to cupcake, but not to gen3 electronics (we have gen 3 on both cupcakes and gen 5 6 on the reprap - see serial 0070-003 on board, identical to pic here) WINDOWS PRONTERFACE ETC
  • ReplicatorG The original and recommended host software for controlling Gen3 Makerbots (like our Cupcake). We succesfully used it for stored s3g files we got with the cupcake, but not yet well made prints from stl.
  • Pleasent3d - (mac only, source is objective-c) for visualizing/tweaking stl and g-code. has a not-yet-working print mode
  • Slic3r stl to g-code converter. Produces G1 with E (that makes our Gen3 v3.0 firmware whine)

Some misc links possibly important later on:

  • flatline on changing thermistor table in firmware for 1mm thermistors.


This is Moty's baby. Currently has a head for 1.7mm filament. We're trying to piggyback on another large order of filament. Possible improvements - all-steel extruder head, plastruder.

Cupcake (this applies to both - I think they are clones)

  • Motherboard is v1.2 (aka Gen3 electronics, based on a sanguino)
    • firmware v3.0 (source link needed)
  • Extruder Controller v2.2 (dc motor)
    • firmware v3.0 (source link needed)
  • They both talk s3g (the boards are Sanguinos), not straight g-code, making them harder to use with newer software (but hopefully easy to fix using pys3g)
  • Cupcake Extruder MK 4

Printing: using ReplicatorG 034 following instructions, and printing from sd card


Howto Skeinforge for cupcake

Log for cupcake (newest topmost)

Cupcake lives.jpg Cupcake lives2.jpg


rebooting efforts

sun 25.11.12

replicatorg28 ( you can find skinforge35 in there as well.

Mon 24.10.12

turning down the speed to 30mm/s seemed to allow a decent test print of the rectangle test thingy. I then tried to do a 'pi' shape for Hanan but things went haywire just as soon as i turned away for a second...

Mon 22.10.12

The thing seems to be losing x-stepper motor counts somehow, each pass gets displaced/offset leading to a consecutive accumulative disaster. some reading hints the xmotor driver should be beefed up, as a alternative i tried slowing the feed rate at 30 mm/s it is still too fast.

Yair and J-ball supreme put up replicatorg28 and skeinforge 35. You choose 'Gcode->edit base profiles->SF35cupcake ABP' from the menu. Then there's a bunch of stuff you can control concerning the build. Before using that combo of replicator ad skeinforge versions the thing lets out a bleating whine like a stuck pig and theres no stopping it short of pulling the plug and usb both.

Settings in the skeinforge window-


The 'flow rate' is the pwm settig (0-255) for the extruder . We have to keep within a small range of 255 since too low values ad the thing wont print.

The'feed rate' is the speed of the table. These


The 'raft' is a base layer. This has its ow feed and flow rates which are multiples of those set i the 'speed' settings.

Thursday 28.6.2012

Moti did a lot more calibration of the hardware (dismantle, rebuild), printed one of the RepG28 examples->self-improvements (spring holder of Z axis).

Important: turn off fillet and clip in skeinforge (still using 35), those two try to move slower causing overflow due to our dc (relatively-)fixed flow rate extruder.

Thursday 21.6.2012
Wednesday 20.6.2012
  • Some false start with installing a g-code capable firmware, teacup. Went back to v2.4 MB v2.6 EC with RepG 28 and skeinforge 35.
  • Started from Cupcake GEN3 ABP profile (automatic build platform)
  • A lot of trial and error with skeinforge.
  • We managed to have a nice print - flat raft and X filled box (20mm calibration). Not sure if the X is what was designed or something wrongly set in skeinforge profile. The parameters cannot be attached to the wiki since they are in zip file, I'll try to convert them to something at some point.
  • Trying to print knot3d failed: raft is good, afterwards material comes out too quickly - flow rate is too fast for feed (movement) rate. and/or need more distance between lines. and/or need more distance between layers (higher nozzle).
Tuesday 19.6 2012
  • tried slic3r produced g-code. Saw the E but tried anyway. Got the familiar whine.
  • looking for new firmware for s3g, or alternatively fixing slic3r to produce something the firmware will take (but need to find firmware sources for that).
Sunday 17.6.2012
  • kubuntu: Installed oracle java from for replicatorg on ubuntu.
  • s3g - not working on the files produced with replicatorg (at least the examples/ but promising
  • created support request for long whine:
  • Found the problem code: G1 X-14.67 Y12.69 Z0.47 F371.901 E5.099
    • Specifically the E5.099 - extrution length - is not compatible with our firmware (v3.0 on motherboard, v3.0 on tool).
  • We tried printing in two methods to remove the E:
    • sed it out, using gcode produced with skeinforge 47 from replicatorg34. Produced too much material. (evidence: layers are too thick, new layer bumps into top most layer, test cube comes out abominationfied).
    • produce gcode with skeinforge 35 on replicatorg29, then copy it over to replication34, and print. similar results. (the gcode is slightly different).
    • produce gcode with skeinforge 35 on replicator34 (different gcode preamble, should be same gcode layers? didn't compare, lacking good gcode-diff, also lacking s3g_to_gcode). The produced code doesn't contain the E, so it works, but the extruder doesn't work. Moty thinks the gcode produces is for a stepper driven extruder, and not a dc (pwm) driver one.
    • we have a irc chanel with people that will try to help
Friday/Saturday 16.6.2012

Printing works. After:

  • Burning two stepper driver boards. They are on the secondary cupcake, the X and Z (written on the cupcake)
  • one cable (z cable) end broke, taken from secondary cupcake.
  • changing the pots on all steppers to be a little above the minimum.
    • minimum worked but caused slippage and stopping after a single layer when printing the penny_op.s3g from the sd card.
  • problems remaining:

AlonLevy 15:49, 20 May 2012 (PDT)

  • replaced ribbon for Z.
  • put back original power supply after discovering 12V touch/no touch problem (didn't mark faulty power supply..)
  • upgraded extruder firmware to v3.0
    • connect USB2TTL cable directly to extruder board
    • use replicatorG
    • instructions (for pastruder but apply just as well for our v2.2 board)
  • extruder problems:
    • started with white, think it's ABS, switch to black (sure it's PLA, much lower 'melting' temperature).
    • some black cable stuck in the middle.
    • need to move the support (called dino), open it up to remove piece.


"If you’re using a MK5 or DC motor extruder {this is what the cupcake has...}, I would recommend calibrating Skeinforge as you hold the DC motor speed at 255 and the maximum reliable flowrate you are able to obtain. You’re going to want to adjust the other variables that are easier to control – the feedrate4 , the layer thickness5 , and thread width6 . Since I like choosing my resolution and there’s not much choice in flowrate, I like to vary the feedrate since it’s easier"


We have a reprap v2.0 (Mendel) with Gen 6 electronics (see serial 0070-003 on board)

RepRap extruder lessons

Cleaning the extruder

  • cut the filament short
  • dismantle the hotend with the stuck filament piece
  • heat the hotend while making sure you do not
  • settings


current problems

  • extruder

small gear slips one revolution when under pressure - small extruder gear is broken will alow slipping under high rpm

thermal element missconfigured, will take alot of time missing 7 degrees to reach target temperature

  • bed

One of the 4 bearings under the bed on the x-carriage isn't making contact.

Log for reprap(newest topmost)

Saturday 12.10.2013
1 (f)

Got an extra hour to work on the Reprap, so I decided to do some test:\ Z speed set to 100mm/min. 25mm move:15.63s, 15.31s X speed set to 1500mm/min. 150mm move: 6.45s, 6.43s Y speed set to 1500mm/min. 150mm move: 6.38s, 6.34s These experiments indicate that the printer is infact responding to g-code movement instructions with a mm/minute base.

It's been discovered that the maximum feedrates do in fact work but are functions of mm/sec and not mm/min For Z to have a maximum feedrate of 100mm/sec and for X and Y to have a maximum feedrate of 3000mm/min the following gcode command can be used M203 X50 Y50 Z1.700 E10000 This should be added to the g code appended before the printing in the slic3r.

Handy G code tricks: Query current position M114

Saturday 12.10.2013 (f)

Last Thursday I put a bunch of time in trying to get the machine running. (1) it seems that wiki logs from the last few months are lost (2) I destroyed the hot end. What a post mortem has shown is that the z-axis isn't moving. The feedrate for z is too high. a manual g-code advance shows that 100mm/min is the fastest that the hardware can cope with in z. the g-code from slic3r wants to run way higher than that (like 4500mm/min). I may need to write a g-code modifier script to force the z-advance to a lower amount for all of the z-advance movements.

Sat 12.1.2013

Z axis fixed by tightening bolts at bottoms of z screws, and adding lockwashers. The ends of the z axes now are actually lower than the machine feet but this is ok since the machine rests on supports. Moti dl'ed reprap software 'printrun' for windows, gui is called 'pronterface' here, testcube still not printed.

One of the 4 bearings under the bed on the x-carriage isn't making contact.

Thurs 10.1.2013

The Z axis is screwy , software canbe run from printrun directory thus:

sudo ./

Wed 9.1.2013

If you can't make it perfect make it adjustable.

Adjusting the Z-axis.

X-axis - wheel was loose on motor, used a M3 set screw that Jeremy got yesterday.

Tue 8.1.13

Niv helping setup the reprap extruder.

printrun: use 250000 baud rate, nothing else works.

extruder: one problem was the notching was not centered. deconstructed, centered, reconstructed.

replaced extruder back to Niv's because of Sefi's not matching out X carriage.

  • remove dust from print surface
  • next time put nylon on anything mechanical that will be left alone for a long time.
  • check plate heating element (time to settle and settling temperature).
  • tighten loose screws in extruder.
  • mcmaster - spend 10s of hours looking at catalog.
  • label the plastics

7.1.12 Jeremy managed to get the gears etc. working correctly, this required a wierd kluge of boring a bolt into a bolt. Niv and Alon apparently got the thing extruding in linux, now I will try getting it going under windoze using repsnapper or the like.

28.10.12 Alon and J tried assembling the extruder. The notches on the hobbed bolt are in the wrong place unless you turn the bolt the wrong way round in which case it will not be turned properly by the captive bolt mechanism. The only way we can figure out how to fix it is to thread into the head of the bolt...long story. anyway hopeuly we can klgue our way into something that works. Amazing how things get out of control with this open source hardware approach - no one seems to use a version control system with dependency list, which I would think makes perfect sense for this stuff.

We have a new variant of "wade's extruder" from Sefi for the reprap (which is an 'orignal mendel'). The firmware got changed so to work with it Alon is getting some info from Sefi .

The plastic stuff seems to slip out of the extruder feeder if its too loose or be crushed if its too tight. Tom

tom sofer

'hobbed' his which makes it grab the feed, thats prob what we need to do - I asked him how to do it but by the time I formulated the question he was already doing something in the 11th dimension to a manifold.

hob this

This link has some pics of someone's problem bolt and this is what its supposed to look like.