General Build Questions / Build Procedures / Common Pitfalls

Discussion in 'OpenBeam Kossel Reprap and OpenBeam Kossel Pro' started by Terence Tam, Dec 24, 2014.

  1. Terence Tam

    Terence Tam Mr. OpenBeam Staff Member Vendor

    Work in progress documentation can be found here:

    Thanks to @skyshard and other early adopters, we also have some very good resources on the forum on assembly instructions:


    Assembly Order:

    This is how I'd assemble the printer:

    1) Build the ball joints first - one of the 360mm OpenBeam extrusions are used as part of the ball joint gluing fixture.

    2) Next, build the vertical linear rail drive "towers". This uses one of the upper vertex pieces as a spacer to set the crucial end-stop position, which determines retraction settings on the auto levelling probe.

    3) Take out the Linear Ball Carriage Plastics from the Chassis Upgrade Kit, and glue the aluminum stand off to the top half of the linear ball carriage. (This gives the glue plenty of time to degas). Set aside.

    4) Take out the Idler Pulley top/bottom and glue in nyloc nuts to 3 of the plastic pieces.

    5) Assemble and glue the lower end effector parts. This is all the major structural gluing for the printer. (The last item will be the exhaust air duct on the PSU).

    6) Build, in no set sequence the following: Bed clamps, Cold End Extruder, Lower Triangle, Upper Triangle, Motor + Pulley set.

    7) Install wires to Brainwave Pro. This is much easier to do without the board installed.

    8) Wire the AC inlet. MAKE SURE YOU SET THE PSU TO 110VAC for 110VAC countries!

    9) Install AC inlet plate, Brainwave Pro.

    10) Install motors and wire route motors

    11) Install linear rails, upper triangle, bolt frame together.

    12) Route end stop wires

    13) Install PSU.

    14) Install bed clamps

    15) Install HBP, glass, carbon fiber swing arms, end effector.

    [Boarding, check back in later...]
  2. KDog

    KDog Member

    Thanks, the above is really good and needed information Terence. I wanted to point out that the video on building the end effector is really quite good. You had made some comments that you didn't like it when you first posted it. It's very clear and the things you struggle with during the build really help point out what we need to be careful with. These types of videos can be so over-edited to make it seem like assembly is a breeze when it isn't. And the video quality is plenty good enough. I hope you make some more when you guys get to building the assembled KSs.
  3. protoserge

    protoserge Member

    [Save some room for printer configuration]

    Computer Software Setup... I'm not quite sure what to do. I already have Arduino IDE, the computer recognized the Brainwave board, I have Repetier Host already and I'm familiar with setting up 3D printers using Marlin firmware.


    Oddly enough, Repetier is now reporting "Connected". I'll confirm and then post a screenshot.

    OK, I went to Windows Device Manager and installed the Brainwave as an Arduino Mega 2560 (you have to go to the advanced driver install method to do this). I am able to operate the printer in Matter Control, but I don't have familiarity with settings to trust much more than just playing with some of the axes.

    NOTE: DO NOT drive X or Y axis in Matter Control unless there is ample space between the end stops and the carriages. It will bind until configuration is set properly.
    Last edited: Dec 25, 2014
  4. Dantzler

    Dantzler Member

    I wasn't quite bright enough to interpret the photo with the vertex as being a spacing spec so I winged it. My end stops are ~66mm from the end of the extrusion instead of 60mm. Is there a way to compensate for this in software instead of me tearing the top end apart and having to move the linear bearing rails, etc. ? Perhaps it can be adjusted in the firmware or using M666? This might explain why the probe tip clips the right hold down at the start of bed leveling.

    I will add a tip:
    When locking down the top extrusions of the lower triangle, flip the assembly upside down on a flat surface (I used a granite countertop) and make sure the extrusions end up lined up with the top faces of the vertexes. There is a little play possible and this will help ensure a more level bed.
    Last edited: Dec 25, 2014
  5. thoughtfix

    thoughtfix Member

    My mechanical build is complete and here are some notes:
    • Even though all the documentation is in progress, it was possible for me to make all the physical parts of this machine with the combination of Vimeo videos, Powerpoint slides, and photos posted (both official and user photos) but this is also my third printer build.
    • Oddly the most difficult part (the part that required a lot of re-doing) was the belt tensioner assembly. I found it actually easier to run the top belt through the assembly with it unmounted and estimate the belt length, then attach the tensioner as a last step. Trying to get the loop to hook up "at tension" was very difficult for me.
    • That loctite is some fierce stuff. It instantly and permanently stains any wood surface on contact. Don't build this on a dining table you intend to keep and leave uncovered.
    • On Windows 8 machines, use either the Brainwave.inf in GitHub or the Arduino Mega .inf, but DISABLE DRIVER SIGNING ENFORCEMENT. If you use Bitlocker, it may be easier to disable Bitlocker temporarily while you go to disable driver signing enforcement rather than type the massive bitlocker recovery key in when you go to "advanced startup settings."
    Now that things are mechanically "done" I tried sending some commands via RepetierHost. I tried MatterControl first, but it doesn't seem to have the granularity of control over settings that Repetier. A raw G29 auto-leveling sequence worked just fine, so I tried heating up the hot end.

    What's that? SMOKE? *power pull" After visually inspecting the end effector after yanking power upon seeing smoke, I determined that the first puff of smoke may have just been "new components" settling in and getting their first taste of heat. Nothing looked damaged or discolored. I set the extruder to 200C again and told it to warm up. It showed that it got to temperature very quickly, but the motor started skipping when I sent it filament. Since there was no resistance on the filment, I assumed that it was too hot. I lowered the effector and setup a thermometer on my multimeter held against the heat cartridge and it seems to be reporting about 50-80C under the temperature reported by Repetier. I tried raising the temperature in Repetier up over 250C (to attempt to get the probe to 200C and extrude some plastic) but at 250C it gave an error that the "temperature was reset" and I had to restart the printer.
    I don't know if this is a software, firmware, calibration, or hardware issue. I may have to wait for the custom MatterControl.

    I do have one genuine hardware problem, though. My Brainwave Pro board LATCHED ON to my USB cable. This is a regular/undamaged USB A->B cable I've been using for Arduino work for over a year but the Brainwave seems to have locked into it somehow. It is NOT letting go. I dare not pull too hard because I don't want to break my brand new board. IMG_0761.JPG
  6. AndyG

    AndyG Member

    This meshes with some behavior I had tonight printing for the first time. I did my research as I'm a 3d printing newbie and PLA is supposed to be 175 - 200 C. OK, so I set my extruder to 200C, let it preheat and try to extrude some filament through it and not getting good results. Requires a lot of pressure when done by hand and when being pushed by the stepper the knurled knob is skipping over the filament.

    Set to 220C and things start working much better and I can print. That seems way too high for PLA so I think something is definitely off. Can't get a good read on it with my IR thermometer. I'll look into my multi meter tomorrow and see if I can get temp off of it.

    EDIT: I also had some nice smoke initially. I think it was the potting compound as it definitely has a different look now than it did prior to heating.
  7. protoserge

    protoserge Member

    I haven't gotten any smoke yet, but I've only run the heater to 80 degrees C ;)

    I put the big finger trap on the bowden tube and harness wire. I like your pocket badge retractor @thoughtfix
  8. Terence Tam

    Terence Tam Mr. OpenBeam Staff Member Vendor

    Sorry, should have warned people about the smoke. I see it on half the hot ends I bring up (and I might have miss it on the other half...).

    We clean the brass nozzles in simple green in an ultrasonic bath for 30 minutes to degrease and clean it of any machining oils, but I think something still remains. The times I've seen it, the smoke comes out of the nozzle itself. But it only lasts a few seconds or so, and then it goes away.

    The potting compound does change color, and on some print heads, swell up over time. I consider it normal - the only goal of the potting compound is to prevent the thermistor from getting tugged out, which is a very catastrophic failure. If the thermistor wire breaks, the firmware registers a break and shuts down. If the thermistor is pulled out, the firmware thinks "Gee, the hot end isn't very hot" and keep applying power until the heater cartridge is cherry red and the hot end melts.

    I have not done a proper temperature check, but I do think that the thermistor reads a bit hotter than normal. On my machine for PLA I've been printing at 235/220 First layer / subsequent layers. You can bring it down to 210, but 200 is definitely too low of a temp. In our redesign, we did move the sensor head of the thermistor a bit closer to where the filament is. I do however think that there are accuracy issues with the thermistor, and on our next printer, I would like to look into using something like a PT100 or a K type thermocouple for temp sensing. (both have their drawbacks too). My guess though is that the actual truth is somewhere in the middle. @thoughtfix do you know what the measurement sensor in your multimeter is? I am guessing it's a thermistor too...

    We have the equipment to do a really good study on actual hot end tip temperature vs what the firmware reports, but I don't have the time for it, and I have a profile that works for printing, so I'm going to drop it in the priority queue.

    WRT USB cable being stuck - this is the 2nd time it happened. It happened on my personal printer too, and it happened after a few cable swaps. Mike and I are discussing our options - likely we'll just cross ship a board for replacement. I want to root cause it to see what's going on, these are brand name Molex connectors that we are using...

    -=- Terence
    thoughtfix likes this.
  9. thoughtfix

    thoughtfix Member

    Very thorough response - thanks! The multimeter is an Amprobe AM-240 and it was bundled with a "K-type thermocouple" in a metal probe. I don't know exactly where in the metal part the thermocouple sits but any good design would have it near the tip. It's rated from -50 to 600C. I also have an IR thermometer "gun" but don't trust my aim on it because I was never able to get a reading over 99C when pointing at the hot end. I think that's a problem wit my aim or range. That particular thermal "gun" is good for ovens, windows, engines, or ducts but no good for little hot ends.

    If you want to replace the board, do it when no one else is waiting for a new Brainwave. After all - it's not hurting any functionality. I can still unplug it from the Surface Pro side :D and if it is replaced I'll ship it back with the USB cable "plugged in" so you can dissect it and see what happened.

    I'll put in more time and energy into testing and investigating and report back on my results. It really is a very beautiful, stable, sturdy, and impressive machine. Once I get it working, it'll become my favorite thing I've ever built. I have a small build gallery at and those photos can be used by anyone for anything. (CC BY-SA 4.0 license)
  10. sandman

    sandman Member

    @thoughtfix Not sure if you fixe it already but your heated bed is not sitting on the bottom of the clamps in the above picture.
    thoughtfix likes this.
  11. Dr.J

    Dr.J New Member

    Quick question as a newbie to the 3D printing world and this build in particular (started yesterday). The rail slides seem to stick more than I would expect. What is the best product and/or technique to lubricate them? I thought I had seen this in a prior blog post but can't find it now. Recommendations? Thanks!
  12. KDog

    KDog Member

    There are a few comments about this on page three of "Unboxing + assembly pics!".
  13. protoserge

    protoserge Member

    The rails already come lubricated. You can clean and oil them again. They won't be perfectly smooth. You will feel the balls moving. That is normal as far as I can tell for this quality of rail.

    I put Vactra #2 way oil on them because that's what I have laying around for my mill and lathe. I then ran the carriages back and forth a bunch of times and they smoothed out a little. Vactra can be pretty stinky, but I didn't saturate the rails.
  14. Carl Quinn

    Carl Quinn New Member

    I'm about halfway through mechanical assembly, having done the ball joints, vertical towers and top triangle. Now onto effector and base triangle.
    I am finding that many of the instruction slides are missing pictures and/or text. Is it just me, or are these just not done?

    Some notes so far:
    • The slide decks are a really effective way to communicate assembly instructions. Having text plus pictures is just right to follow along at any pace.
    • The labeled ziplock bags of screws and nuts are really nice, as they make it easy to know you have the right part for a given step.
    • It isn't always clear which tiny part bag section was needed for each step, and the instruction decks don't always correspond to the subassembly boxes 1:1.
    • It would be nice to have a "begin here" slide deck that showed the overall construction sequence with pictures and a pointer to each of the detailed decks for each subassembly. I tried to follow the build order from above in the forum as well as the order of the DOC numbers, but they didn’t always align.
    • It would be nice to have a couple intro slides in each deck with: (1) the parts & tools that will be used in the subassembly, and (2) a picture or two of the completed subassembly.
    • It’s not clear that the online powerpoint is working well in my browser—maybe Google Slides would be more accessible online for everyone.
    • Specific notes:
      • 00190: Slides 3 & 4 are a great example of a subassembly intro. All slides look OK.
      • 00191: All slides look OK.
      • 00192: All slides look OK. No bearings were left over, contrary to the text. It seems like some tape or clamps around the base of the joint would keep them from splaying before the locktite is set.
      • 00193: All the slides are missing text. Slides 16-20 are missing pictures too. The small screws and nuts seems obvious, and how to measure the switch end as well. Took a few tries to get the rail in the right place. Wire connectors seem brittle when bending to tuck the wire—they might break.
      • 00194: Slides 13-24 are all missing pictures and/or text.
      • 00195: Slides 6, 15, 17, 18 are missing pictures.
      • 00196: Slides 27-37, 39-66 missing pictures,
      • 00197: Slides 2-8 missing pictures, 9-19 missing text, 20 missing both.
      • 00198: Slides 3-8 missing pictures.
      • 00199: Slides 23-31 missing text, 32-33 missing both.
  15. Dave

    Dave Member

    Carl, Yes the instructions are not complete yet, Terence mentioned that in his blog post. I found downloading the PowerPoint slides works far better than viewing them in a browser. I completed my build but it wasn't easy, I agree that having a step by step instructions would be better but in Terence's defense he also posted on the forums in what order to build the kit. If you have any questions and or hit a road block just ask, many of us have been helping each other out with our builds.

    protoserge likes this.
  16. hojalot

    hojalot Member

    I've been using the online powerpoint, but stopped in a few places where the missing pictures or text seem important to finishing the subassembly properly. I'm part-way into the end-effector, but the ExStruder seemed to have all the steps I needed. Some of the slides did not open at all for me, like the Wiring and Systems 00196, and I was prompted to open in PowerPoint which failed since I do not have MSWindows.

    Nothing has totally blocked me from making progress since there are other parts to work on.
  17. protoserge

    protoserge Member

    For some reason pressing the extrude (E+) button is the same as a Z+ operation. Any ideas?

    I know the extrude works by itself. I've put the cable onto the Y-axis motor and
    And don't bother searching. There's stuff spread all over it's easier to just ask. If there is any doubt in what you are doing, don't hesitate to ask. A lot of questions have a few minutes response time. I found the photos Terence has on OneDrive to be extremely helpful. The photos thoughtfix posted here should help too.
    Carl Quinn likes this.
  18. Dave

    Dave Member

    General Question: I want to load the filament, what is the procedure for doing this? Is their a comand to activate the filament motor to manually feed the filament to the End Effector? I am not ready quite ready to print yet since the G29 is failing at the end still but I wanted to at least get the filament loaded up.

  19. protoserge

    protoserge Member

    has a video where you see the actuation of the red lever and the orientation of the filament. What you want to do is hold that lever down and feed the filament all the way to the hot end. Alternatively, you can feed it in a bit and use the motor to extrude (E+ in MatterControl). I'm having problems with my extruder functionality, so I haven't been able to use it yet.
  20. Terence Tam

    Terence Tam Mr. OpenBeam Staff Member Vendor

    We use Superlube 51010.

    We like it because it's doesn't have nasty VOCs and is rated NSF H-1 (ok for food contact). Important as most people have their printers in close vicinity to them.

    -=- Terence

Share This Page