Here is a little robot the kids and I were working on.
We designed it in Solidworks.
Prepared for printing in the MakerWare software.
Printed it on the MakerBot. It failed halfway through and we had to re-print some of the parts.
Here is a closeup of the print.
And some pictures of the finished chassis.
We used custom rubber bands for tires. Threads were even printed into some of the parts so 6mm bolts could screw right in.
Here is a closeup of a 3D printed chess piece. We designed them in solidworks and printed them using a MakerBox Replicator 2 with PLA plastic. It’s on the chess board dining room table.
Here is the chess “board”
Here are the ceramic tiles going into the chess table before the bar top was applied:
If you’ve ever printed with a FDM style 3D printer such as MakerBox, I’m sure you know that the build platform has to be pretty flat and level for you to have a successful build.
The first layer is critical. If you do not get the first layer right, things go bad. I wrote about this a while ago here: https://blog.gahooa.com/2013/08/10/makerbot-replicator-2-tip-first-layer-just-right/
I was having continuous problems with the first layer being too close in the center and too far on the edges of the build plate. It was like a lose/lose on a bigger build. On smaller parts I could calibrate it so it would work fine, but larger parts inevitably ended up being wrong either in the middle or the sides.
Enough is enough. A Mitutoyo Digital Indicator, a Starrett Indicator Holder, a Precision Granite Surface Plate, and time to figure this out.
While there are more conventional ways to do this, I did not have the right equipment. So I took a steel bar and flattened it on a granite surface plate with sandpaper (60 – 2000 grit). It ended up with a very flat mirror finish on the bottom.
I did not take the time to get all of the pits out of the steel, but here is a closeup of the steel after sanding:
Next I measured out a grid on the Acrylic Build Plate. 
Using the digital indicator mounted to the steel block, I was able to take height measurements on the grid.
Here is a wider view, including my high tech data collection process.
Note: I did this with both the Acrylic OEM plate and a Glass Aftermarket Plate.
Once I had the data points collected, I made an exaggerated model in Sketchup, and plotted the points in 3D.
The one that is wildly out of specs is the Acrylic, and the one that has a slight dip in the middle is the Glass.
Both plates were fairly warped in my opinion. The glass was a lot more workable. I eventually bought a CNC machined aluminum plate, but did not take the time to measure it in this same way.
I feel that the manufacturers of low end 3D printers are not paying enough attention to the stiffness of their 3-Axis mechanisms. While a FDM printer does not (should not) experience side or vertical loads during disposition, they still need to be fairly immune to vibration and even the weight of the build plate.
This was a MakerBot Replicator II – an obvious improvement over the MakerBot Replicator, and I’m sure they will continue to improve this aspect of 3D printing.
Here are pictures of a few parts we printed on the MakerBot Replicator 2.
Talking about the MakerBot Replicator 2 3D Printer
Last post I talked about what happens when the nozzle is too far away from the build plate when you are starting a build. This post will discuss what a proper distance looks like.
Remember proper setup:
The print in this discussion was standard quality. I slow it down a little bit to 70mm/sec during printing. I find the MakerBot doesn’t shake so much at the expense of slightly longer print time. I also heat it up a little to 235C which seems to be helpful for me.
Here is what we are building (at a whopping 110mm x 110mm x 100mm):
This is a good sign that the print is getting off to a good start. See how each outline is nicely pressed flat, but no scraping is apparent?
This is the front-left corner, the first printed. Notice how with few exceptions, each line is flattened just enough to merge into the adjacent line?
This is the front-right corner. The build plate must be too low here (not level) or the machine is out of alignment. You can see the nozzle isn’t close enough. What this might result in is a slight lift on a large part, but it is unlikely it will break away and cause problems.
This is the beginning of the second layer, and the time I left for 6 hours. I was comfortable that little would go wrong with the print at this point. Layers adhering correctly, build adhering correctly.
Here is a side-view of the print in progress. This is 0.20mm layer resolution, and it looks fine.
Here is a top-view of the same.
Here is the finished build:
This is stuff I have figured out by trial and error, so I hope these pictures help!
Get a good putty knife, because with the above method, the print will REALLY stick to the platform. I’d rather it over-stick to the tape than come off and fail (possibly breaking the printer).
We’ve been printing using PLA on scotch blue painters tape. I’ve learned a lot about how critical the first layer is. If you get it too thin, then it just barely smears on and will clog up the extruder. That can cause issues with subsequent layers not bonding correctly (in my opinion).
If you get it just right, and your tape is clean (alcohol wiped), then the PLA will stick very well to the tape. Maybe too well.
If you get the nozzle too far away, then the model doesn’t stick. Here is a picture of a bad first-layer (not close enough).
This model peeled up during printing and broke the cooling fan ductwork on the extruder head. Baaaaadddd. If you see a print starting like that, best to abort it and start over.
I was using a piece of foam rubber under my Arduino for the Marble Roller project. Not so nice, right?
Having a brand new MakerBot Replicator 2 on my desk, I decided to, well, make a nice bumper (or 3).
No design time needed as someone else did the hard work: http://www.thingiverse.com/thing:26237
Printed out wonderfully. Fits perfectly.
The Gahooa Perspective 