Timelapse

Video

Yes, yet another timelapse video of a 3d printer on youtube. How original.

Here its printing out a cellphone stand. The stl file can be found, here, for those interested:

http://www.thingiverse.com/thing:38681

In case you’re wondering, my head shows up way more often than I normally would given I’m trying to embed the usb cable into the piece rather than insert it after the fact. The usb cable I had lying around was a little too thick for the port provided in the stl file – really more a problem with the usb cable than the model.

Advertisements

First print

Video

Apologies for the video quality – I was trying a lot of different things before I got the printer to work and rushed into shooting after I finally got it down.

Shown here is the very first print I was able to run from the printer. The printer was able to deposit plastic for every layer it ran on, though by the end of the print the infill settings made the piece of rather poor quality.

Shown prominently is the new hotend – I started with an old bakelite pan handle I’d bought for a quarter at the thrift store. There was a preexisting hole in the handle that was too large for filament, and there wasn’t much room to drill my own hole. Bakelite is also a tricky substance to drill after its already been molded, as it’s prone to crack. As a solution, I filled the existing hole with high temperature sealant to constrict the diameter to the approximate size of the filament would pass through it. At the bottom, I affixed a pipe fitting and acorn nut with a 1/32″ hole drilled through, which was the smallest hole I could drill under my setup at the time.

It turned out after a week or two of calibration that this hot end suffered significant leakage. It did not help things when the bakelite isolator cracked after a failed attempt to reinsert the hot end. This would call for a new hot end, which will become apparent in subsequent posts.

Calibration

Video

Here is the printer in its first iteration. Notable things I would change include the the soldering iron based hot end and the sheet metal arm that holds it.

The soldering iron was mostly just the first thing I had on hand that served as a stand in. My soldering iron broke half way through working on the electronics and I found further use for it. Conceptually, it made sense – the bakelite used within handle would have a thermal operating range very similar to teflon. If anything, bakelite would work better as a thermal isolator as it inherently cannot melt and burns at temperatures comparable to the melting point of PEEK. The nichrome wire from the soldering iron also worked as a nice first order heating element. The 1/8 pipe fitting attached to the soldering iron was a failure to read existing documentation, on my part 🙂 The sheet metal arm I found was far too flexible, causing the position of the hot end to vary. It was soon replaced with plywood alternative.

X-axis calibration

Video

One of the biggest uncertainties going into this project concerned torque. Inkjet printer motors only need to drive a lightweight plastic header and maybe an ink cartridge. Depending on its purpose, the motor of a 3d printer would have to drive a 1kg filament spool of plastic, a bed with a printed object on top, or even an entire axis assembly. It did not help that I was borrowing skeletons from inkjet printers – because an inkjet printer is only intended for horizontal motion, I decided the best way to go was a static x axis driving a mobile z axis mounted on top of it. Doing otherwise with a mobile x axis, such as seen in a standard reprap, would put more weight on the z axis than could be born by a gear/pulley system. This was also very similar to the wolfstrap setup I’d been working with for the base.

Before continuing any further, tests would have to be done. Here, I test the motor driving the x-axis. A 1.5 pound exercise weight I had lying around simulates the approximate weight of the z axis I was working with. The endstop was also tested, which was scavenged from a printer and tacked on to be interrupted by some random ridge on the printer’s structure. An arduino was simply programmed to drive the motor until it reached the endstop, then backup.

The x-axis would turn out to be the only axis which absolutely, positively needed a stronger motor. This is without surprise the only axis which would drive the weight of another axis. The y-axis was perfectly able to drive its weight however the particular motor I’d found used a smaller step size and appeared jittery. Opening another printer could have potentially fixed that issue. The extruder I’d purchased came with its own motor and I figured I would save some effort just sticking with that.