Wednesday, December 22, 2010

RepRap: Droid Mount

My wife and I upgraded to Droid X phones this year. As soon as we got them we realized they outperformed our old Garman GPS and have been using the phones for GPS ever since. My wife quickly identified the need for a droid mount to hold the phone while driving. A few weeks ago I designed and printed the following mount.

It worked decent for a week and then warped! I'm using PLA plastic and its full blown winter right now. My wife had the heat on and the part that holds the phone warped. You can see the warp below. I redesigned using the strength of triangles and increasing the depth of the part.

While redesigning I had an idea. Though the three degrees of freedom design worked fairly well, I wanted a bit more freedom. I made a bundle of 25 strands of floral wire and wrapped it with electrical tape. I then incorporated it in the design. The wire serves a dual purpose. It allows virtually any position to be obtained and it absorbs pot hole shock.

Here it is installed! I'm happy!

If you have a Droid X and would like to make this mount, here it is on Thingiverse. I included the OpenSCad source file. If you want a mount, but have a different type of phone, it shouldn't be too hard to create a derivative of my design for your phone specs.

RepRap: Production Mode: Bed #6 and #7

Bed #7 is the last bed and completes the set. In the next few weeks I will post these sets on Thingiverse and add links to these entries.

I intentionally left out the tooth pulleys from the large batches because I've found that they come out cleaner if printed separately. Bed #6 is just for printing the large z tooth pulleys, and bed #7 is for printing the smaller tooth pulleys that go on the motor shafts. You only need three small pulleys for a Mendel set, but I've found that printing six at a time increases quality by letting them cool between layers. I took some pictures of the pulleys with some extra belt, showing that they are functional. These pulleys are the toughest parts of the print to get right. It's a good idea to have some extra belt lying around to prove they are functional pulleys.

Tuesday, December 14, 2010

RepRap: Production Mode: Bed #2

Because of my background in Computer Science I tend to complicate problems beyond necessity and must step back in order to arrive at a simpler solution.

Problem: How do you divide all Mendel parts into 5 sets of equal print time and bed size? The goal is to have 5 twelve hour print jobs that can be started in the morning and evening.

Complicated Solution: In Computer Science this problem would be classified as a knapsack problem, with print time and bed size equally playing into the "cost" of each item. I more than considered coding one of the knapsack solutions and running it on part data before stepping away.

Simple Solution: Assume that a long print time is always equal to a large bed size with each part. Create a list of parts with name and print time. Order the list by print time. Treat the parts like people and ask them to count off in order from 1 to 5. Group all 1s together for your first set. Group all 2s together for your second set and so on.

The simple solution can be carried out in a spreadsheet in less than a minute, and it gives sets that would take between 11 and 13 hours to print, according to data from It's then only a matter of tweaking the sets by taking a larger part from one of the larger sets and swapping it for a smaller part from a smaller set. I did this a few times until I had 5 sets that should each take almost exactly 12 hours to print.

Here are some images of the first set.

In the next few weeks I will add this series of STLs to Thingiverse.

Monday, December 13, 2010

RepRap: Production Mode: Bed #1

A lot of people are printing Mendels these days. This is really good for the community and for growth. I started printing Mendels for others in November, and have printed and sold two sets so far. I'm currently working on printing my third set of parts and decided to take the time to blog about the whole process. With this being my third go I hope it will be seamless. I will be writing an entry for each bed of parts I print, and releasing the production files I am using as I go. I feel this is one more way I can give back to the community.

The first production file that I will be using was done for me, and is up on thingiverse here. All other production files I am in the process of making myself.

The following images show a production file with all the parts you need to make a Wade extruder. Depending on your feed rate this file will take anywhere from 3 to 8 hours to print.

Sunday, November 21, 2010

RepRap: RepRap is reproducing!...

I finally finished my first set of parts for someone other than myself this week! My extruder has been relatively consistent and I had a reason to put in deliberate effort as they were willing to pay for the parts. The following are a few finished product images of a complete set of Mendel parts with an Adrian extruder. The quality that I'm getting on my Mendel is really encouraging.

Wednesday, November 17, 2010

RepRap: update: aluminum extruder perfection!

I was able to use the extruder I made in the previous entry for 10 hours before it jammed. The jam was due to the PTFE expanding with the pressure of the hot plastic and causing a bubble right before the transition zone. I cut the PTFE open to show this.

I find at times like these, when you've failed after many tries, it helps to talk with another mind. I talked things out over coffee with a friend and we came up with a solution. If you follow Nophead at all you will be familiar with his plumstruder, and you will see where the inspiration came from.

I'm now surrounding the PTFE with an aluminium heat sink that does not make contact with the hot end. This keeps the PTFE cool enough to not expand, and to keep the transition zone short. I'm also using a block of wood to further insulate the hot end from the heat sink.

I've been able to run this setup for over 30 hours without fail.

Sunday, November 7, 2010

RepRap: aluminum extruder perfection!

I made a few design decisions since the last OpenSCAD renders. I decided to stick with the vanilla x-carriage so that others can benefit from the design without having to purchase linear bearings. I also increased the size of the heater block at the suggestion of an engineer friend. The other design changes are a result of trying stuff until I got something working. I am not a mechanical engineer by education, nor do I remember much from thermodynamics, so there was a lot of trial and error before the design matured into the following.

The second to last picture shows me tightening up the whole rig with a 3mm rod used to align it. This is critical. I tightened a few designs without the alignment rod and the filament was really hard to push.

One element of the design that is very exciting is the fact that the top most plate of aluminum is cool to the touch! I think this is a result of three plates.

Finally, the PTFE is inside the hot tip, instead of the other way around. This really does make all the difference.

I have printed for about 4 hours without it failing. I'll leave it running for the rest of the weekend and report back on how it is holding together.

Monday, October 18, 2010

RepRap: another custom extruder...

So... my extruder... broke again. Ha. I don't think I can count the number of times I've said this or read this on another blog. This time the glue holding the PTFE to the PLA drive mechanism loosened and it fell out. While putting it back in I ended up breaking the nichrome wire. Ha. I laugh now, but I really am tired of fixing and fixing my extruder. I decided to put my printer on a metaphorical shelf for a few days and put some thought into my current design's shortcomings. Here is a list of the shortcomings I have identified.

1) It occasionally leaks because the hot end is inside PTFE instead of the PTFE being inside the hot end. This idea is inspired by Nophead's plumbstruder post and Adrian's geared extruder nozzle.

2) Don't use glue.

3) Don't use brittle nichrome wire. Even Makerbot has stopped using nichrome wire in their latest MK5 design.

4) I don't have much room to work with in the current Mendel design.

So, what I've decided to do is completely redesign the xcarriage. This will give me more room. Also, I will completely redesign the extruder so that it doesn't use glue or nichrome while incorporating the PTFE update.

I am attaching some OpenSCAD renders of what I am thinking of making. I won't have all the materials for a few days, so if anyone has feedback on the design I still have time to change it before I pull out my lathe, taps, and my backup Darwin printer. Drop me a line or leave a comment. Otherwise, be prepared to learn from my mistakes yet again. Ha.

Wednesday, September 29, 2010

RepRap: close to a full bed of parts...

Now that my quality is up and I have my spool working properly I wanted to try a full bed of parts. Here are 20 bar clamp parts needed for a Mendel. These parts took 7 hours to print.

Sunday, September 12, 2010

RepRap: Mendel calibration through gears...

The best way I've found to calibrate a RepRap printer is by printing gears. Any flaws show up immediately. Also, if one of your gears break after several prints some of your bad calibration prints could work enough to print a better gear. This weekend I was able to print all the Mendel gears, as well as all the Wade extruder gears. Pictures follow.

I am satisfied with the quality. Now I am ready to print a full bed of parts.

RepRap: Spool Extensions...

The spool I printed up a few days ago works, but has occasionally jammed as a result of the filament coming off of the ends of the spool. To fix the problem I designed some extension tabs in OpenSCad

printed, and installed.

I think this will be a solid solution. Now I just need to upload it to Thingiverse so others can benefit from it as well.