Pictures:

^ RAMses in it’s current state; defunct. I scrapped some bits and pieces for modifications to the Velleman Vertex I’m building, but most of the structure is still there.

Since VT doesn’t let me upload mp4s, here is an avi of the CNC component working. This is the only recording I have of this CNC working before its demise.

Skills used:

3D Printing, Hardware Hacking, Electronics, Mechanical Engineering

Description:

While working on the massive Dynamo Death printer, I got horribly bored and decide to take a break from building the printer… and building another printer.

If you’re like me, you like to have an objective of everything you do (even if that objective is “Because I freakin’ can!”). And the objective of this build was to build a 3D printer as cheaply and quickly as possible.

I went from having all the parts on my desk to a working CNC machine compatible with Repetier in about 6 hours, and had filament extruding in another 2. The total cost came out to roughly $40-60, but I had most of the stuff lying around. Unlike my other printers, I didn’t care about quality, and just wanted it to work. If something broke off, I just hot glued it back together (the way real engineering should be).

Aside from approximately one (1) metric crapton of hot glue to hold the thing together, I scrapped the endstops, motors, and each axis from 4 separate desktop optical disc drives. I used the PISCES 3D printer to make a very minimal Bowden extruder that is basically just a bearing mount with a tension spring and a screw hole, and a I used some spare RAM lying around as structural components. The bed itself was made of laser cut acrylic coated in blue painter’s tape. As I have no heated bed, the only material I have half a prayer of running is PLA and maybe PETG, but whatever. After that I slapped on some Industrial grade googly eyes for good measure and spun her up.

This may be one of the worlds smallest FDM 3D printers, with a build area of 35*35*25 mm. However, with a build area like that, all you will really be able to effectively print is a Ring. Maybe. perhaps A Cherry MX key cap? Not much.

Update: I changed the drivers to A4988s to DRV8825s for more precision, but, blinded by the light of my infinite genius, I forgot to reset the potentiometers, and all of the stepper motors blew out in about 3 seconds. I have spares, but I have to rewire them.  Rest in Pastrami, little fella. 

Pros:

• 5 GB of RAM! That’s more than most of VT’s computers!
• Dirt cheap demonstration of CNC principles.
• Quick 6 hour build.
• A hell of a lot of fun to build.

Cons:

• Fragile; has broken twice
• Resolution is extremely poor.
• Miniscule build area renders this more of a proof of concept than an actual working printer.

Opportunities for improvement:

Seeing as how this build was a joke to begin with, its hard to think of things to do to make it a better joke. Perhaps I should use a smaller hot end so I can get a larger build area, or use different axes.

Conclusion:

With 3D printers getting more and more hyped, the price is dropping greatly, and for people that just want to see the concept in practice (like a high school), this is a pretty stupid and silly way to make it happen. TL;DR: it works, but not well.

Skills used:

Mechanical Engineering, CAD design, Basic FEA, Intensive Design for Additive Manufacturing.

Description:

Whew, it has been a while since I have posted anything, but rest assured, it’s because I have been too busy to post, not because I haven’t been doing anything.

When I wasn’t doing classes, or working in the makerspace I run, I have been, for the greater part of last semester, working on building a 3D printed drone that can both fly and drive.

How the hell did I get myself in this position? Well….

Back in Feburary, the Virginia Tech DREAMS Lab, which is the university’s Additive manufacturing lab, announced the Additive Manufacturing Grand Challenge, which was to use CAD systems and our own knowledge and skills to make a vehicle that could drive and fly for disaster relief missions made almost entirely by 3D printing technology.

In the Frith Lab, the Makerspace in which I work as the Lead Lab assistant, one of my cohorts suggested making a team to compete, since we had some serious experience with the printers in our lab. I gathered a few of the ULAs and a couple of freshmen to assist, and, as I was the manager of the lab already, I was volun-told to be team lead. This year, competition was fierce. 70 teams entered the competition (many were graduate student teams), and although many hit trials and tribulations along the way and dropped out, somewhere around 45 were competing for 10 spots in the finals during a elimination round. Since we had a fairly good idea how to work with 3D printers (some of us had built our own), what could possibly go wrong?

 Everything went wrong. After numerous redesigns, breakthroughs, team agreements, arguments, prototypes and lost sleep, we finally had something polished enough to go to the judges with. Being that we were a team of almost entirely freshmen, we lacked skills in FEA, and bumbled our way through the process by printing parts and putting them in our Instron tensile testing machine to get some sort of data. When it came time to present, we gave our spiel and left, pulling our hair out for the next 5 hours waiting for a decision.

After enough coffee and anxiety to kill a small animal, we got word back: <hackervoice>We’re in!</hackervoice> We started gearing up to finish the files and get them all submitted. The final result, a vehicle we jokingly referred to as the “Frith Prints of Bel Air,” was fairly impressive.

The best part, was that due to our knowledge of the threshold angle (the angle the printer stops printing support material), which we call the “magic angle.” We allowed the entire vehicle to be printed on ANY well-calibrated FDM printer, instead of just a $200,000 SLS printer.

We handled the obstacles really well, and had a few perfect scores. When it came to the integrated obstacle course, we were pretty hyped up and feeling good about ourselves when we spun up our rotors to fly through a window.

But suddenly, catastrophe. The wall sucked our vehicle towards it, clipping the rotors, and causing the entire vehicle to split into three pieces.

We rushed our vehicle to the ER, and dissembled the thing and re-assembled it from some scrap pieces. Our fatalistic pilot was sure that we were finished, given that the vehicle could not be stabilized with the pieces we had remaining. After some intense argument and convincing, we used some ingenuity and a soldering iron to weld together the broken plates, and by the next challenge, we managed to have a working vehicle again.

We scrapped together some points, and when it came to another window, we once again got sucked in and destroyed. This time though, we were ready, and we had the vehicle repaired and flying again within 2 minutes.

The final challenge was the deadlift, where we had 3 attempts to lift as much as possible. By now, the judges had been whispering that they wanted to see our team continue, as our team dynamic was both scratchy and scrappy, but some somehow effective. Since we had 10” propellers, we could do fairly well, but because our vehicle was damaged, we didn’t exactly know the lifting capacity. Having broken the vehicle twice already, our pilot was understandably cautious to not put too much weight on it. To which I called to him, and said “What’s the worst that can happen? We’ll break it! AGAIN!” After disconnecting on attempt 1, and securing 2 lbs of lift on attempt 2, we decided to go for broke, and put as much weight as possible on it for the last attempt, which accounted to close to 5 lbs. We gunned it, and just barely did not get the points, as it did not go quite high enough. Regardless, we placed second in the deadlift category.

After all was said and done, we managed to obtain the title “Zombie Car,” after our vehicle had been seemingly destroyed twice, and still managed to continue to compete. Also, for our team’s dynamic, skill, and over thinking process, as well as an amazing speech given by our team’s presenter, we received the Judge’s Choice Award. To celebrate, our team went home and slept, some of our team having been awake for 36+ hours.

Pictures:

^ Augmented Reality watch testing on Unity3D.

^ Playing Halo: CE on Oculus Rift using my gyroscopic game controller.

Skills used:

Spatial reasoning, Mechanical Engineering, Electrical Engineering, Unity3D, C#, a little JavaScript

Description:

Recently my friends have been bugging me to get an Oculus Rift, mainly to see what I would do with it. Well I finally broke down and bought one. So, here is what I did with it.

To start, I attempted to play Halo: CE on it with my Gyroscopic Game Controller. I have a computer with a decent enough processor to can play halo effectively, and I tried hooking my game controller up to it and playing Halo by aiming. Aside from having issues with Vireio, I ran into the issue that my game controller has no buttons to move, only an accelerometer (at some point I need to use quaternions to make that work, but four-dimensional coordinate systems are scary). Once I had everything hooked up right, the lag was just bad enough to make playing difficult. Oh well. Maybe later.

Also, there are a number of interesting games and tools that can be used with the rift, like VRclay (for those of us that use 3D printers) or any of the games on Oculus Share, and Virtual Desktop has been fun to watch movies, do work, or (my personal favorite), turn on the iTunes visualizer, wrap that 360 degrees around, and be “immersed in the wubs,” as it were.

However, I am more creative than just playing with it the way it was supposed to be used. What if this could be used as a tool?

Following in the footsteps of our lord and savior William Steptoe (hallowed be his name), I hooked up two webcams to the front of the rift, along with a 180 degree fish-eye lens and used Unity3D to create a link to the rift through them. After learning the basics of C# and JavaScript, I started coding up a H.U.D. that displays system time. Not long after I started playing with Unidunio (which allows unity to interact with an Arduino), NyARtoolkit (an augumented reality software), and coding up what could, potentially, be an infinitely expandable build platform. Think about it, you can look at a glyph on your wrist and it blows up into the greatest smartwatch to not actually exist! Need to remember where you put your car in the parking lot? Set a waypoint and follow it back with GPS through Arduino. Don’t know your heartrate? Run a heart rate sensor through Arduino, throw it on the HUD and have it automatically call 911 when your heart rate drops too low, goes too high, or changes at too great a rate. Want to fly? Hook two cameras up to a quadcopter and fly around, displaying the view from the quadcopter to the Rift. The possibilities are endless!

Conclusion:

For those of you who are interested in buying a Rift but are not sure, I must say that, if you are looking to play games on it, wait for the consumer version. However, if you are looking for an immersive space to use as a maker’s build platform, the devkits are extremely useful and fun to play with in that regard.

Pictures:

^ The Inventor model of the the gantry system.

^ The printer in full. The testing transformer can be seen to the lower right of the image.

^ The brilliant arc flash of a 3D printer printing steel.

Skills used:

Autodesk Inventor, Marlin Firmware, Mechanical Design, Electrical Engineering, Welding.

Description:

So I have been interested in 3D printers for a while, and I decided to build one. However, I did not want one a common one like a Delta or a Reprap. Many of those are too small for what I wanted to do. My requirements for a printer are at least 200mm x 200mm x 200mm build area, decent accuracy and fairly cheap to build (< $500). Oh yea, and I wanted it to print in ABS, PLA, and steel. Since I could not find anything within my price range that could print in steel, I decided to make my own. I needed a name for my custom printer, so I called it a PISCES (Progress-of-Ingenuity Sampling and Constructive Engineering System).

Alright, Step 1: Design a method of printing metal. Most additive manufacturing devices that print steel use Selective Laser Sintering (SLS) technology. However, this is completely out of my price range, and moreover, this would not allow me to easily switch between thermoplastics and steel. So what about a crazy idea… Why not print steel in an FDM process? I know what you are thinking: “But Scott, steel is a hard substance with a higher melting point than ABS, you won’t be able to extrude it.” Maybe not, but I can adapt an extruder to become a wire feed, remove the tip, and I can use a DC electric current to melt it at the point of contact with the baseplate. This idea came about after taking a college welding course. The process of MIG welding with flux core welding wire works practically the same way… with a few modifications. This is the basic schematic for the design:

After I had an idea where my project was going, I used Autodesk Inventor to CAD a basic gantry system that has a spool holder on top of the extruder head. I used the Frith Lab’s laser cutter to cut the side plates out of 0.2″ acrylic, and its 3D printers to make the parts. After installing a RAMPS1.4 board and adapting Marlin firmware to fit the need, I started running ABS filament. This worked fairly successfully (it was not perfectly calibrated), but it worked well enough to move onto the next stage of implementing steel.

I then started thinking about how I could melt steel to a ground plate. I put a copper plate on top of the heated bed (so the steel does not stick during testing), and ran that to a microwave transformer, and attached that to a diode so I have a somewhat DC current. As of now this is 2100VAC at 0.57A (the reason for such low amperage for testing is so that I can use easy-to-maneuver, high-gauge wire without it melting). I have a second transformer wound to output the 65A, 18V needed for MIG welding and a couple Schottky diodes on order in order to make a bridge rectifier, but I am waiting to put that on until later.

And then, suddenly, a hiccup in my design. I was using a high voltage transformer hooked up to a HV diode to make a modified, positive-current-only circuit for testing (because it is easy to see when it works). It turns out that leaving the grounded heated bed wired to the RAMPS board while the transformer is running runs an AC current all the way from the extruder head to the bed through the electronics, causing my thermistor to turn into a lightbulb, and then into a fireball. I have since corrected this issue, but I have yet to replace the thermistor, because I want to get everything else running for fear of burning it out again.

There is still a lot of work to do, but progress has been made, and the proof of concept is there, and besides, no good project is ever really done. Now, where to order a few hundred more thermistors…

[The files for this printer have been made available here.]

Pros:

• It was fairly cheap to make, costing only around $400 for a build area of 8.25 x 8.25 x 10, not counting the cost of 3D printed ABS material.
• I was not intentional, but the machine ended up looking, aesthetically, pretty good.
• I learned a huge amount about 3D printers, Marlin, electronics, etc.
• I have more CAD practice then I ever thought I would get on a single project.
• 3D printing steel in an FDM process opens up a world of possibilities unimaginable for thermoplastics (Printing silverware, circuits, etc.).

Cons:

• The Catch-22 of needing a 3D printer to print a 3D printer.
• The plates were cut with a laser cutter, which just is not that easy to get access to.
• The whole case is huge, and takes up more space than it probably has to. This also allows air drafts which may cause ABS parts to peel up.
• The modified DC current used for the welder has damaged the electronics before,
• MIG welding throws sparks in all directions. I need to put a cover on to prevent the entire thing from becoming a major fire hazard.
• The light produced from the welding tip is blindingly bright. Anyone nearby will need to wear welding goggles to prevent blindness.

Opportunities for improvement:

Well for starters, the calibration of steel will be difficult. Heat management will be brutal, especially since the whole machine is made of plastic, and I may need to implement breaks into the G-code in order to allow the part to cool for some time. If the feed rate is too high and the steel gets stuck to the plate without initiating an electric arc, getting the wire off may be a challenge (perhaps adding a new G-gode command to discharge a capacitor to blow up the remaining wire?).

I need to find a way to lower the light output. The electric arc is blindingly bright (ever wonder why welders wear welding helmets?). I can put covers on the side plates to reduce this, but then it would be difficult to see the part. I am hoping that lowering the voltage will correct this, but I may need to implement a light filter into the plate system.

And there there is the whole idea of design for additive manufacturing, which is a field of study centered around how to designing a part to best be build on a 3D printer. Using an electric current adds a whole new level of complexity to the mix, as this requires that the wire feed be the smallest point of contact between the extruder head and the baseplate, so small features are out, and cohesion between liquid metal will be an issue as well.

Conclusion:

It is a good start to an interesting idea. It already runs ABS plastic pretty well, and some safety limitations need to be put in place when dealing with such high amperage, but the proof of concept is there, and I am working to get it up and running with steel successfully.

Pictures:

^ The device remotely accessing my school laptop. Note how windows can be run on the phone fairly easily. The background is a Rainmeter layout.

^Running Autodesk Inventor over Kainy.

^ A close up of the structure

Skills used:

Autodesk Inventor, CatalystEX, Kainy, Additive Manufacturing, Basic Mechanical Engineering.

Description:

Based off of the success of my original 3D printed wrist mounted cell phone case (whew, that’s a lot of words. Let’s just call it a 3DPWMCPC), I decided to fix some of the errors in the old version and add some new stylings and features.

While it is still sized to fit a Samsung Galaxy SII, the first major difference in the 3DPWMCPC is the case itself. For starters, I fixed the placement of the slot in the top to fix correctly, and put a larger hole in the slider to accommodate larger chargers. I also replaced most of the fillets with chamfers in an attempt to make it look more masculine and sleek, and printed it in black for this same reason. The pictures do not do this work justice, and you will need to look at the CAD files to see it properly.

One of the biggest overhauls to the system is actually the addition of a certain software called Kainy, developed by Jean-Sebastien Royer. The software allows remote access and control of computer systems via a cell phone over WiFi. Also, unlike many remote access systems, it provides a direct link to the computer screen, as opposed to just the files. With the additional facts that Virginia Tech has WiFi all over the campus, the capabilities of Google drive, and the ability to leave all of my computers on permanently, I have complete control of all of my computers at any time, (given that I remember to leave them on) from my wrist.

So far, I have used this technology for fun little things like playing Halo: CE, adapting a part with Autodesk inventor and uploading to Google Drive, using it as a clicker for PowerPoint presentations, and moving files that I left on my desktop computer to Drive so I can access them on my laptop. The potential is practically endless.

(The IPTs for this part have been made available here)

Pros:

• Still a cool idea and one of the most useful things I have created to date, as I have used it on a daily basis.
• I can access all three of my computers from my wrist from anywhere in the world, so long as I have internet access.
• It looks really cool, and I have been noticed by a large sum of people because of it.

Cons:

• Kainy saps battery like crazy, and I have seen a drop in battery life since its implementation.
• Changing the fillets to chamfers created a number of points, which poke the user on occasion.

Opportunities for improvement:

I have been dying to get a new phone, and when I do, I will have to hit it with the vernier caliper again to get the measurements right. Once I get that, I want to attach a solar charger and a portable power pack to counteract the power drain by Kainy.

Conclusion:

This has been one of my best inventions, and I am just making it better. Looking towards the future, I am trying to make it better, but it is almost what I want.

Pictures:

^ Playing Halo CE on the device

^ The device with no phone

Skills used:

Rhino CAD, Autodesk Inventor, CatalystEX, Additive manufacturing, basic Mechanical Engineering

Description:

This is a simple 3D printed cellphone case attached to a pair of heavy reading glasses. Designed for use with the Gyroscopic Game Controller.

This was printed on a Stratysys uPrint SE Plus using the software CatalystEX, and assembled using a laser-cut acrylic panel and paper clips.

What really makes this interesting is that it uses a cell phone as a second screen using a software named Kainy, developed by Jean-Sebastien Royer. Kainy remotely accesses a computer running a video game and can remotely control it using the accelerometer in the phone, allowing the user to look around and have that translated into game movement.

While nowhere near the quality of Oculus Rift, it still allows the use of VR technology for approximately $20. There is a little bit of internet lag and no 3D functionality, but it is still very cool to play Halo: CE with it.

[Update 2/18/15: I decided to buy an Oculus Rift, so this is obsolete now.]

Pros:

• It is much lighter then my original idea for a helmet, weighing in at about 1 lb. compared to the original helmet’s 25.
• It works for anything that runs off a computer.
• Surprisingly comfortable to wear.
• Open-air design allows the user to see infront of the device, allowing it to be used outdoors.
• Kainy is an interesting software, and I have been dying to try out new things with this.

Cons:

• It has some slight lag, changing based on location.
• Focusing on a point about 3 inches in front of one’s eyes begins to hurt after about 5 minutes.

Opportunities for improvement:

Focusing on the short distance causes eye strain rather quickly; I don’t think the lenses I have are large enough. I might need to break a magnifying glass to get the kind of focus I need.

Conclusion:

It’s rather quick and dirty. Not wonderful, but better than what I currently have.

Pictures:

^the Device itself, made of clear plastic acrylic

^ A closeup of the Arduino and trigger system

^ The carrying case

^ The device in its case

Skills used:

Rhino CAD software, Electrical Engineering, Mechanical Engineering, Arduino IDE, Computer Science, Vector Geometry

Description:

Ok, so this was a hell of a challenge, and currently is still one. This is an acrylic gun-shaped mechanism, approximately 4 feet long, utilizing the Arduino Esplora’s ability to function as a mouse and keyboard in order to play first-person shooter games more realistically. While this would function quite well with an Oculus Rift, it plays best on a windows computer.

What inspired me to create this in the first place was not for fun, but for exercise. I for one, have an intense hatred for treadmills. Running as fast as you can and getting nowhere… there isn’t even a better metaphor. Since the controller has Euler angles and accelerations in that tilted coordinate system, it is possible to determine accelerations, and thus movement, in all directions. To put it simply, this controller allows one to play a FPS (Halo: Combat Evolved, currently) in a large field (Virginia Tech’s Drillfield), translating real movement into in-game movement. It is much more fun to run around shooting aliens and capturing flags then it is to just plain run.

The system consists of a Parallax gyroscope communicating with the Arduino through a hacked SPI header. The Esplora itself has buttons that control aiming lockup, reload, grenade, action, pause, switch weapon, flashlight, as well as other necessary controls. Firing is controlled by pins that used to be photoresistors, but were cut off for more inputs. Movement is controlled by an accelerometer, which is the only part of this design that is functioning out of whack. Due to the insanely fast refresh rate required to keep the angles correct, I try to minimize matrix operations, so my accelerations are not 100% correct, and that leads to noise and drift. I am currently working on putting a Kalman filter on it, but again, speed is critical to the operation.

To play this, I have a helmet-mounted computer utilizing a low-power intel 4400 Haswell processor to play Halo: CE at minimum settings.

How does one store, carry, and protect such an item? In a gun case of course, with a snarky warning label.

Pros:

• Extremely cool.
• Frees me from treadmills for the foreseeable future.
• Halo: CE is the greatest game ever made and deserves to be played in reality instead of on a flat screen.

Cons:

• I have never felt more sketchy in my life walking into a Virginia Tech residence hall with a gun case.
• I have issues with the movement vectors.

Opportunities for improvement:

I need a Kalman filter to make the accelerations more accurate, and thus make the darn thing actually work.

Conclusion:

An interesting idea, and all the parts are there, but I don’t have the machine shop access or the materials to make it as perfect as I think it should be.

This a cool idea, and I really want to see it done, but damn, it’s a lot of work.