uncategorized « My Site/Blog
Untitled Document


Citi Bike Key

It’s been a while since I’ve posted but with the creation of a site dedicated to larger projects this website got put on hold. With that in mind, this blog can now finally serve as, well, a blog.

If there’s one thing I really hate it’s carrying things (don’t get me started on physical keys). I already carry my cellphone and wallet wherever I go, and people tell me how cellphones nowadays can do anything… except unlock a Citi Bike. Why the company behind the bikes decided to make the docks and keys frustratingly hard to use is beyond me. So, with my hatred of keys and desire to enable my cellphone to do more things, I created this monstrosity:


I should also mention my hatred of my cellphone case which enabled me to take more destructive measures than I normally would. While the design is far from final it gets the job done. The key itself adds four millimeters to the phone (5 1/2 for the spacer and screw), which is essentially unnoticeable unless you’re holding it in your hand. I printed out a base the approximate size of the key so that it could easily swing out and not hit the edge of the uncut rubber. I then printed out two nubs to prevent the key from swinging freely when in my pocket. Everything was then superglued down or screwed in place. For everyone who’s asked why I have painters tape under my case, you now know why (this wasn’t the first project where I attempted to add to my phone).


Sure it doesn’t look the best, but it works. In the name of being more practical, I can sacrifice a couple millimeters of pocket space. As always, please don’t hesitate to ask any questions or tell me how much worse my phone looks.

And visit my new site, if you haven’t already.

Raspberry Pi Action Camera

It’s been a while since I’ve made a post however, that is not due to lack of projects. Almost six years after setting my eyes on the Makerbot Cupcake CNC, I finally purchased a 3D printer and have been putting it to (hopefully) good use. I felt as though I had reached a point where fabricating enclosures was the main reason my projects weren’t going forward. I’ll probably spend another post talking about the printer but I thought I would share my first project (and 4th and 5th prints).

I picked up a GoPro as a safety measure while riding my bike and wanted to get a second so I could record forwards and backwards. The GoPro was a nice start however, it wasn’t really suited for taking extended video (say on a multi-hour ride). I didn’t really like the idea of having to carry multiple batteries around with me and then having to stop to swap them out. Right around this time, a new Raspberry Pi model came out called the A+, a small, stripped-down, low-power version of the A. It looked perfect. A quick search on Google gave me a few results for people who had turned their Pis into cameras, and found a post with some great information on how to go about doing it.

Knowing the idea was plausible, the next step was to figure out how to build it. I’m in a transition period with modeling software so I opted to use some designs already on Thingiverse (just for the prototype). I printed out cases for the camera module and for the A+ and had two great platforms to work with. I then set about wiring up the necessary connections. I didn’t want to alter the Pi too much so I soldered my connections to the bottom of the headers. I also spent some time removing the headers on the bluetooth-serial module to save that extra bit of space. Two holes were then drilled in the back of the case for the start and stop buttons, completing the build. Check out the photos below:





You can also watch this, somewhat, embarrassing video of me demonstrating the camera quality (the video is from the Pi, photo booth is so you can see the camera is shooting and the other monitor shows the video from the camera):

One thing to note is that I had to edit the camcorder script to work with the new 40 pin headers on the newer Pis. GPIO.setmode was changed to board, and the pins were updated from 23 and 24 to 33 and 35. There are a host of additional updates I’d like to make to the py script as well as to the hardware of the camera. Once I’m happy with where the project is I will do a full, detailed, writeup.

Designing a Better Test Box for Fencing

The idea behind this was born of part necessity and part boredom. As primarily an epeeist, I’ve had my fair share of faulty weapons and would end up having to use the connection detector on my multimeter to debug any problems. This was not the most practical approach. However, instead of shelling out $15 for a box with LEDs I decided to go and make my own.

The first thing I wanted to avoid was using two batteries. A regularly used tester would see maybe an hour of on time in a couple years and I really wanted to make the case as light and slim as possible. However, that left me with a different problem. The reason most testers use two batteries (AA) is because the general forward voltage for LEDs between 1.8 and 2.2V, just over the 1.2 or 1.5V you’d find in an NiMH or alkaline AA. There are a few ways to go about boosting the voltage to the necessary levels, but the approach I took was to use a circuit called a joule thief.

I found an easy-to-follow instructable and made a quick trip down to the EE lab to see if there were and toroid magnets which I could use to make the inductor. Lucky, there were. I only managed to get 7ish complete turns around however, it seems to work just fine:

IMG_9708I wanted to see how low the voltage could drop before the LED would become unusable so I hooked up the circuit to my bench supply. I slowly dropped the voltage to the point where the LED was no longer emitting any detectable light which was around 0.3V

DS2000Above is a capture at 0.3V in. The top waveform is from the toroid output to the base of the 3904 and the bottom waveform was of the LED. This makes it much easier to see the LED being powered by higher voltage pulses from the joule thief circuit. The capture below was taken at around 1.5V in and shows the difference in frequency relative to voltage (excuse the different time bases):

Newfile1iHaving solved the issue of the voltage supply à la joule thief, the test circuit is then a simple matter of wiring up three pins to connect to a body cord and then foil or epee. IMG_9712Above is a photo of the body cord that will be connected to the pins in the test box (the test box whose housing needs to be made). The middle pin (B) is connected to the collector of the 3904. The left pin (A) is then connected to a green LED whose cathode is connected to the emitter or the 3904, such that when the circuit is closed the LED will turn on (the weapon is working). The same is done for the right pin (C) with a second LED. Thus completes the wiring component of the test box, I’ll cover the housing in another post when the fab shop reopens.