So, exams are finally over and I’ve had time to get playing with something again. Some friends bought me a NERF Vulcan for my birthday (cheers!) and of course I had to mod it up. The gun itself is now running off a 3-cell lipo pack, which about doubles the rate of fire, and has three ammo belts chained together to give 74 rounds in one continuous burst of fire.

I then hooked up an Arduino and servo motor to a series of cabletied pencils, which can pull the trigger on command. The Arduino and a USB webcam then connect to a laptop which is running motion and a small python script which interfaces to the Arduino and plays sound clips from Portal turrets when motion is detected/no longer detected.

The whole thing works very nicely, shooting down anyone who walks into my room with a rapid burst of darts, and terrifying them with the portal turret sounds.

Check out http://www.flickr.com/photos/randomskk/sets/72157624178093055/ for more photos!

I modified Principia Lab’s servo code from here and motion and Portal turret sounds from Valve’s Portal and the code below for controlling motion and the arduino and playing the sound files:

I’m part of CU Spaceflight, and since we have clearance to launch high altitude balloons from a few nearby locations we often do launches for other people. A lot of the time these launches have a radio transmitter on board, so that people around the country can pick up its GPS position and help keep track of the balloon’s location. We were helping with a launch a few weeks back which didn’t have a radio transmitter, but instead planned to rely on a GSM based system to text the GPS coordinates back. Jon Sowman and I decided to whip up a small radio tracker based on some work that had previously been done by Iain Waugh at CUSF. Six hours later, we’d completed Ferret 1!

It consists of an Arduino with some stripboard stuck on top. The strips on the stripboard were cut down the middle with a dremel, then we soldered some pin headers on to mate with the power and lower digital pins on the arduino. We didn’t need any of the higher digital pins, so the alignment issue wasn’t a problem. We then stuck an EM-406a GPS unit and a Radiometrix NTX2 on top along with a few required resistors and capacitors. The radio requires an antenna, which we constructed out of some mini coax soldered to the stripboard, a small square of copper clad board with a hole drilled through and five 17cm long bits of single core wire. Four of the wires were soldered onto the corners of the copper clad board sticking outwards, while the fifth was soldered to the centre conductor of the coax. This results in a quarter-wave antenna with a ground plane, which is great because almost all the radiation goes downwards in a fairly nice pattern.

Antenna done, we hacked up Iain’s code and flashed that onto the arduino and gave the whole system a quick test with a radio lying around the lab:

This was successful, so we put the whole thing in a box, taped it down and were ready to go!

We duct taped it to the main payload and launched it the next morning.

UKHAS Church Hill Site Project Orion
Uploaded by kannasnakka. – More video blogs and vloggers.

The first launch attempt didn’t go so well, with too little helium resulting in us almost taking out some football players. Luckily we were able to grab the balloon and refill, getting a successful second launch. The video above is from Scott James, whose flight we were piggybacking on. He had an actual video and still camera onboard.

Sadly, shortly after crossing the meridian the GPS lost lock and for a while we didn’t get any new data. Eventually, the data picked up again — but with a suddenly discovered bug in the code! We were transmitting the decimal part of the degrees as as unsigned integer when in reality they were signed, leading to some pretty significantly incorrect results. Luckily we were able to figure out how to determine the actual position from the data it was transmitting (thanks, SpeedEvil on #highaltitude on irc.freenode.net!) and the balloon was successfully recovered.

Ferret is on its way back to us in the post now and hopefully will live to fly another day!

P.S. we got hackaday‘d too!

It’s been a while since I’ve posted anything, so I’ll drop a quick update. Hopefully by the end of the coming week I’ll have finished on my current great big time-consuming project and will be able to write that up.

I’ve been playing with a few things besides that project, though. The first is stress testing my small and somewhat dodgy NIXIE PSU, which amazingly has managed to drive eight tubes:

I plan to finish off designing a somewhat more powerful power supply so I don’t worry about it exploding in my face every time I try using it. A better way to control it than a plain RTC would be nice too – I’m thinking of using a GPS module.

(p.s. yea, I accidentally set the 8 on instead of the 9 and didn’t realise — oops!)

The other thing I’ve recently been playing with is an accelerometer I got for Christmas. It’s the LIS302DL from Sparkfun, a super cheap 3-axis accelerometer that can output on SPI or I²C, has interrupts for freefall and things like the user tapping it or double tapping it, and adjustable sensitivity – 2g or 8g. All that for $20 on a breakout board is pretty incredible!

I hooked it up to an Arduino and after some messing around, was able to get it to read the acceleration data from the sensor over I²C, and then send it to the PC with a timestamp over serial (itself over USB). Then I whipped up a quick Python script to plot the data and decided I’d test something that’s now been thoroughly battered into me: in simple harmonic motion, the acceleration on a particle is proportional to its displacement, which can be shown to mean it is a sine wave. I figured that’s as good a way as any to test the sensor, and so I took the Arduino and swung it as a pendulum from its USB cable.
I then had Python generate the magnitude of the acceleration – the single value showing the resultant acceleration on all three axis. This was plotted against the millisecond value from the arduino and hey presto – a (nearly) perfect sine wave!

Not bad! The accelerometer itself:

(p.s. be sure to view the full sized image for tons of lovely macro sharpness)

arduino_loguino – the arduino code (CC BY-SA 3.0)

python_loguino – the python code (Public Domain)

For bonus points: Given that the Arduino was on the Earth’s surface, and that the millisecond data is accurate, how long was the USB cable the Arduino was swinging from?

]]> http://negativeacknowledge.com/2009/02/nixies-and-accelerometers/feed/ 0 http://negativeacknowledge.com/2008/07/robot-1-prototype/ http://negativeacknowledge.com/2008/07/robot-1-prototype/#comments Wed, 02 Jul 2008 11:57:13 +0000 Random http://negativeacknowledge.com/?p=25

This is my first prototype robot. A friend and I got together and he taped stuff to cardboard while I connected wires up and then we wrote a basic program and the prototype drove!

This version literally just drives forward until the ultrasonic sensor detects an obstacle, then it turns and drives forwards again. We can also rotate the servo the IR sensor is on and read the IR sensor, though we haven’t yet determined what voltages represent what voltages (it’s non-linear, typically).

The drive motors are the two servos I modified for continuous rotation, with an Arduino+Protoshield to control it.

This prototype has since been disassembled and work started on the new version, with a vacuum formed plastic chassis and normal geared motors harvested from an old toy. That one probably won’t get finished for a few months though, so I thought I’d post the cardboard prototype now.

It took a little change in the code to get it working with two matrices, but now it rocks!

Check out the video of it in action:

So I created a PCB for the Space Experiment mockup.

It’s functionally identical to the breadboarded version but is its own PCB, looks a lot better and has some text on it.

It actually took three tries to get it right. The first time, it all worked nicely and I etched it perfectly, only to realise I’d messed up the design!

I’m making the PCB as a shield for an Arduino, so it plugs into the top. I had the layout fine and was initially assuming I’d have the copper on the bottom (since the LEDs have to be on top), but later decided I wanted the copper on top. I quickly did a change layer to top for everything, without realising that I’d left the text mirrored! What printed, if made properly, should have resulted in mirrored text on top. Instead I made it so the text was readable – which means the copper had to go on the bottom for it to connect to the Arduino!

The next revision fixed this but was overetched and would have taken a whole load of fixing up, which is a pity.

Finally, I had it all perfect! All the text is really readable, I have a right angle socket for the LCD connector, and it all worked first time. Lovely!

I did have one new trick in the process this time:

This awesome plastic scourer thing has a nice, easy to use handle and really really tough plastic bristles, which are absolutely ideal for PCBs. Not only can I clean them at the start (and get the copper really nicely scratched up too), but it also strips the toner off like no one’s business. It’s taken about twenty minutes out of the process, especially when the transfer didn’t work and I have to take the toner off.

This was also the first PCB made with my new laser printer, which absolutely rocks – for PCBs, I print at 1200dpi, extra toner, high contrast, onto transparency, from the manual feed tray. I’ve set it up so I can just select the “Samsung_ML2510_PCB” printer on the list and all those options are used, so it’s just a case of hitting print and it does it. All the PCBs transferred really, really well with it! I can definitely recommend the Samsung ML-2510 mono laser for making PCBs at home.

So, I’m part of a team entering the UK Space Experiment Competition, and basically our proposal won out and is now one of six final proposals around the country! This is great stuff but we wanted to put on a display at a stand at a fair someplace, and for this we need something tangible!

The actual proposal calls for a piece of light detecting kit that runs some £1500 and we certainly can’t afford that, so instead I’m using a £0.50 light dependant resistor: pretty crude but works great!

What’s happening is the LEDs are being lit up, the Arduino reads the light level coming off the LDR and shows it on the LCD. When you put your hand or a piece of paper over the LEDs, the light reflects down onto the LDR, increasing the light level! In space, the bits of dust act as the paper and you can detect a whole lot less light.

The next step is vacuum forming a little case for the thing!