We managed to get some footage of the tesla coil running last night, just before we switched over from a static gap to a rotary gap. It looks like the rotor is not quite in sync with the up and down swing of the AC line voltage. So that’s the next thing to fix.

We did the final assemble and checks on our Tesla coil on Saturday night, we rolled it out into the back yard and switched it on…

Nothing is as scary as that first time…

And…

Nothing exploded, imploded, melted, smoked or fried…

We were able to tease 7″ sparks out of the toroid with the safety ‘wand’. We tapped the primary on the recommendation of Teslamap - but we’re not trying to fool ourselves that this thing is tuned. We just built the oscillator tuner based on the 555 chip and we’re going to try and tune it properly.

The static gap is quenched with a tough little AC blower and set to about 0.18 inches. After operating for about 30 seconds the tungsten electrodes did not see very warm at all.

So after getting the tuning as perfect as we can, some of the things we’re going to examine next are:

* Height of the secondary from the primary.
* Height of the toroid from the secondary.
* Spark gap distance / better quenching.
* Addition of a rotary gap. (this is NST based, and so I’m not sure if this is safe…)

Our setup is:

Variac -> RF filter -> NST (12Kv) -> Terry Filter -> Static gap -> MMC (16 series with 10M 1/2 Watt bleeders) -> 0.25 copper tubing primary (15 degrees 12 turns total) -> 4″ 24 gauge wire secondary (19.5 coil height) -> 12″ / 3″ spun toroid.

I just found this clip on Youtube - awesome!

It makes me really happy to see that someone has got the plans working - great stuff.

I’ve been asked to write an article for Make magazine on MIDI controllers using lasers and light. It’s come at the right time, as I’ve been meaning to try out the Sharp GP2D12 sensors I bought from Trossen Robotics. Anyone that has played a Roland synthesizer with a D-Beam controller will recognise these immediately. They fire out a pulse of IR light at 25Hz and then triangulate the position of the reflected dot to get a distance measurement.

Having played around with them for a while, there are two problems that need to be addressed before they can be used for musical applications:

The first is that the output can be noisy - this is in part due to the fact that quite a lot of current is drawn every 40ms when the IR LED is fired - this can get into the output signal as a spike. First off a 300uF tantalum cap and 0.1uF regular cap can be put across the power rails close to the sensor input - this helps to smooth the power going in. Additionally this basic filter circuit can be put across the output to smooth out any remaining spikes.

[Circuit -coming soon!!]

Also only sample the output every 40ms or so, as the output is only updated this often. It’s also possible to filter in software like this:

forever
{
while (!sample_available())
{
do_something_else();
}

average = (average < < 3) - average + get_sample();
average >>= 3;
// do something with the average here

}

This gives each sample an exponential decay.

The second problem is that the output from the device is non-linear - so we need to apply a ‘linearization’ to the output. Thankfully someone has already done the hard part for us - the functions are as follows:

GP2D12:

Range (cm) = (6787 / (V - 3)) - 4

GP2D120 (long range):

Range (cm) = (2914 / (V + 5)) - 1

[Taken from http://www.acroname.com/robotics/info/articles/irlinear/irlinear.html]

Applying this function as either integer or floating point math helps in getting a value that can be further scaled to modify MIDI pitch bend (0-16383), or continuous controller information (0-127)

One final tip - be careful where you place the sensors - putting them too close to a side wall, or in the wrong orientation can affect the reading accuracy, but when used properly these are a valuable addition to any MIDI hackers toolkit.

I made some progress today - it looks like the majority of the components used in the PixArt hookup here:

http://www.kako.com/neta/2007-001/2007-001.html

Are for creating a clean 25Mhz clock signal from a bare crystal. If I buy a crystal oscillator, in place of a bare crystal, then I don’t have to worry about any of that stuff, I can just plug the output from the oscillator straight into the sensor.

I have to film the laser harp for MakeTV this weekend so I have to wait until that’s complete before trashing the WiiMote to get the camera out…

Some of you will have noticed that I was able to modify a WiiMote to do visible light tracking recently. My goal was to produce a gesture tracking system for the laser harp.

This was OK, but having to use Bluetooth, and Windows was a PITA - I needed a hardware only solution that could be wired straight into the harp.

At the recent Maker Faire I talked to Johnny Lee about the amount of information available on WiiMote internals - he directed me to the WiiMote project forum and I was able to locate this post that hinted at how the PixArt sensor could be attached to a microcontroller using I2C.

http://www.wiimoteproject.com/other/red-laser-and-i2c-bus-done-t579.0.html

Post #34 is pretty crucial:

Here’s how I get this to work:
Remove Pixart sensor from wiimote.
The sensor needs a 25 MHz clock signal here’s a circuit and pinout:
http://www.kako.com/neta/2007-001/2007-001.html
Connect the Sensor to your MC (don’t forget the pullups)
The slave address is 0xB0.
Initialise like here
http://www.kako.com/neta/2006-019/wii_ir_test2_src.zip
http://www.kako.com/neta/2006-019/2006-019.html
Now send 0×38
Capture 8 Bytes from slave
Capture 4 Bytes from slave
(I’ve seen this by sniffing the Bus)
I’m not sure about how to interpret the data(Haven’t even tried this yet, maybe it’s like the basic report?)

Please share your experience with the community.
If somebody knows how to remove the IR filter please post it here!
Does somebody know if the sensor can be used with 5V?

Fortunately the good ‘ol Arduino has everything we need - 3.3v output and an I2C protocol, accessible using the Wiring language.

Here’s a nice article on using I2C with an Arduino

Over the next few weeks I’m going to be trying to create an Arduino/PixArt/WiiMote hybrid that is capable of 200Hz blob tracking and MIDI controller output…

Stay tuned!

But in the meantime - check out LadyAda’s excellent new Boarduino USB!
All the goodness of a regular Boarduino, now with the FTDI chip/USB built right in.

This is me (looking slightly startled) with Limor Fried - also known as Lady Ada, head honcho of AdaFruit industries. Creator of the Boarduino variant of the Arduino microcontroller, and a whole bunch of other really cool gadgets and gizmos. Check out the SpokePOV - genius!!

http://www.ladyada.net/make/spokepov/

…and we got to check out his awesome head tracking demo…

WE NEED GAMES WITH THIS!!

Johnny Chung Lee was at the show… It’s great to meet some of the people you collaborate with on the web…

The Tesla guys (and girls) from UCSC invited me to step into the faraday suit and feel the power of 18 bazillion volts swarming over me…

Now I know how a cyberman feels…