cool things at Cool Fusion

Thermomo-Jacket

28.7.07


Click here to find out more!







At around 10 degrees Celcius the LED lights blue. It's nice and warm in my house, so I took a quick walk in the winter air. Brrrr!





Things are heating up- this is around 18 degrees. Temperature is mapped to hue. The saturation and brightness of the colour was kept at 100%.





In the mid 20's the LED glows yellow then orange. This is a really warm jacket, so a fast temperature rise is inevitable.





In the high 20's the LED glows red, or even a little violet. Hot stuff... coming through!




The LED is very bright, and the three colours are produced by three separate LED's, and so I'll need to add some sort of diffusing material to cover it. Otherwise, a colour split occurs. In the picture above, the colour should be orange, but we see separate green and red colours. For the series of pictures above that, I used a small piece of masking tape. Obviously this won't do for high fashion!





Here's the LED itself- its pins have pushed through the jacket, and are wired up on the other side. It's an RGB LED- which means that one can control the individual red, green, and blue lights. The LED and the thermistor (a resistor that changes its resistance with temperature) are connected to the microcontroller board:






A: In the foreground is a power regulator, which provides nice, stable power to the microcontroller. B: This is the microcontroller. C: The brown pillowy thing in the background is the ceramic oscillator- it's the component that provides a 16MHz timing signal to the microcontroller.


Here is how the thermistor was wired up:



The Arduino board has several analog inputs, that require a varying voltage. This voltage needs to change as a function of the resistance of the thermistor. The best way I know to achieve this is with a voltage divider:






As the resistance of the thermistor changes, so does the voltage 'XV'. This value is read by the Arduino board, and then in software I map this to colour. The thermistor is hidden near the LED, next to the wool of the jacket, and so it tends to get very warm very quickly.



Here's another detail of the wiring to the thermistor. The blue wire gives the variable voltage level. Notice how everything is heat-shrunk. This ensures no short circuits can occur as the thermistor moves around inside the jacket.

Cymixics

18.7.07

Cymatics 2- return of the killer cymatics!

This time cymatic frequencies are used to create beautiful patterns, using dyed and un-dyed cornstarch solutions. I call this "Cymixics".

Click here to find out more!

I used an old electromagnet from a fish tank pump, with a metal (Meccano!) plate held about a millimeter above it. On top of this lies a plastic lid that holds the solutions. This sets up powerful oscillations that would be impractical to produce using sound waves from a speaker.


Here is an inert blob of cornstarch suspension, lying on a plastic lid, surrounded by red coloured cornstarch mixture.





As the power is applied to the electromagnet, ripples appear. These indicate tiny currents within the solutions, and these cause mixing to occur.




As the mixing continues, the patterns become more complex.



Eventually the red mixture becomes pink, because both solutions are water-based. If two non-mixing liquids can be found that don't limit detailed patterns because of their surface tension, I'd be keen to try them..



These pictures do it little justice! Click on any of these images to watch the video that these frames were derived from.

This time, instead of the sound that was used to produce the mixing, I've dubbed in some pleasant music that I made. (Who wants to hear the buzzzzzz of an electromagnet, after all?)

Homecooked laser scanner

I recently converted my home-made Meccano 2-axis webcam mount into a laser mount.

Click here to find out more!

The system is made of an Arduino board that sends timed signals to two servo motors, mounted at right angles to each other. This gives a crude x&y pointer, which I can use to 'draw' patterns on any surface.



My desk is illuminated by many vertical sweeps of a 1mW red laser. A Leica C-LUX1 was used in 60 second extended exposure mode.

left-click this image to see a larger version.





This one is a visualisation of 'Brownian motion'. Brownian motion is analogous to a 'random walk'. The walk is not truly random because the same pattern will be generated every time the system is reset.

left-click this image to see a larger version.

This image is a visualisation of the morphology of the 2-axis mount for the laser. It's clear from this image that the unit is more stable when making vertical sweeps than when it makes horizontal sweeps. The increment I used for moving the laser also affects the quality of the scan.

left-click this image to see a larger version.

Now I'm telling the laser to point along 8 different directions. The unsteadiness mentioned in the previous image is seen along any direction that has a horizontal component. The increment used in this picture is small, so overall the quality of the line is smoother.

left-click this image to see a larger version.

My desk is illuminated by an expanding 8 sided spiral. The laser gives a clue as to the type of material it hits, for example the translucent plastic of my computer versus the opaque wood of my desk.

left-click this image to see a larger version.

In a future post: I've added a microcontroller switch to the laser, enabling me to 'pick up the pen' as it were.



Super-sneaky sneak preview: Here's a random dot pattern, playing across one of my clay sculptures.

left-click this image to see a larger version.

Cymatic Akira-fu

3.7.07

Click here to find out more!
Have you heard of "Cymatics"? Have you ever stood close to a bass-bin at a rave, or concert, or audio enthusiast's house? Then you'll know that sound waves, at the appropriate volume and frequency can move physical objects. There are two major categories to this- moving particles (for instance talcum powder or mushroom spores, and liquids, such as cornstarch suspensions).



Here is an inert blob of cornstarch suspension, lying on a plastic lid. Such a suspension is known as 'non-linear'. One practical outcome of a non-linear substance is that it may behave as either a solid or a liquid, depending on how it's interacted with.





As the sound is applied, ripples appear. I used an Ensonic ESQ-1 to produce the sounds. It's a 1987 vintage, and so the capacitors are 20 years old, and hence the pitch of the sounds tends to drift. With a higher quality sounds source I could control the blob much more accurately.






As the sound's volume is increased, tendrils rise up from the cornstarch- with a little imagination it's easy to see limbs! Yikes!









It's even possible to produce fleeting 'archways'.








With yet more volume the tendrils become longer.









These pictures do it little justice! Click on any of these images to watch the video that these frames were derived from.
WARNING: LOUD!!

LED storms

left-click this image to see a larger version.




In this capture, you can see an ordered set of rows of LED's, as well as more of a storm in the foreground.

left-click this image to see a larger version.

So what have I learned from this? Well, for a start I think in the next version I'll space the LED's more closely, which will give a better vertical resolution.


(Bear in mind that the goal for the project is to actually be able to create text and images with this technique. The technique is nothing new, there are kits available that one can build that do this sort of thing. Stay tuned for a future post for my take on it..

Weird flower forms

These images were made by running Keith's image stacker over a series of images generated by a simulation I developed in Java.

In this simulation, each pixel in an image is assigned to a particle, and placed at the appropriate location on screen to recreate the image. The image is then stirred, in four places.

left-click this image to see a larger version.



I started with an 'original' image. Here we see Justy's cranium in full effect.

This image was idly made one day with my graphics tablet, and represents a thirst for knowledge- or at the very least a search for more screen real estate.




Each pixel was converted to a particle, and then 'stirred' in four areas:

You can see how the image 'tears'.

(to avoid this I would need to supersample the image, creating more than one particle per original pixel.)

I then ran Keith's Image Stacker run over all of the images, tears and all.





Here's another one from a different stir pattern:

The same starting image was used.

left-click this image to see a larger version.




.. and one with less stirring applied:



left-click this image to see a larger version.






The same 'less stirring' idea applied to the second stir pattern:



left-click this image to see a larger version.

Moonlit Skating

In the first shot my Skatelight is seen at the bottom of the image.

The Skatelight is a 12V LED strip powered by a small 12V battery.





All of these images were taken over 60 seconds, and in all of the shots I'm holding a small LED torch.

As I skate down the bridge I'm 'windmilling' my left arm, pointing the light back at the camera.





In this shot, the motion of the LED torch is less regular; I was thinking "firefly".

Diffuse inter-reflections can be seen on the concrete to the left.

You can see the atmospheric and light pollution of Sydney to the North.





Similarly you can see reflections on the metallic grille of the bridge to the right.

LED Skating

This photo features a display I developed that uses an RGB LED, two hi brightness white LED's, and two UV LED's.




This image is of a standard brightness 8x7 matrix of LED's that I handmade about 4 years ago.

It's waved in front of the camera for about 15 seconds.





About 2 years ago I developed a display that simulates the action of sand being dropped into piles.

The code is less than 2k; the chip I used is an AVR2313 10MHz microcontroller.

Individual frames of the simulation can be seen moving across the frame. The refresh direction of the display is easily seen also.





The display has a mercury switch that can be used to randomise the display.

In this shot you can see the randomisation taking place.

Skatelight

Here are some extended-exposure shots taken at night.

To find out more, don't forget to check out my Slamless site. Specifically, the Skatelite section


The underside of the Skatelight, showing both lights blazing.

At one point I considered using an accelerometer to detect when I was decelerating and light the red LED's when that happened.

One challenge with electronics plus skateboards is one of robustness.




The lights are individually switchable- here's a shot of just the 'headlights' in action.

Over the 60 second exposure I try to paint a nice picture with light.

In this one it's easy to tell where the end of the line lies.





In this one I tried to do some space-filling, imagining the action of a bubblejet printer.

I've turned the 'headlights' off to keep the image red.





In this shot, both lights were on, and there is enough background illumination to show the driveway more clearly.

Can you work out which way I was traveling?






Digital cameras are more sensitive towards the red end of the electromagnetic spectrum and this fact is easily verified in these mixed light shots.








Each path that I take is like an individual's signature, and I can't help but be reminded of graffiti tags.








This one shows a clear bump in the road.

Look at the path of the three red lights to the lower left of the picture.

Can you see where the lights trace out a little hump?






The three individual LED's are seen clearly in this shot.

The question is, what was I doing on the board to get the arcs of light?