It’s much harder than it may appear but for the sake of my kids who so badly wanted to see it finished, I prevailed.
Thanks guys, this was a fun Christmas break activity to do with my three oldest kids--we each made one....
What is better than an acoustic levitator with a laser shining down the path of the nodes? TWO lasers--one on each end. My first laser began to get dim so I replaced it and while I was at it (and while I had ordered a package of 10 lasers, https://www.amazon.com/gp/product/B071FT9HSV/) I figured I'd put one on each end. Simply wire them in parallel to the same 5-volt connection that powers the Arduino on the motor driver. I would suggest that if anyone is going to drill a hole in the center of the printed frame, start with a small hole and work your way up using every drill bit size in-between, otherwise your drill will catch and you won't get a smooth hole. My first hole was rough and I had to hot-glue the laser but the second one friction fit nicely, with a 15/64" hole.
With an upgraded heatsink I've had mine running for 57 consecutive hours (with 4 balls levitating) and the (larger) heatsink is only warm (96F/36C). I can't imagine that another day or two will make any difference but I'll report in if anything changes. The balls haven't fallen down during those 2 days either (which is not always the case as some days I'll come in and they will have fallen--possibly a power flicker, who knows).
I bought the kit from Makerfabs, which came with the 9V power supply and I didn't make any adjustments but I didn't actually pay attention to the peak-to-peak voltage output of my motor driver, since the levitator worked fine and the waveforms looked good when displayed with the oscilloscope probes connected to an extra transducer. I just tried to pull the motor driver output up on my oscilloscope but couldn't get it to sync with the signal and show up clearly. FYI: If you guys want to buy a heatsink that is made for this transistor, SparkFun has one: https://www.sparkfun.com/products/9576
I'll check back on on the long running time now that I have a suitable heatsink. I purchased a 90's model oscilloscope off of Craigslist for only about $100 that worked perfectly to verify my transducers. I first tried the Arduino polarity testing method and it was not 100% successful and I did find one faulty one too, which turned out that it had some adhesive on the cup of the transducer (either hot glue from me or from manufacturing, I'm not sure).
That's exactly what I used too and I sorted through them to pick out the smallest ones as well. You have to wait for the sharpie to dry (and evaporate) otherwise they will be too dense to levitate well.
I found that the heatsink on my motor driver would get rather hot, too hot for my comfortability. Not quite too hot to touch, but hot enough that it was almost uncomfortable to squeeze the heatsink with my fingertips. I took a spare heatsink I had saved from deconstructing an old TV, modified it so it wouldn't touch any of the circuitboard components, tapped a hole in it, applied some thermal compound to the chip, and now the heatsink only gets warm:
It has been my intention to leave mine running constantly on my desk and I've had it run for a week or more at a time now. As I just commented above, I installed a larger heatsink on the motor driver chip as I thought the factory heatsink was getting too hot for my comfort--now it only gets warm.
I added a 5V 5mW laser to mine, which lights up the end ball and has the double purpose of helping you place the first ball too. You can power the laser with the 5V output of the motor driver, in parallel with the Arduino, and since the laser only pulls 1 mA it works just fine.
I built the kit over Christmas break--thank you for making the kit available--it was just not easy to get a low quantity of the transducers at an affordable price without buying 500 and I'd wanted to build one since I saw the first instructable. Sometimes it is difficult to place the first ball. If you drop them in with a tweezer it shoots them off to the side if you don't get the spot correct (and appears that something may not be working correctly)--I found a trick to locating it without the frustration (or confusion)--thread a ball with a fine needle and lower it into place and locate the centerline and nodes and then place the second one with a tweezer...
First of all--good instructable and I appreciate your willingness to receive feedback--I learn so much from others. One additional thing you could do is make the pins look nicer by sanding more off the ends (so the end grain isn't so course).
Asier, thanks for taking the time to respond. What is the radius of your bowl? I don't have software to open your print files. I have another question too- why did you choose a square wave as the sound you use? If I remember correctly, square waves are much more difficult to produce, having many high frequency components in them in order to reconstruct. Why didn't you simply choose a sine wave as the sound file? Thanks again, I AM going to build one of these!
Must the radius of the bowl be exactly the same as yours? Could I simply find a spherical bowl approximately the same size and use that instead? I know the bowl would need to be spherical so that the radius is uniform and the transducers are the same distance away from the focal point.
Very nice, I restored one of these but the switch wouldn't work and I was required to take the switch apart, clean, and reassemble (which was a bit difficult). Check out this link if you need a detailed article on rebuilding an old drill switch: http://www.shareyourrepair.com/2015/01/how-to-fix-switch-and-change-cord-on-thor-model-4199-independent-pneumatic-tool-co.html
You can find shark teeth in Kansas too, here's my collection:
LED Cube 8x8x8