Float a ping pong ball at will

[ClockGuy]

If anyone remembers J Collins ping pong levitation system (http://www.instructables.com/id/Float-a-Ping-Pong-Ball-at-Will/) I'm looking for someone who has replicated it. I want to use it as a vertical clock and was wondering if the fan can be controlled in such a way as to reproducibly achieve the same ball height so that the height can be controlled by a clock in order to assign height to time markings along the tube. Would need ball to indicate time to within +/- 30 minutes. Thanks.

[JColvin]

Hi @ClockGuy,

I have moved your question to a more appropriate section of the Forum.

I personally haven't used the floating ping pong ball as a vertical clock though it does sound like an interesting application for it. Realistically, the main thing you would want to get some sort of real time input to the system and then apply some sort of division or scaler (depending on how you are getting your time, whether it is from the current length of time the system board has been run or the current time from a server (like in this thread) and then based off your conversion put the values into the heightControl function. Once the time "rolls over", whether that is at 12 o'clock to 1'clock or 23 hours to 0 hours, you could have it then drop the ball to be at the bottom of the tube.

Let me know if that made sense.

Thanks,
JColvin

[ClockGuy]

Hi J, thanks. Yep, much like what you said. For argument sake let's assume I am working with your identical configuration, i.e., a 4 foot high tube. If the clock tells 12hr time that means there are  roughly 4" along the scale between each hour marking. So as long as the ball settles in to the right height (i.e., doesn't bounce around too much) then maybe it will be no cruder than a sundial. Do you know off hand how much the ball bounces around at any given steady state height? Your video is what got me excited but because the fan frequency s always changing in response to the accelerometer I can't even begin to guess how stable the height would be at a fixed fan on/off frequency. The scale will be established empirically to determine the fan on/off frequency needed to establish each steady state 4" marking (I can't recall if your system provides info on the frequency?) Then I need a clock that will set that frequency corresponding to each hour's marking along the scale. The default position of the ball will be at rest, i.e,. at the bottom of the tube. The fan will be activated only when a human requests the time. The fan will then blow for enough time to settle the ball into a steady state position so that the time can be read in a relaxed manner. Then the fan shuts off until the next request for time. It's going to be a trial and error experiment. To establish proof of principle your "levitator" would only need to be slightly modified, i.e., replace accelerometer with a rheostat (I'm naive about electronics so I pulled that word out of a hat)  that would allow varying the frequency in a step-wise manner in order to get an initial sense of whether it could function as a clock. If it looks promising, but not ideal, I might then dive deeper into this and try altering the performance by playing with the ball size (diameter of tube will be adjusted accordingly), material, etc.. But that's all fluid mechanics or something like that. Cheers. 

[JColvin]

Hi @ClockGuy,

From what I recall, the ball is actually pretty stable after the fan speed is set (+/- 1 inch or so). In terms of the on/off frequency, it looks like I used milliseconds in terms of turning it on and off in a 100 millisecond timeframe. There is an idle timeout through of 6 seconds (same value for 6 seconds in a row) so you would want to remove that functionality from the code though. The rheostat (more commonly called a potentiometer in this field) will work fine as a replacement for the accelerometer.

Thanks,
JColvin

[ClockGuy]