I have been thinking about this idea for a couple weeks and I would like some help with it. I pretty much sketched out what it would look like, and how it would work. Now for the questions:
1)What kind of fans are compatible with arduinos that can have its speed adjusted?(If the fan is powerful, that would be a bonus)
2)What kind of weight scale can I use to link to the arduino and determine the fans speed?
3)Would there be problems with keeping it in place?
(deleted)
First off, this is only going to work if you have a fairly uninterrupted flow of air; surrounding the fan with a closed "capsule" is probably not going to work well. A tube that was open on top would probably be a better option.
Secondly, your "objects" are going to have to be either spherical, or have some kind of oblate-spheroid shape (with one of the long ends weighted) in order for it to be stable within the airstream.
Thirdly, I am not sure that there is necessarily a linear correlation between the weight of an object (such as a sphere) and the airflow speed; the proper speed might have several variables attached to it - temperature of airflow, amount of turbulance in airflow, material of object being floated in the air column, etc. But - perhaps this is part of the project that you haven't explained (it would make a good science fair project, certainly).
If you find that you do need to use a mains voltage fan, you could vary its speed using a TRIAC, but it may be safer and cheaper to purchase a commercial fan speed controller; these also use a TRIAC, but have a potentiometer to vary the gate interruptions, and are designed for mains voltage. Then, control the rotation of the potentiometer with a servo.
Finally - while you haven't stated it anywhere, please don't claim anything about the "Bernoulli Principle" or "Effect" being the cause or reason behind the object floating in the airstream; this explanation is currently discounted by current scientific understanding (see also references 62-68):
Good luck!
Dear cr0sh,
Is there anyway we could correlate the weight scale to the fan speed(weight scale would substitute the potentiometer).
Interesting project....clever! I'm not sure that weight/mass alone is enough to determine rate of upward force needed to maintain lift.
Dear 1ChicagoDave,
I am pretty sure you are right. I am trying to come up with some sort of algorithm that would solve this problem. In the algorithm I would have to take the weight and probably determine the velocity of a certain sized fan from the weight to keep the object in the air.
The factor which you haven't accounted for is surface area, unless all your objects will all be the same size and shape. The airflow needed to keep a sheet of paper afloat is far less than if the same sheet of paper is folded 8 times.
brandong512:
Dear 1ChicagoDave,
I am pretty sure you are right. I am trying to come up with some sort of algorithm that would solve this problem. In the algorithm I would have to take the weight and probably determine the velocity of a certain sized fan from the weight to keep the object in the air.
Got it.
As Henry_Best pointed out, you'll likely need to limit your algorithm to similarly shaped items.
If you're planning to keep this project rather small in size (at least at first), most/many computer fans have built in speed control using PWM on a separate wire (separate from power & ground) much like a servo. And some of those things can move quite a bit of air...especially when confined within a tube.
Good luck!
This would be far more practical if you used position sensing to control the fan speed rather than just using a fixed pre-calculated fan speed, but for a feasibility test you could simply control the speed manually to see whether you have any hope of getting things to stay in place. It'll also be very sensitive to the size and shape of the object versus the weight, so some things will be much easier to keep supported than others. Ping pong (table tennis) balls seem to be a popular starting point for this type of thing. I suspect you'll find it's easier to get working if you get rid of the plastic cover.
Henry_Best,
Thanks for reminding me about that. I totally forgot about the surface area. Anyways, I am going to limit the objects shapes inside this dome, to rectangular like shapes(ex: popsicle stick). If anybody has this algorithm to keep an object in the air, with a certain surface area, feel free to share it.
PeterH,
I have also been thinking of that and came across this video.Ball and Plate PID control with 6 DOF Stewart platform - YouTube I would just have trouble integrating this with my project but if you know how to, feel free to reply.
1ChicagoDave:
brandong512:
Dear 1ChicagoDave,
I am pretty sure you are right. I am trying to come up with some sort of algorithm that would solve this problem. In the algorithm I would have to take the weight and probably determine the velocity of a certain sized fan from the weight to keep the object in the air.Got it.
As Henry_Best pointed out, you'll likely need to limit your algorithm to similarly shaped items.
And similarly sized.
If you're planning to keep this project rather small in size (at least at first), most/many computer fans have built in speed control using PWM on a separate wire (separate from power & ground) much like a servo. And some of those things can move quite a bit of air...especially when confined within a tube.
Good luck!
As 1ChicagoDave pointed out, wouldn't it be simpler to just measure the height of the object in the tube and adjust the fan accordingly. Too high - slow the fan. Too low - speed up the fan. There would then be no need for any weighing.
....just measure the height of the object in the tube and adjust the fan accordingly. Too high - slow the fan. Too low - speed up the fan. There would then be no need for any weighing.
Like an air-powered version of the old magnetic levitation circuit! Yeah...
(**I just finally learned how to embed videos in posts. Please excuse my ambitious overuse for now. 
)
Thank you again Henry_Best. That is actually very smart. We wouldn't need to do anything about weight but just coding the fans to go faster or slower when it is at a certain height(which I will have to find something to measure height with).
1ChicagoDave, I actually was trying electro magnets for levitation before but not everything is made out of magnets. Thanks for the suggestion anyways.
I was actually just referring to the IR part of that project. It's using the IR emitter & sensors to tell if/when the floating item is too high or too low, and it adjusts the strength of the magnet to compensate and keep it at the correct level. You could use that.
Say, use two IR sensors - an upper & a lower.
• If the upper beam breaks, the fan is too strong causing the ball to float too high. Slow the fan speed.
• If the lower beam breaks, the fan is too weak causing the ball to fall too low. Speed up the fan.
....or some derivation of such.
1ChicagoDave, Sorry for my misinterpretation. That is a smart idea, I was also thinking if I could have a laser pointing at the object from the bottom or the top that would get its distance. The Arduino would then read the distance and make the fans go faster or slower. Also, do you know if there are any specific fans I need or can I just use any fan? Thanks.
There are IR distance sensors, made by Sharp I believe, that Sumorobots use.
"This SHARP distance sensor bounces IR off objects to determine how far away they are. It returns an analog voltage that can be used to determine how close the nearest object is. Comes with 12" long 3-JST interface wire. These sensors are good for short-range detection. For over 1 m distance, we suggest using sonar sensors
To use, connect black wire to ground, red wire to 5V and white wire to analog input. The analog voltage out will range from 3V when an object is only 4" (10 cm) away and 0.4V when the object is 32" (80 cm) away"
The initial premise here was completely off-kilter; it's not just a matter of weight and surface area, it is about density. The airflow to levitate a cannonball and that to levitate a styrofoam ball of precisely the same shape and dimension will differ by many orders of magnitude.
Granted that the objects are not so disparate, it is still impossible to predict from weight. Simply servo-controlling the fan speed according to the height will be the (only) answer, but even that is not so simple and requires a comprehension of control theory. It gets really scary about a third of the way down that page.
While Paul is entirely correct that control theory can get quite frightening in its math to do right, you can do it the wrong way and take much much longer and probably result in similar if not ideal results basically by fiddling with it. I would start with Bang-Bang control (If its too high, turn it down, if its too low, turn it up) suggested above, which you can technically do with the laser/IR distance finder (Find 2 thresholds), but horizontal gates might be easier, though this control will result in some awkward bobbing. Just keep in mind that your objects need to be big enough that the range finder or gate will always see it and can't sneak past. This method will be easier to implement relatively properly without knowledge of control theory, though a PI or PID loop is ideal, and while we can point you in the right direction for the basics (See wikipedia page), you are on your own in implementation since we will not be able to figure out your system's response over the internet, as each system is different.
In terms of a DC fan, you will just need to find something that is the correct size for your tube/object (IE a 3" diameter tube should have a about a 3" diameter fan). Make sure it has a PWM input and find the one with the highest flow rate possible. I could be wrong, but you should be able to ignore the back-pressure restrictions as long as you do not block the inlet with anything (Filter, tube, etc) Just leave it with as much open air as possible on the inlet. I'd look at Digikey, Mouser, Newark, or Allied Elec (Or your choice of parts distributor).
There exist MANY examples of "control" systems using Arduino online...code and all. Whether it's maintaining the altitude of an RC quad copter, the temperature of a room, or the moisture level of the soil in someone's garden. No need to reinvent the wheel here.
In fact, this particular application is extremely similar to maintaining altitude of a quad copter. And, I doubt that many users of that technology have a much greater understanding of control theory than "if it 'bounces' too much I need to adjust this...if it reacts too slowly, I adjust that other thing". The main difference here being that instead of the rotors being attached to the object being kept aloft, they'll be separate from it. Both will have a similar mechanism for control — If it's climbing, slow the rotor (fan)...if it's falling, increase rotor speed.
If I were attempting this project, that's where I would start looking. Look at the code and sensors used with DIY Drone type projects. Figure out how they translate the sensor input into control output. Then, 'massage it' into working with your choice of sensors and control (example: IR beams or sonar sensors & PWM fan control). In fact, you could even use the same rotors, motors, and ESCs (electronic speed controllers) that are used with RC 'copters. They even sell custom Arduino-based boards to work with these items. If done right, theoretically, as long as any object to be levitated falls within the minimum & maximum limits of the system...it shouldn't matter much what shape, size, or weight it is. It should still work. Albeit, perhaps with a few presets or moderate adjustments...and the magic words "in theory".
"Arducopter" or "APM" (Ardupilot) are good search terms, and one very well documented form of this type of Arduino control.
Probably not something to be thrown together overnight. But, a neat project which should teach you a lot....and result in a very interesting conversation piece when finished.
Somewhat ironically, if all one wanted to do was float a ball over a fan, all they would need is a ball and a fan. No math or microcontroller needed. Google "ball floating fan". Or walk into a WalMart in the summer — usually at least one beachball floating on top of fan somewhere. 
Paul__B:
The initial premise here was completely off-kilter; it's not just a matter of weight and surface area, it is about density.
A sheet of paper has exactly the same density whether folded or flat, yet a flat sheet will require less airflow to keep it afloat. Also, density is a function of weight and volume. Surface area is a function of shape and volume. Therefore, weight and surface area, assuming a ball shape, are sufficient to determine how fast the airflow needs to be.
The airflow to levitate a cannonball and that to levitate a styrofoam ball of precisely the same shape and dimension will differ by many orders of magnitude.
Now imagine that the styrofoam ball is huge and the same weight as the cannonball. It will still take less airflow to keep it up because of the increased surface area. Even less air flow would be required if the same amount of styrofoam was shaped in a sheet. I.E. Same density but a different, larger, surface area.