Science Fair Ideas

This will be a collection of unusual ideas for somewhat difficult science fair projects - hopefully not a repeat of other sources.

• Use colored LEDs and an old phototube to duplicate Millikan's photoelectric experiment to calculate Planck's constant. Watch out, however; most LEDs put out quite a broad spectrum even though they appear monochromatic. You may need additional filters. There are pitfalls galore and discussing them would make for a great write-up. Use a CMOS op-amp to hold the current in the tube near zero by adjusting the voltage across the tube (simple feedback circuit). Read the voltage as a function of light color striking the tube. One possible schematic is fairly simple as seen below. Instead of using positive and negative supplies, a dual, 5 V and 12 V supply is used; either approach is fine. Both supply's' commons are connected to the circuit common (small triangle ground symbols).
  
   
  
  The photo shows a couple of typical phototubes, one plugged into a socket with a banana plug adapter and partially inserted into a PVC plumbing "T" painted black.  My particular tubes have the pickup wire bent in a rectangular shape, allowing maximum light to reach the cathode, but most have a simple center wire as depicted in the schematic. They all suffer from the problem of light hitting the wire, causing undesired emissions or a "reverse current".

  An LED holder made from more plumbing parts and an earphone socket is at the top of the "T". The different colored LEDs are mounted in earphone plugs that plug into the PVC adaptor. The left end of the "T" has a removable plug that allows for inspection of the light hitting the tube. The LEDs are powered by a variable power supply (not shown) and each has a 220 ohm resistor in series. The variable supply allows the brightness to be varied. A current meter in series with the LED supply is a good idea; keep the current below 30 mA.
  
  Filters may be added over each LED to purify the spectrum. Unfortunately, effective filters can be expensive! But, it turns out that colored cellophane can make a big difference. (Packs of colored sheets of cellophane are sold for arts and crafts projects.) Use a few layers of red cellophane for the red and yellow LED (the yellow cellophane doesn't seem to attenuate other colors much). Use a couple of layers of blue for an ultra-violet LED. Adjust the light intensity for the maximum reading, too. The other spectral components will get through but by dimming the LED, the reading will be controlled by the dominant spectral line.  Here are some quick test results with and without cellophane filters:
   

  Color	Red	Yellow	Green	Blue	Ultra-violet
  With Filter	0.19 V	0.65 V	0.7 V	-0.26 V	1.6 V
  No Filter	0.27 V	0.27 V	0.21 V	-0.13 V	0.95 V

  
  There are some surprises in the data. For one, the voltages are low, but the actual behavior of one of these phototubes isn't ideal. A major undesired effect is most pronounced with the blue light. Notice the voltage is negative! Apparently the little collector wire  in the tube emits more electrons than the target cathode at this particular color. At other colors, the wire also emits electrons and this reverse current lowers the voltage readings. The reverse current also lowers the impedance of the tube which is handy; no high-value resistor is needed in the circuit. When the tube is illuminated sufficiently, its impedance drops fairly low. The op-amp is really setting a voltage that balances the forward and reverse currents. Despite the shortcomings of this particular tube, the trend that a higher retarding voltage is needed for a shorter wavelength can be seen, especially with the cellophane filters.

  

  The photo above shows a typical test setup. You can see the LED power supply, the meter, different colored LEDs in banana plug housings, bits of colored cellophane, the phototube housing and the electronics box. The box was made such that it can be used for a variety of projects. The power is supplied by a dual voltage molded supply (not shown) and comes in via the black cable. A panel lamp indicates power but the leads are plugged into the breadboard so that the light may be used for other purposes in the future. A dual banana jack on each end is used for external connections. A solderless breadboard is permanently fastened to the inside lid (see earlier photo). The breadboard is pushed to one side to allow for the addition of switches or other controls on the top in the future. This is a great way to build special projects that might have a limited useful life!

  
  
  One word of caution! Those breadboards are insulated on the bottom by paper and they can exhibit a tiny amount of leakage between pins. Bend pin 2 of the op-amp up into the air and connect the tube (black wire from banana plug in my case) and capacitor directly to the op-amp leg with solder.  The prototype case is sitting on an aluminum sheet with a grounding clip connecting the two. This ground plane helps reduce line voltage related interference. It might be a good idea to connect an aluminum cake pan to the ground and lay it over the whole affair when making measurements. The pan will also help block ambient light. I wrapped electrical tape around the base of the tube so that the plumbing pipe fits snugly.

  It doesn't really matter which way the tube is connected; the meter voltage will simply reverse polarity.

  Search the Internet for "Einstein, Millikan and the Photoelectric Effect" by Richard Keesing. His excellent and interesting paper describes many of the problems with the experiment and would make an excellent reference for a science fair project. You won't be able to measure Planck's Constant to any degree of accuracy with this setup and his paper will explain why.

  
   

• You have probably heard how one's hair can stand up right before a lightning strike due to the electric field. Can that same field lift light objects off the ground such as leaves or perhaps fine dust the way a charged pocket come can lift bits of paper? Are charged clouds attracted to the ground? Look up Millikan's oil drop experiment which used an electric field to move around charged oil droplets. Even if water droplets are attracted to the ground, they will move very slowly through the thick air; read about Stokes Law. Calculate that a force equivalent to 10g might move a typical cloud droplet 10 microns wide at only 10 feet per hour. But what is the typical force from a typical cloud charge? What other previously unnoticed effects might these large static electric fields have? It would make a great display!
  
  
  
  The photo shows a couple of plates made from thick aluminum foil attached to a Wimshurst electrostatic generator. Little "leaves" of thin aluminum foil will fly up with a few cranks of the generator.
   
• Use n-acetyl cysteine (NAC - found in the vitamin section) to preserve the bright red color in some fall leaves. (The Chinese Tallow is a good choice.) Grind up leaves in a little water with and without a pinch of NAC (put it in before grinding), pour the red liquid into test tubes, cap them and watch the difference in how long it takes it to turn brown. The difference should be minutes vs. days or weeks. The leaves should be immersed in water and NAC while grinding. Here is a photo of a typical setup:
  
  
  
  The tube on the right was with the NAC and the difference is apparent after only a few minutes. After two hours the pink NAC tube was even brighter after solids settled out and the untreated liquid was completely dead brown.
  
  Try other antioxidants. Do they work as well? If not, why? What is the effect of temperature? What is the red color, anyway. Some think it is an antioxidant that the leaves produce in a last-gasp effort to stay alive after the tree cuts their food and water supply off! Does the NAC, a powerful anti-oxidant, react with the free radicals before the red substance, thereby preserving the color? Will NAC protect whole leaves? Why or why not? The timing isn't right for the typical science fair so consider duplicating the colors in the tubes with food coloring, simply for display purposes. Good photos should suffice. What else can be preserved? What else oxidizes and spoils as a result?
   

• Design and build an "infrasound" microphone capable of detecting frequencies below 20 Hz. The transducer could be a electret microphone with unusually low frequency response. A couple of web articles recommend the Panasonic WM-034BY for its unusually low frequency response. There are several interesting articles on the web regarding possible designs. Interpreting the data using FFT freeware. Maybe playing the data back speeded up to bring the frequencies into hearing range would be interesting. This is not a particularly easy project! An interesting "far infrasound" microphone may be had by simply amplifying the output of a pressure transducer like the Motorola MPX100A (see http://www.techlib.com/electronics/barometer.html#Electronic%20Barometer).
  
  If you want to try something a bit more difficult, consider making one of these:
  
  
  
  This could be called a "cansformer", I suppose. This low-frequency transducer boosts the pressure wave that a small microphone element sees to give much more sensitivity, especially to low frequency audio. This large cookie tin has a soft plastic diaphragm stretched across the open end with a circular piece of steel sheet cut from the bottom of a peanut can glued to the center to act as a stiffener. The red cookie tin lid has a hole cut in the center to let in the sound. The metal stiffener will move up and down with low-frequency sound waves by an amount that is proportional to the can height (ignoring the stiffness of the diaphragm). By connecting the center of the diaphragm to the center of a smaller diaphragm on a much shorter can, the pressure changes in the smaller can will be a multiple of the pressure changes in the larger can. If the smaller diaphragm moves the same distance as the larger one (and it will if they're mechanically connected), that distance will result in more pressure change since it is a higher percentage of the total height of the can. The larger surface area of the bigger diaphragm provides the extra force needed to move the smaller diaphragm against the higher pressure. This pressure transformation gives more output for a given audio volume and, therefore, better signal-to-noise performance without resorting to banks of microphones.
  
  
  
  The photo above shows the diaphragm peeled back, revealing the smaller mint tin mounted on a strip of wood. (The little microphone off to the side is only there for comparison purposes.) The mint tin also has a diaphragm made from the same soft plastic material (scavenged from a zipper bag for a comforter). Soft rubber would also work (perhaps from a toy tom-tom drum). The smaller diaphragm has a smaller disk cut from another peanut can and the lid of the mint can has a fairly large hole. The metal disks are glued into place facing each other and the mint can support is positioned so that a small, powerful magnet bridges the gap when the large diaphragm is in place. Lay a straightedge across the mouth of the big can to help position the support. The top of the magnet should just touch the straightedge. The magnet mechanically connects the two metal disks and makes it easier to assemble and disassemble the unit. Make sure the disks are made from a magnetic material. Forcing the lids over the plastic stretches the plastic nicely, leaving a smooth, tight diaphragm.
  
  A small hole is drilled into the bottom of the mint can and an electret microphone is mounted over the hole with epoxy such that the microphone hole is exposed to the inside of the can:
  
  
  
  My idea was that the microphone must have some sort of internal, precision pinhole to allow the pressure to slowly equalize across the electret element. This slow leak will relieve pressure differences and keep the mint can at ambient pressure. The leak needs to be slow so as not to attenuate low-frequency response. Now the leak also balances the pressure in the much larger volume of the small can so the time constant must be very long. Once the epoxy cures, the diaphragm is stretched over the can and the lid is squeezed on, stretching the diaphragm to a nice, unwrinkled state. If the hole in the mint lid is big enough, the steel disk may be glued into position after the diaphragm is stretched. A little glue was applied to the underside of the edge of the diaphragm before assembly to help make a good seal.
  
  The larger can could probably stand to have a pressure relief hole, too. But it can be difficult to make a leak that isn't too fast so some experimentation may be required. The prototype happened to have a slow leak so no hole was needed. Pressing in the big diaphragm and then releasing it after a minute results in a slow return to the flat position, perhaps taking 30 seconds. The motion of the large diaphragm can be hard to see, so temporarily glue a pointer stick to the diaphragm and let it rest on the lip of the hole. Slight changes in the height of the diaphragm will cause the end of the stick to move up and down against a scale.
  
  The gain of the prototype is only about five which is lower than expected. The lower gain is probably due to the attenuation caused by the stiffness of the diaphragms, especially the smaller one, and possibly the relative diameters of the cans. The gain and sensitivity were checked by positioning the transducer in front of a large speaker driven by a signal generator set to 15 Hz. The second microphone glued to the wooden support gives a much smaller signal than the one mounted to the mint can, perhaps by a factor of six.
  
  Make sure to seal the connector for the microphone cable. I used potting wax on the inside of the connector but epoxy would also work.
  
  Most sound cards only go down to about 20 Hz so another type of input may be desired. One possibility I am considering is to use the amplified and low-pass filtered microphone signal to drive a voltage to frequency converter with a center frequency well above the frequency to be monitored, perhaps 2 kHz. The audio will FM modulate the 2 kHz and FFT software can see the variation as sidebands around the center frequency.
   

• Ordinary incandescent light bulbs can serve as bolometers and thermometers! Put that refrigerator light to use as a thermometer when the door is closed! An ordinary 60 watt lamp can heat a filament in another clear bulb to 400 C; what can the sun do? Does sunlight heat a filament enough to impact its lifetime when the lamp is off? Despite the intense heat of the lit filament, can one detect the additional temperature added by another bright light or the sun as a decrease in the bulb current? (Hint: A #387 bulb current changes about 0.1% when illuminated by a 100 watt lamp.) How much outdoor incandescent lamp energy is saved by warm weather? Bulbs are everywhere; think of some unusual bolometer applications or measurement opportunities. Can our lamps spy on our movements in our homes? Or watch for flames? Can a traffic signal bulb see car headlights? (Well, before LEDs!)
  
  To see the phenomenon in action, connect a clear night light bulb to an ohmmeter. It should measure around 400 ohms (for a typical 4 watt bulb).  Hold the bulb under an incandescent desk lamp and watch the resistance quickly climb. You can use a night-light socket to hold the bulb but make sure that it is the type with a simple power switch and not the automatic type. Higher wattage bulbs will work but the resistance is much lower and the change is harder to measure. For example, a 50 watt spotlight bulb measures about 25 ohms and only changes about 2 ohms when illuminated by a desk lamp.
  
  The schematic above is a fun illustration of the phenomenon, a lamp flasher that relies on the light from one lamp changing the resistance of the other. The two bulbs are placed close together, but with a small gap so that a piece of paper may be slipped in between to block the light. When the bulb in the collector turns on, it heats the filament of the other bulb, driving the voltage up on the negative op-amp input. When the voltage rises high enough, the op-amp switches the lamp off and the sensor bulb begins to cool, reversing the process. The 10 megohm resistor provides hysteresis so that the switching is quick and complete. The selected resistor is chosen so that the resistor and potentiometer set to the half-way point are the same resistance as the bulb on the left. The pot value isn't critical but the higher it is, the more touchy the adjustment. The bulb will change resistance when bumped so it is a good idea to mount the pot so that an observer can adjust it to get the lamp flashing. Mount a piece of paper or plastic on a slide so that the light can be blocked without bumping the lamps. Simply tapping on the table can upset the critical resistance. The prototype will blink for several minutes before a fine adjustment is required. Any op-amp suitable for 30 VDC operation and most NPN transistors will work. Bulb substitution is more tricky. Choose a high voltage, low current bulb like the #387 to get a reasonably high bulb resistance and adjust the power supply accordingly.
  
  
  
  Here it is, flashing away. Slipping a piece of paper between the bulbs stops the flashing.
   
• Human Sensitivity to Temperature Changes: A thermoelectric heating/cooling module would be characterized and a driver circuit would be developed to create rapid, small temperature changes. Simply applying a step in voltage to the device will result in a fairly slow exponential temperature change but a more complex pulse with a large peak at the beginning could help slew the temperature more rapidly. The test subjects would be tested for response time as a function of step size, direction, pulse shape, or whatever else occurs to you.
   
• Fire Ant Baiter: Design a baiter made from biodegradable materials that only allows tiny ants in and keeps water out - something that could be dropped from planes by the thousands. Properly designed holes would block water via surface tension. The thing should probably float and work equally well upside-down. Maybe shaped like a flying saucer with holes along the edges. It probably isn't practical but the project could be great anyway!
   
• UFO explanations: I have seen some great UFOs but, fortunately for me, I was able to figure out what they actually were! See the article, "UFO's Explained - For the Most Part" for a couple of ideas and research other resolved sightings for other potential demonstration material including illusions. Demonstrate possible explanations to some of the more famous sightings. Perhaps a Saran Wrap "boundary layer" could be stretched across the inside top of a cardboard observation box and various light emitters could be behind a cardboard barrier such that only their reflections can be seen above. Tilting the plastic wrap would give the illusion of movement.
   
• Study electronic components for non-traditional applications. (see bolometers above) For example large power transistors exhibit extremely low voltage noise making them great for ultra-low noise preamplifiers. LEDs exhibit a variety of useful characteristics besides the light including a very low noise forward voltage. Zener diodes exhibit areas of "negative resistance" and will exhibit strange oscillatory behavior - especially at low currents. Many reverse-biased transistors (emitter-collector) exhibit a large negative resistance region and can make interesting oscillators.  Use sliced-open ceramic capacitors to measure humidity. Ferrites have a host of odd characteristics and non-linear behavior.
   
• Ion Chamber experiments can prove fascinating! If you don't have a radiation source for your ion chamber, make a gadget like this:
   

  

  An ordinary computer fan draws air through a piece of dust cloth and highly radioactive radon "daughters" accumulate on the paper. The radiation will last for about two hours after the fan is turned off.  This filter was used in a controversial experiment that shows how one might use the collector in an experiment. I would recommend another theorem since suggesting something positive about second hand smoke might not be popular among judges! Long-term variations in radon levels might be a good project. Can radon daughters be swept up with charged plates or plastic electrets? How quickly does radon build up in a sealed camper's tent? How about a dirt-floor storage shed or barn? Can you pick up radon variations as a function of earth tremors, thunder, barometric pressure change, etc.?

   

• Does a DC current flowing in the soil affect plant growth, seed germination? AC current? How about plant parasites?
   
• Some plant parasites travel via pollinators (spider mites on bees, for example). What strategies might prevent these mites from making it onto the home garden tomato plant? Would "fake" flowers trick a certain number of mites into exiting early when the bees investigate the blooms? Might a sticky substance be placed below blooms prevent the mites from making it to the leaves or do the mites just jump off in midair? Could a fixed, stiff flower allow for the high-speed microscopic photography of the mite's "deplaning" ? Can farmer ants be fooled into "poor farming techniques" while tending their aphid "crop"? Perhaps smells, sounds, lights, or other stimuli might affect the ants' behavior.