Setting Up the Millikan Oil Drop Experiment
Most people treat this like it is some arcane physics ritual. It is not. You have two metal plates, a way to atomize oil, a light source, a telescope or camera, and a voltage supply. The goal is simple. You suspend a tiny oil droplet between the plates and measure how fast it falls when gravity is the only force, then how fast it moves when you apply an electric field. From those two velocities you can calculate the charge on the droplet. The first thing that goes wrong is almost always the atomizer. You press the bulb and expect a fine mist. What you usually get is a handful of giant drops and a cloud that takes forty seconds to clear. I spent two afternoons chasing this before I realized the nozzle was clogged with old oil residue. A quick soak in acetone and a fresh squirt of stopwatch-clean synthetic clock oil fixed it. Use synthetic oil. Mineral oil evaporates too slowly and leaves gummy deposits that ruin your nozzle within a week.
Running the Millikan Oil Drop Experiment Correctly
Here is how the actual measurement works. Turn off the voltage. Watch a single droplet fall through the eyepiece crosshairs. Time how long it takes to travel between the two marked lines. That gives you the terminal velocity under gravity alone. Then turn the voltage on and adjust it until the droplet hovers or rises at a steady rate. Time that fall or rise too. With those two velocities, you apply Stokes' law to find the radius of the droplet, then use the balance of gravitational and electrical forces to extract the charge. The formula looks clean on paper. In practice you are working with droplets around 0.5 to 2 micrometers in radius, and Stokes' law starts to drift at that scale. You need the Cunningham correction factor, which accounts for the fact that air is not a continuous medium at those sizes. Without it your charge values will be systematically off by roughly three to five percent. Look up the correction factor table and apply it to every measurement. It takes fifteen extra seconds per drop and saves you from publishing garbage data. I once spent an entire lab session chasing what I thought was a calibration error. My charges kept coming out as multiples of 1.6 times ten to the minus eight coulombs instead of 1.6 times ten to the minus nineteen. I had forgotten to convert millimeters to meters in the velocity calculation. The numbers looked beautiful. They were wrong by exactly a billion. Triple check your unit conversions before you trust any result.
Another thing nobody warns you about: Brownian motion. Your droplet is never perfectly still. Even when the voltage is dialed in to hold it, it jitters. If you time a hover over five seconds and the droplet drifts up and down by a full graduation mark, your uncertainty is going to be huge. The workaround is to measure rise and fall over multiple intervals and average them. Take at least twenty drops. Discard any that evaporate noticeably during the trial. A shrinking droplet changes its terminal velocity mid-measurement and ruins the calculation. The voltage supply needs to be stable. I have seen cheap bench supplies ripple enough to make a droplet oscillate vertically. That makes timing impossible. A battery-powered supply or a well-filtered DC source works better. If you are using an function generator setup, add a large capacitor in parallel to smooth the output. Five hundred microfarads across the plate supply is usually enough to kill the ripple without affecting your ability to adjust voltage quickly. Environmental factors matter more than you would think. Air currents from HVAC vents, people walking past the bench, even the heat from the illumination lamp can create convection currents that push droplets sideways or change their fall rates. Shield the apparatus. Turn off nearby fans. Let the lamp warm up for ten minutes before you start so the thermal gradients stabilize. Your data quality improves noticeably once the setup stops drifting.
Get the Full Details
There is a limit to what this method can tell you. You cannot measure charges smaller than a single elementary charge because the droplet simply will not hold just one electron stably in most conditions. You also cannot reliably distinguish between charges that differ by less than about one elementary charge on larger droplets. The uncertainty in the radius calculation dominates. This is why Millikan focused on small droplets and many trials. A single drop is not enough. The statistics come from the distribution. If you need to measure individual electron charges directly without this indirect method, modern alternatives like the photoelectric effect setup or electron beam deflection tubes give cleaner results for introductory labs. The oil drop experiment is historically important and teaches you about experimental technique, error analysis, and the patience required for precision work. It is not the most efficient way to demonstrate quantization of charge if that is your only goal. The original Millikan setup used x-ray ionization to change the charge on droplets mid-experiment. You can approximate this with a small radioactive source like a weak americium-241 strip, but most teaching labs skip that and just rely on natural ionization from the atmosphere or a UV lamp. The tradeoff is that you get fewer charge-changing events, which means more drops at a single charge state and slightly less elegant data. It is fine for a classroom demonstration.
Common Pitfalls and Practical Notes
Drop size selection is the bottleneck. Too large and the charge is so big relative to the mass that the electric field needed is tiny and hard to control precisely. Too small and Brownian motion swamps your timing. Drops in the one micrometer range are the sweet spot. If your atomizer is producing mostly larger drops, dilute the oil. One part oil to ten parts solvent like isopropyl alcohol helps break the surface tension and produces finer mist. The solvent evaporates quickly and leaves clean oil droplets behind. Timing method matters. Manual stopwatch timing introduces reaction time of about 0.2 seconds per trigger. For a drop falling in three seconds, that is a seven percent uncertainty. Using a photogate or a camera with frame-by-frame playback cuts that down to under one percent. If you are doing this in a teaching lab without video equipment, practice your stopwatch technique until your reaction time is consistent. You can calibrate it by timing a pendulum with a known period and adjusting your trigger offset. Plate parallelism is easy to ignore and hard to fix once your data is already bad. If the plates are not perfectly parallel, the electric field is non-uniform and the droplet will drift sideways as it rises or falls. This creates an apparent change in velocity that has nothing to do with charge. Check parallelism with a feeler gauge at multiple points. Adjust the mounting screws until the gap is uniform within 0.1 millimeters across the plate area you are using. I have thrown away entire data sets because I did not check this and blamed my calculations instead.
The measured value of the elementary charge from this experiment typically lands between 1.5 and 1.7 times ten to the minus nineteen coulombs in undergraduate labs. The accepted value is 1.602 times ten to the minus nineteen. If your average is outside that range, you have a systematic error. Air pressure affects the correction factor. Room temperature affects the viscosity of air. Altitude affects both. Record your ambient conditions and use standard air viscosity tables adjusted for your temperature and pressure. A five degree Celsius error in temperature reading can shift your result by nearly one percent. For anyone actually building this, the plates are usually aluminum or brass circles about five centimeters in diameter spaced two centimeters apart. The voltage ranges from zero to five hundred volts depending on droplet size. The illumination comes from a narrow beam, often from a projector or laser pointer with a cylindrical lens to create a sheet of light. The observation is done through a telescope or a smartphone camera mounted at the side. The whole apparatus fits on a standard optics bench. Downloadable resources for this are sparse because most universities treat it as a custom lab. Your best bet is the American Journal of Physics archives, which have several papers detailing improved versions and error analyses. The Halliday Resnick Walker textbook appendix also has a complete derivation if you need the full mathematical treatment. Online simulations exist but they cannot replicate the frustration of actually getting a droplet to hover, which is where most of the learning happens.
