FAQ
TL;DR: Easiest method: use a funnel into a single-file chute and an off-center IR beam-break that gives ~50% output duty; “No moving parts.” [Elektroda, Stephen Van Buskirk, post #21665461]
Why it matters: You can count ping pong balls accurately without cameras or RFID using low-cost sensors and simple mechanics, ideal for hobby and production lines.
Quick-Facts
- Off-center IR emitter–detector (e.g., QEE113/QSE113) produces a ~50% duty pulse per ball; easy to count. [Elektroda, Stephen Van Buskirk, post #21665461]
- Use a pipe slightly larger than the ball plus a solenoid or top gate to guarantee single-file flow. [Elektroda, Mark Harrington, post #21665460]
- Batch counting by weight works if ball mass is consistent across the lot. [Elektroda, Sarah Harris, post #21665456]
- Adding a tiny steel bead inside each ball enables metal/inductive detection on a chute. [Elektroda, JAWED MATEEN, post #21665465]
- RFID was ruled out here as unreliable and expensive for the target use. [Elektroda, Shimi tapiero, post #21665455]
Quick Facts
- Off-center IR emitter–detector (e.g., QEE113/QSE113) produces a ~50% duty pulse per ball; easy to count. [Elektroda, Stephen Van Buskirk, post #21665461]
- Use a pipe slightly larger than the ball plus a solenoid or top gate to guarantee single-file flow. [Elektroda, Mark Harrington, post #21665460]
- Batch counting by weight works if ball mass is consistent across the lot. [Elektroda, Sarah Harris, post #21665456]
- Adding a tiny steel bead inside each ball enables metal/inductive detection on a chute. [Elektroda, JAWED MATEEN, post #21665465]
- RFID was ruled out here as unreliable and expensive for the target use. [Elektroda, Shimi tapiero, post #21665455]
What’s the simplest non-camera way to count ping pong balls?
Feed balls through a funnel into a single-file chute and place an IR LED–phototransistor pair off-center at the exit. Each ball blocks the beam, producing a clean pulse you can count with a small microcontroller or counter IC. No moving parts are required for basic operation. [Elektroda, Stephen Van Buskirk, post #21665461]
How do I prevent double counts when two balls arrive together?
Add a mechanical gate near the inlet, or use a twin-gate: one holds the next ball while the lower gate releases one at a time. Size the pipe just over ball diameter to enforce single-file flow. As one member put it, “ONE BALL ENTERS ONE BALL LEAVES.” [Elektroda, Mark Harrington, post #21665460]
What parts do I need for a beam-break counter?
Use an IR emitter (e.g., QEE113), a phototransistor receiver (e.g., QSE113), a pull-up resistor, and a microcontroller or simple counter/display. Mount the pair off-center so the output toggles during passage, producing a ~50% duty signal per ball. Shield from ambient light with a short, matte-black tube. [Elektroda, Stephen Van Buskirk, post #21665461]
Can I count balls by weight instead of sensors?
Yes. Place a known number of balls on a scale and compute counts from mass if ball mass is consistent. This is best for batch checks, not flow counting. Mechanical vibration and variable ball mass can reduce accuracy. [Elektroda, Sarah Harris, post #21665456]
Is RFID a good idea for counting ping pong balls?
Not for this scenario. The OP tested the concept and found RFID both unreliable and costly. Unless you require identity per ball, use optical or simple mechanical methods for better robustness and price. [Elektroda, Shimi tapiero, post #21665455]
How should I place the sensor to get reliable pulses?
Mount the emitter and detector off the centerline so the ball partially interrupts the beam. This geometry yields a pulse that is on about half the time as the ball exits, which simplifies edge detection in code. [Elektroda, Stephen Van Buskirk, post #21665461]
How fast can such a system count?
Throughput depends on chute slope and gating. With a single-file feed and a gate, you get one pulse per release with clear separation. The off-center placement creates a stable, ~50% duty pulse, improving reliable high-rate counting without moving parts. [Elektroda, Stephen Van Buskirk, post #21665461]
What about using capacitance to count?
You can pass balls between two metal plates and monitor capacitance changes as they transit. This avoids optics but requires careful shielding and calibration. It’s a creative path if you want contactless sensing beyond IR. [Elektroda, Steve Lawson, post #21665463]
Could sound or vibration help with counting?
Yes. One idea is to let balls hit a slanted plate and detect each impact with a microphone, then count the hits. Signal processing must reject echoes and multiple bounces. Protect the sensor from mechanical shocks. [Elektroda, Steve Lawson, post #21665463]
How do I build it? (3-step quick start)
- Funnel balls into a pipe slightly larger than ball diameter; add a top gate for singles.
- Mount an off-center IR emitter–detector across the exit to create a beam-break.
- Feed pulses to a small microcontroller and show counts on a simple BCD/LED display. [Elektroda, Mark Harrington, post #21665460]
What failure modes should I anticipate?
Watch for two balls entering together, false counts from ambient light, and jams at diameter transitions. A gate or twin-gate mitigates doubles; shrouds and alignment mitigate light leaks. Regulate drop speed to maintain spacing. [Elektroda, Mark Harrington, post #21665462]
Can I embed something inside the balls to improve detection?
Yes. Insert a small steel ball bearing inside each ball, then detect them with a metal detector or inductive coil as they pass a checkpoint. This enables counting without optics. [Elektroda, JAWED MATEEN, post #21665465]
Do I need moving parts?
Not necessarily. An off-center IR beam-break with a funnel provides clean pulses with no moving parts. Add gates only if you see double-feeds or need stricter control. “No moving parts.” [Elektroda, Stephen Van Buskirk, post #21665461]
How should I display the count to users?
Use a microcontroller (e.g., small 8-bit MCU) driving a BCD or LED display for readability. Trigger a buzzer or light when a target count is reached. Keep wiring short and shielded near sensors. [Elektroda, Mark Harrington, post #21665460]
