A chip that keeps time, gates that decide, and a digit the room can read.
Kept under the counter by Ray Kessler — ohmsville.com/blog/meet-ray-kessler
Chips. A 555 that keeps time better than any pair of transistors will, gates that decide, a counter that remembers how many and a display that says so. These are the builds that used to fill the back half of a 200-in-1 manual, and they are the ones that look most like the inside of something you own.
Tap the button and the lamp comes on and stays on for about a second, then goes out by itself. Tapping again during the second changes nothing.
9 V battery, NE555 from the drawer, 100 kΩ, 10 µF from the drawer, 1 kΩ, pushbutton, 330 Ω, red LED
Wired this way the 555 waits. Pressing the button lets the capacitor start filling through the 100 kΩ, and the output stays high until the capacitor reaches two thirds of the supply, which takes about 1.1 times R times C, around a second here. The length of the pulse is set by those two parts, not by how long you hold the button.
One shot, continued
Open it on the bench: ohmsville.com/bench#a=one-shot
Learn more: Classroom, Capacitors and time, paragraph 2 — ohmsville.com/classroom#capacitors-and-time-2
The lamp is lit at rest and goes dark while you hold the button. That is NOT.
9 V battery, NAND gate from the drawer, pushbutton, 10 kΩ, 330 Ω, red LED
A NAND gate answers no only when both of its inputs say yes. Tie its two inputs together and there is only one input left, so its answer is always the opposite of what it is given. That is NOT, and it is the smallest useful thing a gate can do.
NOT, from a NAND, continued
Open it on the bench: ohmsville.com/bench#a=inverter
Tap the button and the digit goes up by one: 0, 1, 2 … 9, and then round to 0 again.
9 V battery, decade counter (up) from the drawer, 7-segment display, pushbutton, 10 kΩ
The counter adds one to its number each time its clock pin goes from low to high, and your finger on the button is the clock. The four wires to the display carry that number in binary, and the display turns it into a digit. After 9 it has nowhere to go but back to 0.
Count by hand, continued
Open it on the bench: ohmsville.com/bench#a=count-by-hand
Hold the button and the digit blurs through 0 to 5; let go, and whatever it lands on is your roll.
9 V battery, 555 timer from the drawer, 0.01 µF capacitor from the drawer, 4 × NAND gate from the drawer, decade counter (up) from the drawer, 7-segment display, pushbutton, 1 kΩ, 2 × 10 kΩ, 330 Ω, red LED
The electric die, continued
A 555 timer free-runs on its own at about seven thousand cycles a second, and an AND gate only passes that clock to the counter while the button is held, so the digit races while you hold it and stops the moment you let go. A second AND gate watches for the count of six and resets the counter to zero the instant it appears, so the display only ever shows 0 through 5 — a true 1-to-6 die needs a counter that can be preset partway through its count, which this one cannot do. Nothing here is random: the timer and the counter are both exactly predictable, and the only unpredictable part is the moment your thumb lets go.
The electric die, continued
Open it on the bench: ohmsville.com/bench#a=the-electric-die
Two identical circuits, two buttons: tap yours as fast as you can, first digit to show a nine wins.
9 V battery, pushbutton, pushbutton from the drawer, 2 × decade counter (up) from the drawer, 2 × 7-segment display, 2 × 10 kΩ
Each pushbutton pulls its own counter's clock pin high, and a resistor pulls it back down when the button lets go — the counter adds one every time that pin rises from low to high. A fast or worn tap can bounce the contacts and count twice, while holding the button down only ever produces one edge, not a run of them, because a counter counts rising edges, not seconds held. First player whose digit reaches nine wins, though nothing here latches at nine — a tenth press wraps either digit back to zero, so the win is called by the players, not the circuit.
Race to nine, continued
Open it on the bench: ohmsville.com/bench#a=race-to-nine
Two buttons pulling opposite ways: yours climbs up from zero, theirs counts down from nine. First one to reach their end — nine for you, zero for them — wins, and both buttons go dead on the spot.
9 V battery, pushbutton, pushbutton from the drawer, decade counter (up) from the drawer, decade counter (down) from the drawer, 2 × 7-segment display, 2 × 10 kΩ, 4 × NAND gate from the drawer, 2 × NOR gate from the drawer, relay and 3 × 1N4001 from the drawer
Tug of war, continued
Each decode's own output drives the relay coil through its own diode, no transistor between them. A CMOS gate's output is 250 Ω, the coil is about 500 Ω, and 9 V across that clears the 10 mA pull-in easily — this solver measures roughly 11 mA whichever decode fires alone, with the other decode's diode blocking its low output from robbing that current. Once the coil pulls in, its contact cuts power to both buttons, so whichever count won — A's nine or B's zero — is the last thing either counter shows. That is the difference from race-to-nine: there the players called the win; here the circuit enforces it from either end of the rope.
Tug of war, continued
Open it on the bench: ohmsville.com/bench#a=tug-of-war
Hold the button to wind the wheel up, let go, and the digit races, slows down on its own, and settles — or press the stop button and call the number yourself.
9 V battery, pushbutton, pushbutton from the drawer, 555 timer from the drawer, 470 µF capacitor from the drawer, 10 µF capacitor from the drawer, decade counter (up) from the drawer, 7-segment display, 100 Ω, 1 kΩ, 10 kΩ
The fairground wheel, continued
The 555's timing string is fed from a 470 µF tank capacitor instead of the battery, while the chip's own supply pin — and so its three-volt and six-volt thresholds — stays on the battery. Charge the tank and let go: as it drains through the 10 kΩ and 1 kΩ, each charge starts from a lower supply while the six volts it must reach stays put, so every step takes longer than the last. The first measures 86 ms, the thirty-eighth 465 ms, and by six seconds the tank cannot reach the threshold at all — the wheel stops and holds. Feed that same string from the battery and the period is flat at 83 ms forever.
The fairground wheel, continued
Open it on the bench: ohmsville.com/bench#a=roulette
Top row and bottom row lit, middle row dark: eight of the twelve LEDs on, by wiring alone.
9 V battery, LED matrix 3×4 from the drawer, 4 × 330 Ω from the drawer
Two bars, steady, continued
A matrix LED lights wherever a driven row crosses a grounded column: current has a path, anode to cathode, through that junction’s diode and nowhere else. Tie the top and bottom rows to the battery and all four columns to ground through their own resistor, and every junction those rows cross lights — both rows solid, the middle row dark, eight of the twelve on. Picking one dot instead of a whole row needs the columns driven one at a time — which is exactly what a real sign this size does, scanning rows fast enough that the eye adds the frames back together, the same 88,200 steps a second this solver runs.
Two bars, steady, continued
Open it on the bench: ohmsville.com/bench#a=matrix-face
Learn more: Classroom, Diodes and LEDs, paragraph 1 — ohmsville.com/classroom#diodes-and-leds-1
Row 0 blinks on and off by itself, a count at a time — off a clock and a counter, no finger on the board — and it comes up dimmer than the row that’s just tied to the battery.
9 V battery, 555 timer from the drawer, 0.01 µF capacitor from the drawer, decade counter (up) from the drawer, LED matrix 3×4 from the drawer, 4 × 330 Ω from the drawer
One row, off one pin, continued
Counting is easy — the counter already ran in the electric die. Lighting a row off one of its bits, instead of off the battery, is the new thing: the bit still lights the row, but it lights it through its own 250 Ω, and that resistor is real even though it never shows up as a part on the board. Half the brightness at the same column resistor is the measurement that proves it, and it is the reason a bigger sign never drives rows straight off the logic — it drives them through a transistor, which is exactly where this chain is headed next.
One row, off one pin, continued
Open it on the bench: ohmsville.com/bench#a=matrix-one-row
Learn more: Classroom, Diodes and LEDs, paragraph 1 — ohmsville.com/classroom#diodes-and-leds-1
Only one row of the matrix lights at a time — row 0, then row 1, then row 2, cycling on its own off the same 555-and-counter clock as the one-row build — and each lit row is brighter than a row driven straight off a counter pin.
9 V battery, 555 timer from the drawer, 0.01 µF capacitor from the drawer, decade counter (up) from the drawer, LED matrix 3×4 from the drawer, NOR gate from the drawer, NAND gate from the drawer, NOR gate from the drawer, 3 × NPN from the drawer, 3 × 1 kΩ from the drawer, 4 × 470 Ω from the drawer
The screen scans itself, continued
A real sign does not keep every row lit all the time — it scans, one row at a time, fast enough that the eye still sees the whole picture. That is the new idea here: matrix-one-row lit a single row off a single bit forever, but a full display needs every row taking a turn, and a turn means picking exactly one row and shutting the other two off, count by count. The follower is what makes each turn worth taking — about 12.2 mA per LED against roughly 5.4 mA read straight off a logic pin, more than double the light for the same three counts.
The screen scans itself, continued
Open it on the bench: ohmsville.com/bench#a=matrix-scan
Learn more: Classroom, Diodes and LEDs, paragraph 1 — ohmsville.com/classroom#diodes-and-leds-1
The same scan, wired exactly the same way, with one part swapped: the 555's timing capacitor is ten times bigger. The counter never speeds up or slows down on its own — the clock feeding it does, and every row's dwell stretches by very nearly that same factor.
9 V battery, 555 timer from the drawer, 0.1 µF capacitor from the drawer, decade counter (up) from the drawer, LED matrix 3×4 from the drawer, NOR gate from the drawer, NAND gate from the drawer, NOR gate from the drawer, 3 × NPN from the drawer, 3 × 1 kΩ from the drawer, 4 × 470 Ω from the drawer
Slow the scan down until it comes apart, continued
matrix-scan proved that scanning beats a row wired straight to a counter pin. matrix-flicker asks the other question: what sets how fast the scan has to run? Swap in a capacitor ten times larger and nothing about the row-select or reset logic changes — only the 555's own clock slows down, and the row dwell follows it almost exactly. That is the whole idea of a refresh rate: it lives entirely in the clock, and a real sign has to keep that clock above about 150 Hz (50 refreshes a second across three rows) or the scanning starts to show.
Slow the scan down until it comes apart, continued
Open it on the bench: ohmsville.com/bench#a=matrix-flicker
Learn more: Classroom, Diodes and LEDs, paragraph 1 — ohmsville.com/classroom#diodes-and-leds-1
The same three-row scan as build 3, still stepping row 0 → row 1 → row 2 on its own — but the picture itself no longer holds still: two of the four columns wink dark and relight as the counter climbs, so what looks like one scan is really three different frames taking turns, four counts, four counts, then two.
Unhook the reset, and one clock draws two pictures, continued
9 V battery, 555 timer from the drawer, 0.01 µF capacitor from the drawer, decade counter (up) from the drawer, LED matrix 3×4 from the drawer, NOR gate from the drawer, NAND gate from the drawer, NOR gate from the drawer, 3 × NPN from the drawer, 3 × 1 kΩ from the drawer, 2 × 470 Ω from the drawer, 2 × 220 Ω from the drawer
A scanned sign does not have to show one picture forever — the same rows, the same clock, and two bits that were sitting idle are enough to carry a second signal riding on top of the first, one frame at a time. It costs exactly one wire: pull the reset and the counter free-runs through all ten counts instead of stopping at three, so the frames arrive on the counter's own schedule — four counts, four counts, then a short two — not a designer's.
Unhook the reset, and one clock draws two pictures, continued
Open it on the bench: ohmsville.com/bench#a=matrix-frames
Unhook the reset, and one clock draws two pictures, continued
Learn more: Classroom, Diodes and LEDs, paragraph 1 — ohmsville.com/classroom#diodes-and-leds-1
A yellow LED that never sits still: two clocks running at once, and the brightness is their sum.
9 V battery, 2 × NE555 from the drawer, 10 kΩ, 100 kΩ, 4.7 kΩ and 22 kΩ from the drawer, 2 × 10 µF from the drawer, 2 × 330 Ω, yellow LED from the drawer
Flickering candle, continued
Two 555 timers run as two separate clocks, one blinking about four fifths of a second and the other about three times a second, and both feed the same LED through their own resistor. Off when both are low, half lit when one is high, full when both are. Because the two rates do not divide into each other, the pattern takes about five seconds to come round again, which is long enough to look like a flame.
Flickering candle, continued
Open it on the bench: ohmsville.com/bench#a=flickering-candle
Learn more: Classroom, Oscillators, paragraph 3 — ohmsville.com/classroom#oscillators-3
Two red LEDs blinking the opposite way round: when one is lit the other is out, about three times every two seconds.
9 V battery, NE555 from the drawer, 10 kΩ, 47 kΩ from the drawer, 10 µF from the drawer, 2 × 330 Ω, red LED, red LED from the drawer
An LED only lights when current goes through it the right way, so one hangs from the supply down to the chip's output and the other from the output down to ground. When the output is low the first has a path and the second does not; when it goes high they swap. One pin, two eyes, never lit together.
Eyes in the bushes, continued
Open it on the bench: ohmsville.com/bench#a=eyes-in-the-bushes
Learn more: Classroom, Diodes and LEDs, paragraph 1 — ohmsville.com/classroom#diodes-and-leds-1
A rattling burst of flashes for about three quarters of a second, then nothing for about the same, over and over.
9 V battery, 2 × NE555 from the drawer, 10 kΩ, 100 kΩ, 2 × 4.7 kΩ from the drawer, 2 × 10 µF from the drawer, 330 Ω, 1 kΩ, red LED, speaker
The lamp sits between two chip outputs rather than between an output and a rail, so it can only light when the two differ: the fast chip high and the slow chip low. The slow chip switches the fast chip's flicker on and off, which is a burst and then a pause. Voltage is a difference, and a lamp only lights across one.
Lightning, continued
Open it on the bench: ohmsville.com/bench#a=lightning
Learn more: Classroom, What electricity is, paragraph 3 — ohmsville.com/classroom#what-electricity-is-3
Hold the button to wind it up. Let go and the tone groans downward for ten seconds, slower and slower, until the door is shut.
9 V battery, pushbutton, 470 µF from the drawer, 22 kΩ from the drawer, NE555 from the drawer, 10 kΩ, 0.1 µF, 330 Ω, speaker
The 555 always trips at the same two voltages, a third and two thirds of the supply, but here the timing capacitor is fed from the 470 µF instead of the battery. As that big capacitor empties, each refill takes longer, so the pitch falls and falls. When it can no longer reach two thirds, the chip never trips again and the door is shut.
Creaky door, continued
Open it on the bench: ohmsville.com/bench#a=creaky-door
Learn more: Classroom, Capacitors and time, paragraph 2 — ohmsville.com/classroom#capacitors-and-time-2
Keep a light on the cell and the shop is quiet. Break the beam — slide the light toward the moon, or cover the camera — and it screams.
9 V battery, 100 kΩ, CdS cell, Q1, Q2, relay and 1N4001 from the drawer, NE555 from the drawer, 1 kΩ, 10 kΩ, 0.1 µF, 330 Ω, speaker
The photocell and the 100 kΩ make a divider, two transistors amplify what the middle does, and the relay pulls in when the cell goes dark. The relay's contact does not carry the sound; it carries the 555's supply. No light means the chip has power and screams; light back means the chip goes dead.
Tripwire scream, continued
Open it on the bench: ohmsville.com/bench#a=tripwire-scream
Learn more: Classroom, Sensors, paragraph 5 — ohmsville.com/classroom#sensors-5
Hold the button: the shaft spins to 3,813 turns a minute, but bolted behind a 48:1 gearbox the wheel only turns 79 times a minute — 0.270 metres a second across the bench — drawing 58 mA.
9 V battery, pushbutton, motor with a gearbox and wheel from the drawer
One wheel turns, continued
A bare shaft settles at 20 mA because there is almost nothing pushing back on it. Bolt on the gearbox and wheel and the gearbox becomes the load: its friction alone costs 36 mA on top of the 20 the motor always draws, which is the 58 mA the meter reads. The car's own weight barely registers next to that — rolling resistance is twenty times smaller than the gearbox's own friction, so it is the gears, not the road, deciding how hard the motor works. What 58 mA buys is real: seventy-nine turns a minute at the wheel, a fraction of the shaft's 3,813, but forty-eight times the twist behind each one.
Open it on the bench: ohmsville.com/bench#a=one-wheel-turns
Learn more: Classroom, The loop, paragraph 2 — ohmsville.com/classroom#the-loop-2
Flip the slide switch the moment you wire it up and the wheel reverses clean — 3,809 rpm one way, 3,809 rpm the other, no drama. Let it spin a while first and then flip, and the motor fights its own back-EMF instead: 947 mA, nearly twice the 510 mA that stalls it, and it burns in about 12 ms. Same switch, same wiring — timing is the only difference.
9 V battery, slide switch, 10 kΩ resistor, NPN transistor, two relays, motor with a gearbox and wheel from the drawer
Slide switch ON.
Reverse with two relays, continued
A stalled motor pulls 9/(17.14+0.5) = 0.51 A, five times the 2N2222's 0.1 A rating, so transistors decide and the relays' contacts carry it instead — each coil draws just 18 mA at 9 V. Reversing supply polarity while the motor still spins adds its back-EMF to the battery instead of opposing it, driving current toward (9+E)/17.14 — 947 mA here, nearly twice the 510 mA stall figure — until it slows to a stop, a real RF-500 trap called plugging. This switch does not wait for you: flip it from rest and there is no back-EMF yet to fight; flip it at speed and the next reversal burns through it.
Reverse with two relays, continued
Open it on the bench: ohmsville.com/bench#a=reverse-with-two-relays
Learn more: Classroom, Transistors as switches and amplifiers, paragraph 4 — ohmsville.com/classroom#transistors-4
Cover the photocell and the wheel spins: the relay pulls in once the light drops to about 76 lux. Uncover it and the wheel keeps going until light climbs back to about 212 lux before the relay lets go.
9 V battery, 47 kΩ, 22 kΩ, CdS cell, Q1, relay and 1N4001 from the drawer, motor with a gearbox and wheel from the drawer
One eye, one wheel, continued
The photocell divides against a 69 kΩ base resistor — 47 kΩ and 22 kΩ in series, since nothing in the box lands between 47 kΩ and 100 kΩ alone. At the bench's 250 lux default it reads ~10 kΩ, Q1 off, coil draws ~5 mA, under its 10 mA pull-in. Covered, at 8 lux it's ~157 kΩ; the coil crosses pull-in at ~76 lux, drawing ~11.6 mA in full dark — under the 2N2222's 100 mA rating but not enough to turn a wheel itself, so the relay's contact, not the transistor, ties +9 V to the motor. Light back up, the coil doesn't let go at 76 lux again — it takes ~212 lux, needing only 4 mA to stay in versus 10 mA to pull in.
One eye, one wheel, continued
Open it on the bench: ohmsville.com/bench#a=one-eye-one-wheel
Learn more: Classroom, Sensors, paragraph 5 — ohmsville.com/classroom#sensors-5
Set it on the track and it steers itself: whichever eye is over the black line keeps that side's wheel turning, so the moment one side drifts off, that wheel stops and the car swings back onto the line.
9 V battery, 47 kΩ ×2, 22 kΩ ×2, two CdS cells, Q1, Q2, two relays and two 1N4001s from the drawer, two motors with a gearbox and wheel from the drawer
The line follower, continued
Each channel is the same 69 kΩ divider as one eye, one wheel — 47 kΩ and 22 kΩ in series — reading ~10 kΩ off the line at the bench's ~250 lux default and ~157 kΩ on it, at the 8 lux the line reads. That's the same dark figure that crossed the ~76 lux pull-in there, so each wheel pulls in solidly on the line and lets go at the same ~76 lux pull-in / ~212 lux drop-out. Wire an eye to the far relay instead of its own and the correction reverses: lose the line under the left eye and it's the right wheel that keeps turning while the left one — inside the turn — stops too, swinging the car further off the line instead of back onto it.
The line follower, continued
Open it on the bench: ohmsville.com/bench#a=the-line-follower
Learn more: Classroom, Sensors, paragraph 5 — ohmsville.com/classroom#sensors-5
Turn the speaker on and drag the heat: 207 Hz at freezing, 663 Hz in the room, 1 667 Hz on a 50 °C day.
thermistor from the drawer, 1 kΩ resistor, 555 timer from the drawer, speaker
The thermistor stands in for the 555's timing resistor, the way the photocell stands in for one in the light-controlled tone. Hotter means fewer ohms, so the capacitor refills faster and the pitch climbs; colder means more ohms and a lower note. The circuit reports temperature as a sound.
The siren that knows it is hot, continued
Open it on the bench: ohmsville.com/bench#a=heat-siren
Learn more: Classroom, Sensors, paragraph 3 — ohmsville.com/classroom#sensors-3
Hold the plates and it hums at about 70 Hz. Slide toward damp — nervous hands — and the hum climbs to a 670 Hz whine. Let go and it stops.
touch plates from the drawer, 10 kΩ resistor, 555 timer from the drawer, 0.01 µF capacitor from the drawer, speaker
The plates stand in for the 555's timing resistor, and skin is a resistor whose value depends on how damp it is. Dry hands are close to a megohm and charge the capacitor slowly, a low hum; sweaty hands are nearer a hundred kilohms and charge it fast, a high whine. The circuit measures sweat, not truth.
The lie detector, continued
Open it on the bench: ohmsville.com/bench#a=lie-detector
Learn more: Classroom, Sensors, paragraph 3 — ohmsville.com/classroom#sensors-3
Set the pot near the middle — about 68.4 flashes a second — and the grey ring becomes one horse, standing still.
9 V battery, motor from the drawer, 555 timer from the drawer, 1 µF capacitor from the drawer, 10 kΩ resistor, 10 kΩ pot, 330 Ω resistor, red LED
The first movie, continued
The strobe runs from 48.0 flashes a second at one end of the pot to 129.7 at the other, and the disc turns 68.4 times a second. Set the flashes to 68.4 and every one arrives with the disc back where it started, so the same drawing is lit each time and the eye is handed one steady picture. The lamp cannot do this job — its filament is still glowing when the next flash is due, and it measures a flat dim glow instead of a strobe. The LED can, because it has nothing to cool down.
Open it on the bench: ohmsville.com/bench#a=the-first-movie
The first movie, continued
Learn more: Classroom, Oscillators, paragraph 1 — ohmsville.com/classroom#oscillators-1
Same board, hands on the pot. Turn it down to 65 flashes a second and the horse gallops forward at 35 frames a second; turn it up to 72 and it gallops backwards at 43. The motor never changed speed.
9 V battery, motor from the drawer, 555 timer from the drawer, 1 µF capacitor from the drawer, 10 kΩ resistor, 10 kΩ pot, 330 Ω resistor, red LED
The horse obeys the clock, continued
The disc turns 68.4 times a second whatever the pot does. At 65.5 flashes a second each flash arrives with the disc 2.9 turns-per-second ahead, and twelve frames times 2.9 is 35 frames a second forward; at 72.0 the flash is 3.6 early and the same arithmetic runs backwards at 43. The horse is not being driven by the motor. It is being driven by the difference between two numbers, and the pot owns one of them.
Open it on the bench: ohmsville.com/bench#a=the-horse-obeys-the-clock
The horse obeys the clock, continued
Learn more: Classroom, Oscillators, paragraph 2 — ohmsville.com/classroom#oscillators-2
Throw the slide switch and the wheel goes straight to 3,813 turns a minute. Throw it back and it stops. There is nothing in between.
9 V battery, slide switch, motor with a gearbox and wheel from the drawer
Slide switch ON.
Flat out or nothing, continued
A switch has two states and so does this car. Close the loop and the battery puts its whole nine volts across a motor that settles wherever its own friction says — 3,813 turns a minute at the shaft, 58 mA — and open it and there is no loop at all. To have any speed in between, something in the loop has to take some of the volts away, and a switch cannot: it is either a wire or a gap.
Open it on the bench: ohmsville.com/bench#a=slot-car-rail
Learn more: Classroom, The loop, paragraph 2 — ohmsville.com/classroom#the-loop-2
Slide the rheostat and the wheel follows it: 3,785 turns a minute wide open, 2,440 at halfway, 1,067 with all 100 ohms in. Hover it and the current never moves off 58 mA.
9 V battery, slide switch, 100 Ω rheostat from the drawer, motor with a gearbox and wheel from the drawer
Slide switch ON.
The trigger, continued
A motor takes the current its friction asks for and no more, so putting a hundred ohms in front of it does not throttle the amps — it stays at 58 mA the whole way down the slider. What the rheostat takes is volts, and volts times amps is heat: 0.33 watts in the coil at the bottom of the travel, and 0.88 watts for the moment before the car is moving, when a stopped motor has no push-back and the loop carries 133 mA. That is more than three times what the kit's quarter-watt knob can shed, which is why a throttle is a resistor built to take the heat.
Open it on the bench: ohmsville.com/bench#a=slot-car-throttle
Learn more: Classroom, Resistance and Ohm's law, paragraph 2 — ohmsville.com/classroom#resistance-and-ohms-law-2
Slide the world's light down to a moonlit 3 lux and back up to a bright 2,000: the digit goes up by one. That is one car, one lap.
9 V battery, CdS cell, 10 kΩ resistor, decade counter (up) from the drawer, 7-segment display from the drawer
The start line, continued
The cell is the bottom half of a divider under a 10 kΩ resistor, so its resistance decides the voltage on the clock pin. Lit at 2,000 lux the cell is about 1.9 kΩ and the pin sits near 1.4 V, which a 4000-series chip reads as low. Put a car in the beam and the cell climbs to its dark limit of 300 kΩ, the pin goes to 8.7 V, and that low-to-high edge is the count. Between the two the chip does nothing at all — its input has forty per cent of hysteresis built in, which is what stops a lap counter from counting the flicker of a room light.
The start line, continued
Open it on the bench: ohmsville.com/bench#a=slot-car-lap-counter
Learn more: Classroom, Sensors, paragraph 1 — ohmsville.com/classroom#sensors-1
Two triggers, two cars, one battery. Put lane two on the brakes and lane one does not move a hundredth of a turn a minute. But bring a second car onto the rails at all and lane one goes from 3,785 turns a minute to 3,772.
9 V battery, slide switch, two 100 Ω rheostats from the drawer, two motors with a gearbox and wheel from the drawer
Slide switch ON.
Two lanes, continued
A motor takes the current its own friction asks for, and a throttle spends volts rather than amps, so each lane draws 58 mA whatever its trigger is doing. Two lanes is 115 mA out of the same cell — braked or flat out, the meter on the battery reads the same — and the half ohm of resistance inside the cell turns that into 58 millivolts of sag on the rails. So a second car costs you fourteen turns a minute the moment it joins, and after that the other driver cannot touch you. The race is fair for the same reason the throttle gets hot: what a rheostat takes out of a motor is volts, not current.
Open it on the bench: ohmsville.com/bench#a=slot-car-two-lanes
Two lanes, continued
Learn more: Classroom, The loop, paragraph 4 — ohmsville.com/classroom#the-loop-4
Wire it and wait. Nine and a half minutes of nothing, then the buzzer chirps for a third of a second. After that, one chirp every five and a half minutes, from wherever you hid it.
9 V battery, NE555 from the drawer, 1 MΩ from the drawer, 1 kΩ, 470 µF from the drawer, buzzer from the drawer
The annoy-a-tron, continued
A 555's period is a resistance times a capacitance and nothing else. The capacitor fills through the 1 MΩ and the 1 kΩ in a row, which takes 326 seconds, and empties through the 1 kΩ on its own, which takes a third of a second. The buzzer is wired from the supply to the output, so it sounds only during that short half. Same chip as the pad beacon, same three parts, one resistor a thousand times larger, and a tone becomes a chirp nobody can find.
Open it on the bench: ohmsville.com/bench#a=annoy-a-tron
The annoy-a-tron, continued
Learn more: Classroom, Capacitors and time, paragraph 2 — ohmsville.com/classroom#capacitors-and-time-2
Leave it in a corner and put the light out. Two red eyes start blinking the opposite way round, about three times every two seconds. Put the light back on and they both go out and stay out.
9 V battery, 100 kΩ, CdS cell, Q1, Q2, relay and 1N4001 from the drawer, NE555 from the drawer, 10 kΩ, 47 kΩ from the drawer, 10 µF from the drawer, 2 × 330 Ω, red LED, red LED from the drawer
The fake bug, continued
A photocell is a resistor that reads the room: about 10 kΩ at the bench's 250 lux, and hundreds of thousands of ohms in the dark. The 100 kΩ above it makes a divider, two transistors multiply what its middle does, and the relay pulls in below about 76 lux — and what the relay's contact carries is the chip's supply, so in a lit room both LEDs read zero. In the dark the chip runs, and because the two LEDs hang off its one output the opposite ways round, one is lit exactly when the other is not.
The fake bug, continued
Open it on the bench: ohmsville.com/bench#a=fake-bug
Learn more: Classroom, Sensors, paragraph 5 — ohmsville.com/classroom#sensors-5
Press Set: the LED turns on. Let go — it stays on. Press Reset: the LED turns off. Let go — it stays off. One bit of memory, held by two gates arguing with each other forever.
9 V battery, pushbutton (Set), 330 Ω resistor, red LED, 2 NAND gates from the drawer, 2 10 kΩ resistors from the drawer, pushbutton from the drawer (Reset)
The Loop That Remembers, continued
This is the same trick behind every byte of RAM you have ever used, just two gates instead of millions: an output wired back into an input, so the circuit keeps repeating its last decision to itself forever. Letting go of the button proves nothing is stored anywhere outside the loop — the loop itself is the memory, and it never loses the argument with its own last answer.
Open it on the bench: ohmsville.com/bench#a=heritage-flip-flop
The Loop That Remembers, continued
Learn more: Classroom, The loop, paragraph 1 — ohmsville.com/classroom#the-loop-1
Break the light beam and let it clear again: the display climbs by one. Do it again and it climbs to two.
9 V battery, 7-segment display from the drawer, up counter from the drawer, 10 kΩ pull-up resistor from the drawer
The Beam That Counts Itself, continued
A relay and a counter's clock pin both just want one clean high-low-high swing; what happens after is the only difference. The same amplifier that snapped a relay shut can just as easily step a counter, because neither the sensor nor the two transistors know what they are driving — only that the swing is clean enough to read as a single edge, not a flicker.
The Beam That Counts Itself, continued
Open it on the bench: ohmsville.com/bench#a=heritage-optical-counter
Learn more: Classroom, Sensors, paragraph 5 — ohmsville.com/classroom#sensors-5
Generated from the modules the bench runs on: every circuit in this booklet is the circuit the simulator solves, and every word beside it is the word the manual panel prints.
Built 2026-10-29 from commit 92334ff.
The Heritage builds in this booklet re-express ideas from vintage Radio Shack “Science Fair” kit manuals. Those manuals are the work of Tandy Corporation; the circuits, drawings and words here are ours. Read the originals:
© Ohmsville. The booklets are ours to give away; the kits they are named after are not ours at all.