Joburg Token Comparators: Optical Sensor Cleaning Methods
Dirty optical sensors on a Joburg token comparator will bleed your slot floor dry faster than any software exploit, because a contaminated photodiode reads false reflectance values and the validator starts rejecting legitimate tokens while waving through slugs. 😤 The comparator head in these units is not a sealed laboratory instrument—it is a grimy, dust-blasted optical gate sitting inside a cabinet that sees twenty thousand hands a week, and the sensor window is the first component to betray you.
The core failure mode is straightforward: a film of carbon dust, skin oil, and coin residue accumulates on the emitter and receiver lenses, scattering the infrared beam and shifting the acceptance threshold. On a token comparator running dual-wavelength discrimination, that drift means the optical signature of a genuine token no longer matches the stored reference curve, and the unit defaults to reject. You will see operators crank up the sensitivity pot to compensate, which is the worst possible move—now the machine accepts anything with a vaguely metallic glint.
Why Wet Wipes and Compressed Air Are Killing Your Comparators
I have walked into countless Joburg arcades where the "maintenance" routine is a blast of compressed air followed by a baby wipe dragged across the sensor window. That is not cleaning, that is vandalism. Compressed air drives abrasive grit deeper into the optical cavity and scratches the polycarbonate lens, while alcohol-based wipes leave a residue haze that permanently alters the optical sensor transmission curve. 🙅 Once that lens is hazed, no recalibration recovers the original discrimination margin.
The correct method is dry-first, solvent-second, and never abrasive. Start with a soft camel-hair brush to lift loose particulate from the emitter and receiver faces, working in one direction only so you do not grind debris across the window. Follow with a lint-free microfiber swab dampened—not soaked—with isopropyl alcohol at ninety percent concentration, and wipe the lens in a single continuous pass. Let it flash off for a full sixty seconds before you power the comparator back up, because residual solvent vapor will fog the sensor and give you a false reject loop that looks exactly like a calibration fault.
The Reflectance Baseline Trap
Here is the part most technicians miss: after cleaning, the comparator's reflectance baseline has shifted, and if you do not re-teach the reference token set, the unit will run on a stale threshold and start rejecting good tokens within a week. You must run a full reference cycle with three known-good tokens across the entire acceptance bandwidth. Skip this and you have cleaned the lens only to hand the operator a machine that fails on every third coin.
Watch the emitter current draw during that reference cycle. A clean lens on a healthy emitter should pull a stable, narrow current band—any widening or flicker means the photodiode is degrading and the cleaning was cosmetic, not corrective. 😤 That current signature is your early warning, and ignoring it is how a fifty-rand sensor becomes a five-thousand-rand comparator replacement.
Frequency of Service and Environmental Reality
In a high-traffic Joburg venue, the optical window needs attention every fourteen days, not the quarterly interval in the service manual. Dust load in these cabinets is brutal, and the token comparator sits directly in the path of it. Build the cleaning into a fixed fortnightly rotation, log the emitter current reading each time, and replace any lens showing visible scoring immediately—a scratched window will never hold a stable token comparator threshold no matter how often you swab it.
Do not trust a visual inspection alone. A lens that looks clear to the naked eye can still carry a micro-film that shifts infrared transmission by four percent, which is enough to wreck your acceptance rate on worn tokens. Always validate with the reference cycle and the current-draw check before the unit goes back on the floor.