The Coin Acceptor Sometimes Rejects Real Coins But Accepts Washers — What’s Happening?

It was a Tuesday morning in Manila, around 10:30 AM. The operator of a 40-machine arcade in Quezon City had been tracking an unusual pattern across eight of his fishing game cabinets. For the past three weeks, those specific machines were reporting 22% less coin revenue than the adjacent units despite identical foot traffic. He started doing manual coin drops during off-hours — inserting one 5-peso coin at a time and watching the credit register. Three out of ten genuine coins were being rejected outright. But here was the part that caught his attention: he found three metal washers mixed in with the rejected coins in the return tray. These washers were 24mm in diameter, almost exactly the same as a 5-peso coin, and roughly the same weight. Someone had been feeding these into the machine. And based on the wear marks on the washer edges, it had been happening for weeks. The machine that was supposed to reject worthless metal had been giving credits for it while simultaneously refusing real currency. The operator’s revenue loss across those eight machines alone exceeded 75,000 pesos over the three-week period.

What This Problem Looks Like in Practice

The symptoms of coin acceptor tampering typically present as a constellation of small, individually unremarkable anomalies that operators often dismiss one at a time. When you add them together, the pattern becomes unmistakable. The most common indicator is selective rejection: the acceptor refuses valid coins while accepting foreign objects. This selectivity is rarely random. Cheaters calibrate the mechanism to tolerate a broader acceptance window on certain parameters — coin diameter, weight, metallic composition — which creates a side effect of rejecting genuine coins that fall at the edges of the original calibration range.

Operators in Thailand have reported another variant. In a Pattaya arcade, the owner noticed that one specific machine type — older models using the CC-16 comparator-style acceptor — would accept 10-baht coins that had been slightly filed down on the edges. The filing reduced the diameter by approximately 0.3mm, just enough to fall outside the acceptor’s valid range for 10-baht but inside the calibration window that had been widened by the tamperer. The same machine would then accept metal slugs cut from brass tubing that matched the new, broader acceptance parameters. The operator discovered this when he opened the cash box and found seventeen brass rings mixed in with legitimate coins — a loss of roughly 850 baht across a single weekend.

Another symptom is inconsistent acceptance behavior that varies by time of day. When the cheater is present and actively feeding slugs, the machine appears to work normally. But during regular customer hours, when genuine coins are being inserted, the rejection rate climbs noticeably. This time-based pattern is one of the strongest indicators that someone has modified the acceptor’s calibration, because the modified parameters are optimized for the cheater’s slugs, not for genuine currency. One operator in Cebu City tracked rejection rates by hour and found that between 2 PM and 6 PM — when the cheater was known to operate — rejection rates were 3%. Between 6 PM and midnight, when regular customers filled the venue, rejection rates jumped to 18%. That 15-point gap was the direct cost of the tampered calibration.

How Coin Acceptor Tampering Works — The Technical Layer

Electronic coin acceptors used in arcade machines operate on a multi-parameter validation system. When a coin passes through the acceptor channel, it passes through a series of sensor stations that measure specific physical properties. A typical comparator-based acceptor measures at least three parameters: diameter, thickness, and metallic signature. More advanced multi-coin acceptors add weight measurement and sometimes optical pattern recognition. The acceptor’s microcontroller compares each measured value against stored reference values for each coin type it is programmed to accept. If all measured values fall within the acceptable tolerance ranges for a known coin, the acceptor sends a credit pulse to the game board. If any parameter falls outside the range, the coin is routed to the reject channel.

The calibration manipulation attack works by adjusting the tolerance ranges stored in the acceptor’s memory. Every electronic acceptor has a calibration mode — typically accessed through a specific button sequence or DIP switch configuration on the acceptor unit itself. When the cheater gains physical access to the acceptor, they enter calibration mode and feed their chosen slug material through the acceptor repeatedly. The acceptor learns the physical profile of the slug and stores it as a valid coin type. The problem is that this process also expands the overall acceptance window. Once a brass washer is programmed as a valid token, the acceptor’s discrimination threshold widens. Genuine coins with parameters at the edges of the original narrow tolerance range — slightly worn coins, coins with minor manufacturing variations — now get rejected because the system’s ability to distinguish between similar objects has degraded.

A second method involves physical modification of the acceptor channel itself. Some older mechanical-comparator acceptors use a physical gate and magnet system where the coin must have the correct diameter to trip the gate and the correct metallic composition to pass the magnet test. Cheaters can bend the gate arm slightly outward, widening the acceptable diameter range. They can also weaken or reposition the magnet so that non-ferrous slugs pass through without triggering the reject mechanism. I inspected one such acceptor from a Cebu machine where the gate arm had been bent approximately 1.2mm outward using needle-nose pliers — a modification that takes less than 30 seconds with the acceptor removed from the machine. The magnet had been repositioned approximately 4mm further from the coin path, reducing its effective field strength at the coin’s position by roughly 40%. The result was an acceptor that would accept any metal disc between 23mm and 27mm in diameter, regardless of composition.

The economics of this attack are straightforward to calculate. A cheater investing 200 pesos in brass washers from a hardware store can generate 5,000 to 8,000 pesos in credits before the operator notices, assuming they spread their activity across multiple machines and time periods. The washers cost approximately 3 pesos each. The 5-peso coin they simulate is worth 5 pesos. Each successful washer insertion nets 2 pesos in value. At 40 insertions per hour across three machines over four hours, the cheater gains 960 pesos in credits. The operator loses the same amount plus whatever prizes or payouts those credits generate. Over a month, with the cheater operating three days per week, the total loss approaches 11,500 pesos — nearly ten times the cost of replacing the acceptor unit outright.

How to Check Whether Your Acceptor Has Been Tampered With

Detection starts with systematic testing. The most reliable method is a controlled coin drop test using known-genuine currency. Take 50 coins of the same denomination from a bank — not from the machine’s cash box, since those coins may include damaged or foreign currency — and feed them through the acceptor one at a time while counting both acceptances and rejections. A properly calibrated acceptor should accept at least 95 out of 100 genuine coins. Acceptance rates below 90% warrant immediate investigation.

Next, run a slug test using an object you know should not be accepted. A standard metal washer of similar diameter but lighter weight is a good test piece. Try 10 insertions. Any acceptance of the washer means the acceptor’s discrimination has been compromised. Do this test during operating hours when you can observe without drawing attention — cheaters sometimes monitor for inspection activity.

Physical inspection of the acceptor unit itself often reveals tampering. Look for tool marks around the acceptor’s access panel or calibration button cover. Many acceptors have tamper-evident seals or stickers over the calibration access points. If the seal is broken, punctured, or partially peeled back, someone has been inside the calibration settings. On the acceptor channel, examine the gate arm for any visible bending — even a half-millimeter deviation from the factory angle is significant. Use a caliper if you have one available.

Check the magnet position relative to the coin path. Most acceptor designs have the magnet mounted in a fixed bracket or slot. If you see glue residue, tape, or mechanical displacement around the magnet housing, the magnet has been repositioned. A quick field measurement: the gap between the magnet face and the center of the coin channel should match the factory specification for that acceptor model. Significant deviation — more than 1mm — confirms tampering.

For electronic acceptors, check the programming. Enter calibration mode yourself and review the stored coin profiles. If you find a coin profile that does not match any legitimate currency denomination you accept, someone has programmed a slug. Delete any unrecognized profiles and recalibrate using genuine coins. Some acceptors maintain a count of calibration entries — if the calibration count is higher than what you or your staff have performed, unauthorized calibration has occurred.

Track per-machine revenue by day and look for patterns. A machine that consistently underperforms its neighbors by more than 15% in coin revenue while showing similar play counts deserves investigation. The coin box itself may tell the story — foreign objects in the cash box are not always accidental. A mix of washers, tokens from other arcades, or foreign coins in quantities that cannot be explained by occasional customer error indicates systematic exploitation.

What You Can Do to Prevent and Address Coin Acceptor Exploitation

Prevention begins with physical security. The coin acceptor should not be accessible without opening the main cabinet door, which should be locked with a non-standard key. Standard cam locks found on many arcade cabinets use key codes that are widely available — 751, CH751, and similar common key codes can be purchased online for under 100 pesos. Replace these with restricted keyway locks. The cabinet door itself should have a tamper-evident seal or security tape across the seam so that unauthorized opening is immediately visible.

For the acceptor unit specifically, consider models that store calibration data in encrypted memory and require a programming key or software tool to modify profiles. Some newer acceptor designs from manufacturers like MEI and CoinCo include audit trails that log every calibration event with a timestamp. Even if you use a simpler acceptor, adding a small security sticker across the calibration access point creates a visual deterrent — a cheater seeing an intact sticker may move to an easier target rather than risk detection.

Regular calibration audits should be part of your maintenance schedule. Once per month, perform the 50-coin acceptance test I described earlier and document the results. If acceptance rates drift downward over time, you are catching the problem before it becomes severe. Train your floor staff to check cash boxes during collection — they should know what genuine currency looks like and should flag any foreign objects immediately. Photographing unusual items found in cash boxes helps build a record that may identify patterns or specific perpetrators.

If you do find a tampered acceptor, replace it immediately rather than attempting to recalibrate a compromised unit. A modified gate arm or repositioned magnet is a mechanical change that recalibration cannot fix. The replacement cost of an acceptor — typically 1,500 to 4,000 pesos for a standard unit — is almost always less than the revenue loss from two weeks of exploitation. Keep the tampered unit as evidence and document the serial number, modification details, and date of discovery.

For multi-site operators, standardize your acceptor models and calibration procedures across all locations. When every venue uses the same hardware with the same settings, anomalies become much easier to spot. A machine rejecting 22% of coins stands out against a baseline of 3% across your other locations. Centralized tracking of per-machine acceptance rates makes it possible to identify tampered units before the monthly revenue report arrives.

Q: Can a coin acceptor really be calibrated to accept metal washers?

A: Yes. Most electronic acceptors enter a learning mode where you feed a sample of the target coin or token through the unit multiple times. The acceptor measures the physical properties of whatever you feed it. If you feed it ten brass washers, it learns that brass washers are valid tokens. The process is designed for arcades that use custom tokens instead of currency, but the same mechanism can be exploited to program slugs. The calibration mode is typically protected only by a button combination or switch setting, not by any authentication mechanism.

Q: How much money does a typical coin acceptor tampering scheme cost an operator?

A: Based on cases I have documented across Southeast Asia, a single tampered machine operated by one cheater three to four days per week typically produces losses of 8,000 to 15,000 pesos per month in direct credit fraud, plus additional losses from customer frustration when genuine coins are rejected. When the tampering affects multiple machines, as is common when an organized group is involved, monthly losses across a 30-40 machine arcade can exceed 50,000 pesos. The indirect losses from customer complaints and reputational damage are harder to quantify but are real — customers who lose money to a faulty acceptor rarely return.

Q: Why would a machine reject real coins after being tampered?

A: The tolerance windows get wider during tampering, which means the acceptor becomes less selective overall. But this widened window shifts the acceptance boundaries. Coins that were at the edges of the original narrow tolerance range — slightly worn, slightly underweight, or with minor diameter variations from minting — now fall outside the new boundaries. The acceptor is effectively rejecting coins it would have accepted before the tampering because the calibration reference points have shifted. It is not that the machine is actively discriminating against real coins — it has simply lost the ability to correctly identify them.

Q: Can I prevent this by using tokens instead of real currency?

A: Tokens reduce some forms of exploitation but do not eliminate the calibration manipulation risk. If a cheater gains access to calibration mode, they can program any physical object as a valid token. The advantage of a custom token system is that your tokens are physically unique — a cheater cannot simply buy washers from a hardware store, they would need to manufacture or acquire copies of your specific token. Combined with restricted calibration access and tamper-evident seals, a custom token system does reduce vulnerability compared to currency acceptors, but it is not a complete solution by itself.

Q: How often should I inspect my coin acceptors?

A: Visual inspection of tamper-evident seals and physical condition should be done weekly. The 50-coin acceptance rate test should be performed monthly for each machine, with results recorded and tracked over time. If you operate in a region with known coin mechanism tampering activity, consider bi-weekly testing. The cost of these inspections — roughly 15 minutes per machine per month — is negligible compared to the potential losses from undetected tampering.

What to Do Next

If you are seeing coin acceptor behavior that does not match normal operation — selective rejection, slug acceptance, or revenue patterns that cannot be explained by foot traffic — start with the systematic testing approach I described. Document what you find. Photograph any foreign objects discovered in cash boxes and any physical evidence of tampering on the acceptor units.

I encourage you to send photographs of suspicious acceptor modifications or unusual items found in your cash boxes. Real-world documentation helps build a reference library of tampering techniques that benefits the entire operator community. If you have questions about specific acceptor models or calibration procedures, reach out — I have worked with most major acceptor brands used in the Asian arcade market and can often identify issues from photographs alone.

The key point to remember is that coin acceptor tampering is almost never a one-time event. If someone has successfully modified one machine, they will target others. The sooner you identify the pattern, the less it costs you.