Your Cash Drawer Stopped Opening. It Is Almost Never the Drawer.

The drawer stopped opening on a Tuesday morning and everything else at the counter is fine. The receipt still prints. The card reader still takes payment. The software says the sale went through. But the till stays shut, and somebody is now making change out of a coffee mug and a takeaway container.

Almost every time we get called out for this, the fault is not in the drawer. The drawer is the one part of the chain that mostly cannot break in the way people assume it has — and understanding why turns a mystery into a short list you can work through yourself before you call anybody, including us.

The drawer has no brain. That is the whole article.

A cash drawer under a shop counter is a metal box with a lock, a spring-loaded catch and a solenoid — an electromagnet that yanks the catch open when it gets a jolt of power. There is no computer inside it, no network connection, no firmware and nothing to configure. It cannot be "offline". It does not need a driver. It does not know your point-of-sale software exists.

What it is connected to is the receipt printer. On the back of essentially every thermal receipt printer sold for a counter there is a port the manufacturers call the drawer kick connector, and the drawer plugs into that. When the software rings up a cash sale it tells the printer to print a receipt and, in the same instruction, to fire that port. Epson's technical reference guide for the TM-T88VI states the mechanism in one line: "A pulse output is sent to drawer kick connector pin 2 or pin 5, and you can open the drawer." Star Micronics calls the same thing a "Peripheral Unit Drive Circuit" and specifies its output as 24 V, maximum 1.0 A.

So the chain is: your software, then the printer, then a short cable, then an electromagnet. Four links, and the drawer is the last one. A drawer that will not open is nearly always a message that went missing somewhere in the first three — which is why "the drawer is broken" is the least likely explanation and the one everybody reaches for first.

Two minutes before anything else: is it simply locked?

Every cash drawer has a physical lock, and the key that came with it is very often taped to the underside of the drawer where nobody has looked since it was installed. Toast's own support guidance gives the tell in one sentence: "A horizontal key hole means the drawer is locked. A vertical key hole means the drawer is physically unlocked." If the lock is horizontal, no amount of electricity is going to help, and somebody turned it — usually the person who closed up.

Now do the test that splits this problem cleanly in half: turn the key and open the drawer by hand. This is the single most valuable thirty seconds available to you, because it separates a mechanical fault from a signal fault, and those go to completely different places.

If the key opens it, the mechanism is fine and you have a signal problem — everything below applies. If the key does not open it, Toast is blunt about what that means: "If the cash drawer does not open with the key, the mechanical mechanism is jammed or broken. This is a hardware issue." That is the one case where it really is the drawer, and the usual culprit is something wedged at the front of the till insert — a coin under a bill weight, a rolled receipt, a pen. Some drawers also have a manual release underneath; Toast documents one for its own drawers and explicitly notes it is "For Toast cash drawers, not APG", which is a useful reminder that this varies by maker and you should not go poking at the underside of a drawer you have not looked up.

The most useful question on this page: did the receipt print?

If the drawer is printer-driven — and unless you specifically bought a USB drawer, it is — then the printer has to be working perfectly for the drawer to move at all. Square states the dependency plainly: "A cash drawer must be connected to a receipt printer to open automatically." If not, "you need to manually open the drawer using the cash drawer key."

That dependency is stronger than most people expect. The drawer does not fail gracefully when the printer is unhappy. A printer that is out of paper, has its cover unlatched, is showing an error light, has dropped off the network, is powered but not selected, or is mapped to the wrong printer profile in the software will simply never send the pulse. The sale completes, the software is satisfied, and nothing happens at the counter.

So before you touch a setting: print something. If the receipt does not print, you do not have a drawer problem at all — you have a printer problem, and fixing the printer will fix the drawer for free. Check the paper, the cover latch, the error light, the cable, and whether the printer still holds the network address the software expects. We have a separate piece on receipt and office printers that vanish after a router change or a new Wi-Fi network, and that is a surprisingly common way for a counter to lose its drawer.

The receipt prints and the drawer stays shut: now you are in the settings

This is the classic version of the fault, and it is almost always a setting rather than a wire, because a printer will happily print all day with its drawer output switched off. The setting lives in two places and you may need to check both: the printer driver on the computer, and the printer profile inside the point-of-sale software.

On Windows with a Star printer, Star's own help centre gives the path: Settings, then Bluetooth and devices, then Printers and scanners, select the printer, Printer Properties, and the Device Settings tab. There you set "Peripheral Unit Type" to "Cash Drawer", and then choose when it fires under "Peripheral Unit 1" — Star describes the two options as "Document Top: Opens the cash drawer before each receipt prints" and "Document Bottom: Opens the cash drawer after each receipt prints". If that first menu is not set to Cash Drawer, the port never fires and everything else you try is wasted effort.

On the software side, the same switch exists again under a different name. In Square, the drawer follows the receipt printer's profile: you need to have "created and enabled a printer profile for your receipt printer", and Square specifies that you must "ensure that you have enabled Receipts under the 'Use this Printer For' setting — this will ensure your cash drawer will open automatically." A printer profile set up for kitchen tickets or order tickets but not receipts will print beautifully and never open the till. This is the single most common thing we find on a counter where somebody recently added a second printer.

Peripheral Unit #2, and why it is not a shot in the dark

Buried in Star's setup instructions is a line that reads like guesswork and is not: "If the drawer fails to open try selecting Peripheral Unit #2." Every troubleshooting page on the internet repeats it. Almost none of them say why it works, and the why is the most useful thing on this page, because it tells you when to expect it to work and when not to bother.

The drawer kick port is not one output. It is two. Star's hardware manual for the TSP100 series lays the connector out pin by pin: pin 2 is "DRD1 — Drive signal 1", pin 5 is "DRD2 — Drive signal 2", both outputs, with drive power on the pins between them. Epson documents the same two-output arrangement from the other side, specifying that "the load, such as a drawer kick solenoid, must be connected between pins 4 and 2 or pins 4 and 5 of the drawer kick connector."

And here is the sentence that explains the whole thing, from Epson's own manual: "Whether or not pin 2 or pin 5 operates the drawer kick connector depends on the connected cash drawer." In other words, the printer can fire either of two wires, your drawer's solenoid is wired to exactly one of them, and if the software is firing the other one, the printer is doing precisely what it was told and the drawer is correctly ignoring it. Changing peripheral unit 1 to 2 is not a superstition. It is telling the printer to shout down the other wire.

We will not tell you which number maps to which pin on your particular model, because neither manufacturer states that mapping in a single sentence we can quote, and a confident wrong answer here sends you rewiring things you should not touch. What matters is the shape of it: there are two wires, your drawer is on one, and flipping that setting is a legitimate two-second test rather than a random poke. If it works, write down which one, because the next person to set this counter up will need it.

The port that looks like a phone socket, and what both manuals say about that

The drawer cable ends in a modular plug — the same family of clicky plastic connectors as a telephone cord and an Ethernet lead. It is not either of those, and both major printer makers thought this was dangerous enough to warn about in more than one place.

Star prints it on the machine. The hardware manual describes a symbol placed near the peripheral drive connector meaning "Do not connect this to a telephone", and repeats it as a caution: "Do not connect a telephone line into the peripheral drive connector. Failure to observe this may result in damage to the printer." Epson carries the same warning about its own port — "Do not insert a telephone line into the drawer kick connector. Doing so may damage the telephone line or printer." — and lists it among the safety cautions at the front of the manual.

The warning runs in the other direction too, which is the one people actually trip over. On the TM-T88VI the drawer kick connector and the Ethernet connector are neighbours on the lower rear of the printer, and Epson explicitly says: "Never attempt to connect the drawer kick cable or a standard telephone line cable to the LAN connector." It also warns against a third mix-up on the same panel — "Do not insert a Type-B USB connector into the LAN connector or the drawer kick connector. If it is inserted, the connector, printer, and the system may malfunction."

The practical upshot for a counter: if the drawer stopped working right after somebody moved the printer, cleaned behind it, or ran a new network cable, look at the back of the printer before you look at anything else. Three ports that accept three similar-looking plugs, in a dark space under a counter, reached blind. It is not a stupid mistake; the manufacturers wrote warnings about it because everyone makes it.

The cable is not universal, and the drawer end is Ethernet-shaped

Here is the trap that survives a printer upgrade. APG, which makes a very large share of the cash drawers under American counters, states it in one sentence on its own site: "It is important to note that not all printer manufacturer's drawer kick ports are pinned alike. This means there are many different cables that could possibly be used for your setup."

Their connectivity guide shows what that means in practice. Epson's TM series and Star's HSP, SCP, SP and TSP series both take APG's CD-101A cable — so those two swap cleanly. But Cognitive TPG, HP's A776, A798 and A799 receipt printers, Fujitsu's FP510 series, Toshiba's SureMark and Wincor's TH series all take a different part, the CD-102A. Same connector, same click, different wiring. A shop that replaces a dead Epson with a secondhand HP and keeps the old cable has a drawer that does nothing and a cable that fits perfectly.

The connector shapes are worth knowing too, because they explain a mistake we see a lot. APG's pinout documentation lists the printer end of these cables as a "6P6C Mod Plug" — a six-position modular plug, the phone-cord shape — while the drawer end plugs into an "8P8C Modular Jack", which is the same physical socket as Ethernet. Your cash drawer therefore has a socket in it that will happily accept a network cable. A network cable in that socket is wired for something else entirely and will not open anything.

So: never assume a cable that fits is a cable that works, and if you are buying a replacement, buy it by printer make and model rather than by the look of the plug. If you have a spare of the correct part, swapping it is a genuinely good test — drawer cables live on the floor, get rolled over by chairs and yanked when the printer is pulled forward, and they do fail.

12 volts, 24 volts, and the drawer that was fine at the old shop

The other thing that has to match is voltage, and this one is invisible from the outside. APG builds the same drawer with different interface boards inside it, and the boards are voltage-specific: the MultiPRO 320 is the 24 volt version, the MultiPRO II 420 is the 12 volt version, and the MultiPRO III 520 is the dual one — APG notes that "The 520 Interface can support both 12VDC and 24VDC environments."

On the printer side, the counter-standard receipt printers run the higher figure. Epson lists the TM-T88VI supply voltage as DC 24 V plus or minus 7 percent, with "Maximum 1A for drawer kick driving." Star specifies its peripheral drive output as 24 V, maximum 1.0 A. The fact that a major drawer manufacturer sells three different interface boards, organised by voltage, is itself the message: this is a real compatibility axis, not a detail.

It matters most in exactly the situation where nobody checks it — a drawer inherited from a closing business, bought used online, or carried over from a previous system. The drawer looks identical, the cable clicks in, and the pairing is wrong underneath.

Epson also publishes the conditions a non-Epson drawer must satisfy, and the consequence of ignoring them, so we will quote it rather than paraphrase: the resistance of the solenoid "must be 24 [ohms] or more or the input current must be 1A or less", you should "be sure to use the 24V power output on drawer kick connector pin 4 for driving the equipment", and if the drawer does not meet the stated conditions, "devices may be damaged." That is the manufacturer's language, and it is the reason we will not tell you to just try a drawer and see. If you are mixing brands, look the drawer's interface up first.

The printer that beeps instead of opening the drawer

This one sounds invented and is straight out of Epson's manual. On models with a built-in buzzer, the buzzer is driven from the same drawer kick connector as the drawer — and the two cannot share. Epson states it flatly: "Do not set the buzzer to beep for pin numbers used for drawer operations. The buzzer and the cash drawer cannot be operated by one pulse signal."

It gets more concrete than that. Epson's guidance is "When using a cash drawer, connect a cash drawer operated by pin 2. If you have to use a cash drawer operated by pin 5, change the DIP switch settings for the buzzer circuit." And for the optional external buzzer — the one a busy kitchen or a back-room counter often gets added — the manual is unambiguous about the trade: "When the optional external buzzer is enabled with the memory switch (customized values), a cash drawer cannot be used. Be sure to disable it when you use a cash drawer."

There is a physical version of the same warning: "Do not connect both the optional external buzzer and the cash drawer to the printer at the same time by using a branched connector."

So if your drawer died the same week somebody added an audible alert to the printer, you have found it. That is not a coincidence, it is a documented either-or, and the fix is a settings change rather than a purchase.

Two drawers on one printer, and the rule both makers state

Plenty of counters run two drawers off one printer for two cashiers. Both manufacturers document the same limit in almost the same words. Epson: "Two driver transistors cannot be energized simultaneously." Star: "It is not possible to drive two drives simultaneously."

Star adds a duty-cycle condition — the drive must satisfy ON time divided by total time of no more than 0.2 — and Epson gives the operational version of the same rule: "Leave intervals longer than 4 times the drawer driving pulse when sending it continuously." In plain terms, that port is meant to be pulsed briefly and left alone, not hammered. Software that fires the drawer repeatedly because a cashier is clicking the button is working against the hardware.

There is also a behaviour that looks like a fault and is not. Toast documents it: "If another cash drawer on the same printer is still physically open, a new drawer will not open until the open cash drawer is pushed shut." If your second drawer is intermittently dead at the busiest moment of the day, go and look at whether the first one is sitting open.

And if you are building a two-drawer counter, the splitter is not generic either. APG lists a general dual-drawer splitter and a separate part specifically for the Epson drawer kick port, described in their guide as switching in series. Buy the one for your printer.

The other wire: the drawer tells the software when it is open

The drawer kick connector is not only an output. There is a return path, and it explains a second family of complaints that nobody connects to the first.

Star lists pin 6 on its connector as "DRSNS — Sense signal", with its direction marked as an input — that is, information travelling back into the printer. Epson describes the same thing from the drawer end: "When the drawer open/close signal is used, a switch must be provided between drawer kick connector pins 3 and 6", and "You can also check the open/close status of the drawer by checking the signal level of the drawer kick connector pin 3." APG's pinout confirms that its interface boards carry a status switch with ground, common, normally-open and normally-closed contacts alongside the solenoid connections.

That is a real switch inside the drawer that closes when the drawer shuts. It is how your software knows the till is open, why some systems refuse to start a new sale while the drawer is out, and how cash-management reporting knows when the drawer was accessed without a sale. It is also, importantly, a separate set of wires from the ones that fire the solenoid.

Which produces a fault that looks bizarre and is not: a drawer that opens perfectly but whose status the software gets wrong, or gets none of. APG's own cable list shows that not every cable carries every contact — some parts wire the status switch, some do not. If your drawer opens on demand but the software constantly thinks it is open, or never notices that it is, you are looking at the status pins and probably at the cable part number, not at the solenoid.

If your till is a tablet, this is a different animal

Everything above describes a printer-driven drawer, which is still the most common arrangement and the one all the manuals are written for. But a lot of newer counters in Southern California are a tablet and a card reader, and there the drawer may not be printer-driven at all.

Square, for instance, supports USB cash drawers that connect directly to Square Stand, Square Terminal, Square Register, Square Handheld, and phones or tablets running Square with a reader. A USB drawer is its own device to the software, with its own connection to check and its own settings — and none of the pin-2-versus-pin-5, cable-part-number, drawer-kick material applies to it.

So the first question before you troubleshoot anything is: which kind do I have? Follow the cable out of the back of the drawer with a torch. If it lands on the back of the receipt printer, you have a printer-driven drawer and the chain above is your list. If it lands on the tablet, the stand or a hub, you have a USB drawer, and the printer is irrelevant to the problem — which is worth knowing before you spend an afternoon on printer settings.

There is a purchasing version of this too, and it catches people. If you are running a compact tablet setup and you want a printer-driven drawer, you need a receipt printer in the mix for the drawer to open automatically at all. Buying the drawer first and working out the rest later is how a shop ends up with a very expensive manually-operated metal box.

What we will not tell you, and why

A few things come up constantly in the search results for this problem, and we are naming them rather than repeating them.

Any figure for how often each cause is to blame. Several pages will confidently tell you that a given percentage of dead drawers are cable faults or solenoid failures. We could not trace a single one of those numbers to a study, a service dataset or a sample size, so we have written this as an order to work through rather than a league table.

Any instruction that involves opening the drawer's case or the printer's. Star puts a label on the screws for a reason and states the reason: "Individuals, other than service personnel, should not remove these screws. High voltage areas in the case can be dangerous." A solenoid is a coil across a mains-powered device's output. Diagnosing it from the outside — key test, cable swap, settings — is a job for a shop owner. Getting inside it is not.

Any electrical claim about your specific model beyond what we have quoted. Pin assignments, voltages and duty limits are published per printer and per drawer interface, and we have quoted a TM-T88VI, a TSP100-series Star and APG's interface documentation because those are what we could read first-hand. Do not carry those numbers to a different make and assume they hold. That is the exact mistake the whole cable section is about.

And any repair price. What this costs depends entirely on which of the four links turned out to be broken, and three of the four are usually free to fix.

The order to work through it, on one page

One: turn the key. Horizontal is locked. If the key does not open it either, stop — that is mechanical and the rest of this list does not apply.

Two: print a receipt. If it does not print, fix the printer and the drawer will very likely come back with it.

Three: look at the printer's error light, its paper, its cover latch and whether it is still at the network address the software expects.

Four: check the printer profile in your point-of-sale software — specifically, that the profile is enabled and set to handle receipts, not only tickets.

Five: check the driver setting on the computer: peripheral unit type set to cash drawer, and a fire point chosen (before or after the receipt).

Six: try the other peripheral unit. This is the two-wire thing, not a superstition, and it takes two seconds.

Seven: look at the back of the printer with a torch. Confirm the drawer cable is in the drawer kick port and not in the network port, and that nothing else got moved when the counter was last cleaned or rewired.

Eight: swap the cable if you have a known-good one of the correct part number for your printer make.

Nine: if the drawer or the printer is new to this counter, check that the drawer's interface voltage matches the printer before you go further.

Ten: if a buzzer or audible alert was added to the printer recently, that is very likely your answer — the drawer and the buzzer share the port.

Eleven: if it opens fine but the software misreports its state, that is the status switch and the cable, not the solenoid.

Twelve: write down what fixed it and tape it inside the cupboard under the till. The next person setting up this counter will be trying to remember it at 8am on a Saturday.

Where we come in

This is ordinary small-business work and we do a lot of it across Southern California — shops, salons, cafés, workshops, clinics and anywhere a counter takes cash. A dead drawer is usually a single visit: work the chain from the software down, prove which link is broken rather than replacing parts hopefully, get the correct cable in, and leave the settings documented so the next change does not undo it.

We will also tell you honestly when there is nothing to bill for. A surprising share of these end at step four or step six — a printer profile that lost its receipts tick, or a peripheral unit setting that changed when the driver was reinstalled. If that is your situation, we would rather say so on the phone than turn up.

And if the reason you are reading this is that the whole counter is unhappy rather than just the till — the card reader will not connect, the receipt printer keeps dropping off the Wi-Fi, the copier stopped scanning to email — call us and say which. Counters tend to fail together, because they usually share one network and one afternoon when somebody changed something.