Prevent power wastage, fire risk, and ruined tips with this soldering iron alarm.
BUILD TIME: 3 Hours + 3D PRinting Time
DIFFICULTY RATING: Intermediate
Some people have great memories and presence of mind, and always turn our equipment off when they're done with it. For the rest of us, particularly after a long making session where a soldering iron may not have been used for a bit, or has been used intermittently, leaving a soldering iron on and walking away is all too easy.
You know the situation: You have been making something, involving soldering but then moving on to other parts of the process, leaving the iron on to use once in a while. Then, when you're done, you haven't used the iron in a bit and completely forget about it.
At best, this results in wasted electricity. More often, it results in tip degradation, shortening the useful life of your soldering iron tip and reducing its performance. Sometimes you can clean and restore the tip, but if left on overnight, you may well be replacing it. In serious but rare cases, soldering irons left on have resulted in building fires. That's costly and annoying in a commercial or industrial environment, but potentially deadly in a home workshop where everyone is likely asleep by the time a fire starts.
To solve this problem, some more expensive soldering stations feature protection systems built in. This often takes the form of a timer, which cuts the power or puts the iron into a low-temperature mode after a given time where no buttons or controls have been touched. We're yet to find a station that protects itself another way, or a stand-alone iron that has any protection, although both are probably out there.
Our project is timer-based too, but instead of relying on control inputs, ours will use both a temperature sensor and a proximity sensor to control the timer. This way, normal use of the soldering iron will restart the timer, meaning you never have to remember to touch a button or dial, or end up in the middle of a join then suddenly realise the iron is cooling. On top of that, it means the project can be retrofitted to any station, or a soldering iron stand so that non-station irons can benefit. We've included a piezo buzzer and flashing LED alert options, but also included an NPN transistor output for other control options. While it is illegal for anyone unlicensed to do any mains wiring, fixed or plug-in, there is nothing stopping you from hacking into the remote for a remote controlled power point, or similar.
We're not using a microcontroller for this project, opting instead for discrete components. The heart of the circuit is a CD4060 14-stage binary counter, which has its own internal oscillator set by an external R/C network. We're using an LM393 dual comparator to handle the inputs from the sensors, and a CD4081 AND gate to control whether or not the alarm goes off after the timed period. The alarm itself is based on the ubiquitous NE555 with a piezo buzzer and high-brightness LED attached. The NPN output is controlled directly by the AND gate and therefore will not oscillate, while the piezo buzzer and LED will.
PRINCIPLES AND CRITERIA
Our first step was to establish some principles and criteria for our build. Principles are overall concepts that govern a design process, while criteria are the specific points the design has to address. Another way to think of it is that principles are often, but not always, philosophical, while criteria are usually specific needs or engineering considerations. For example, we placed 'Accessibility', meaning designing so that as many people as possible would be able to build it, in 'principles'; while the need for a timer went into 'criteria'.
From the outset this was intended for publication, so we looked at circumstances other than our own. Soldering irons come in a variety of sizes, made of differing materials. Some holders are enclosed, some open, though most can be accessed even if enclosed. Some holders have a heatproof collar that holds the end of the barrel, while others have the end of the pencil or handle slide into the holder. The variety increases when concerning holders for hand-held irons, as the stands for these are often generic and not from the iron's manufacturer. Soldering irons and pencils are of different lengths and diameters.
All of this meant that our design principles were:
- Adaptable. It had to be able to fit or attach to any soldering iron holder and cope with different sizes and types of iron. This needed to include induction-heated and ceramic element irons, which rendered magnetic sensing unsuitable.
- Accessible. We could not use any specialised parts or manufacturing processes. We accept 3D printing these days is accessible, because those makers who don't have their own 3D printer can access a plethora of services with no greater barrier than uploading an .stl file and using an online store. PCBs are another story.
- It had to be based on protoboard, rather than a dedicated PCB. While we're heading toward the point where anyone can get their own PCBs made easily, many makers are not confident using online services to have boards made. There is still a process more complex than 3D print ordering between downloading a Gerber file from our website and ordering your boards from a chosen supplier, and many still aren't confident with this. The day when PCBs are just as accessible for everyone as 3D printing is not far away, but it is not here yet.
- It had to be based on discrete components rather than a microcontroller. While a microcontroller can do almost any job well, it was really overkill for this, and added a layer of complexity in terms of possible freezes or hang-ups. While we ended up building in a reset button for different reasons, we determined that discretes were more appropriate.
- It had to be powered on its own. No hacking of soldering station power supplies, which would rule out hand-held irons anyway. Batteries won't give enough longevity, and you cannot use a soldering iron away from power anyway, so there is no penalty here. Well, you could use gas or battery irons, but gas makes noise and is unlikely to go unnoticed when left on (and usually does not go into a holder), and battery irons generally have a trigger and cannot be left on.
Our design criteria, the specific points (as opposed to principles above) that had to be addressed, were:
- It had to sense when an iron was placed in the holder.
- It had to sense temperature to help determine if an iron had been left on.
- It had to have a reliable but adaptable timer.
- It needed a versatile way of alerting the fact that an iron had been left on.
Our initial thought was to use a Hall effect sensor to detect the presence of the soldering iron. Many irons have steel barrels but all metals have some effect on existing magnetic flux if the sensor is biased, and we hoped the electromagnetic field generated by the electric heating coil could also be sensed. Alas, the sensor cannot be placed close enough to a hot iron to even figure out if this was viable. Cold, with no power applied, we were able to make a Hall effect sensor do the job, but it had to be placed only a few millimetres away from the iron. Not only did this make it unviable for a hot iron, but the bias magnet also influenced the steel wire holder. On top of that, the electromagnetic field surrounding the iron did not radiate far enough, and because some irons use DC and others AC to heat the coil, EMF wasn't viable anyway.
We could not use a limit switch to detect the iron, as many holders for hand-held irons hold the hot barrel and not the handle. This presents the dual problem of the switch heating up beyond its limits, and the heatproof holder material being generally brittle and hard to machine a receptacle for the switch into.
With the aforementioned differences in EMF from AC and DC heater coils, ceramic elements and RF heating, RF detection was not a suitable avenue either. The choice of detector was directed by these concerns toward an optical or ultrasonic sensor. For simplicity, we went with an optical sensor. We chose to try a raw optical barrier first, without using modulation. If this was good enough, it removed the need for a modulator and demodulator to be built. We chose a basic phototransistor which is sensitive to IR light but also picks up visible light, and an IR LED.
Using this combination means the sensor can be arranged in different ways on different holder designs. In a fully enclosed holder, both emitter and receiver can be shielded from light inside the case near the handle end of the receptacle. They can be spaced away from the heat source enough, and could work in the horizontal plane. For non-enclosed holders, the emitter and receiver can be mounted on the front face of the holder so they sense the handle. In this case, we expected to need to mount the receiver downwards to avoid excessive visible light capture. This system allows for the most versatile mounting, as the emitter and receiver can be simply glued on, all the way to a custom 3D-printed mounting.
The temperature sensor similarly had to mount near the end of the holder, where the iron tip sits, and this will be different for every iron. As heat rises, this does not have to contact the iron at all. While we could use something like a non-contact IR sensor for the sake of it, this level of complexity is unnecessary when a simple Negative Temperature Coefficient Resistor (NTC Thermistor) will do the job just fine. This can be mounted in the air over the iron tip with its long legs spacing any mounting brackets suitably away from the hot area.
Both the temperature sensor and light sensor needed a comparator circuit for adjustment. The temperature sensor definitely needs adjustability, but giving the light sensor an adjustable threshold helps the user set the sensitivity so that it is effective at detecting the iron but not triggered by reflected or ambient light. This level will vary on every single workbench.
The timer had to be reliable but not overly complex. We settled in the end on the CD4060, a 14-stage binary counter with its own built-in oscillator. This removes a separate clock circuit and also reduces both board space and parts count. Having an integrated oscillator adds to reliability as well. We hadn't really considered the CD4060 for anything before, and we liked it so much we ended up dedicating this month's Classroom to it. Because of that, we'll gloss over the operation of it here. If you want the full details, the full description is in Classroom elsewhere in these pages.
To maintain alert versatility, we decided to have multiple outputs. Visual alert is via a high-brightness LED, which necessitated an oscillator, so we used a 555 which can source or sink 200mA. This also enables the unit to be configured so that the buzzer is in sync with the LED. On top of that, we added a transistor output to the AND gate so that you can connect a relay or trigger another circuit. More on this in 'where to next' at the end.
The whole thing is built on two prototyping boards that are a near-copy of the solderless breadboards most makers prototype on. The reason for this was manyfold but in most cases, these were publication considerations. Veroboard is great, but showing reliably in print just where to cut tracks and make links is very hard. Additionally, most readers are familiar with the layout of the regular solderless breadboards. You can use the same spacings, transfer designs straight from solderless to solder breadboards, and the diagrams are familiar to boot. The only catch is that the solder versions we had are slightly shorter than the solderless ones, and you'll see an effect of this in the layout.
CIRCUIT DESCRIPTION
The first section of the circuit is made up of an LM393 dual comparator. Connected to one comparator is a trimpot on the non-inverting input, while the inverting input is connected to the junction of the emitter of an NPN phototransistor and a 1kΩ resistor. This resistor to ground ensures that there is sufficient voltage to feed the input when the phototransistor is conducting, as the input of the comparator has too much impedance for current to flow if the phototransistor's emitter was connected directly to it. The trimpot allows for the tuning of the sensor so that ambient light triggering can be avoided.
The other comparator in the package has a potentiometer connected to its non-inverting input, while the inverting input is connected to the junction of a voltage divider formed by a 4.7kΩ resistor and a 5kΩ NTC thermistor.
The threshold set for this sensor by the potentiometer needs to be continuously variable, hence choosing a front panel-mountable potentiometer rather than the set-and-forget trimpot of the light sensor. The reason for this is that, in addition to needing adjustability to suit different irons, we expected ambient temperatures to affect the sensitivity. The threshold will need to be lower on a 4°C winter night than on a 45°C summer day.
The LM393's outputs are open-collector NPN transistors, and therefore cannot source current. We use them to control the bases of two PNP transistors, allowing active high signals to be given. The temperature sensor sends its high when the sensed temperature is above the threshold, representing a hot iron being present in the holder.
This is fed to an indicator LED and one side of the AND gate. The other output, from the light sensor, also has an indicator LED and gives a high output when the light beam is connected This represents the iron being out of the holder. The high is sent to the reset pin of the CD4060 counter, keeping it idle while the iron is being used. Breaking the light beam by placing the iron in the holder removes the high from the reset pin, and starts the timer.
Next along is the CD4081 quad AND gate. We're only using one of the four gates but there are no smaller packages. Even the dual AND gates in the 4000 series are in 14 and 16 pin packages but they have more inputs per gate. The AND gates in the 4081 are dual-input, and one input is fed by the temperature comparator output mentioned above, while the other is fed by the output of the CD4060 counter. The output goes to the second board, the output section of the circuit. More on this in a moment.
The CD4060 is a 14-stage binary counter/divider, with ten of its fourteen stages accessible as output pins. Being binary, each output is high after twice as many clock cycles as the output before it. This makes it very useful for long duration timing. Additionally, each output will be double the time of the one before it, meaning you can choose from quite a few different times without any component changes. By default, the outputs for the oscillator values we chose are 1800 seconds or 30 minutes; fifteen minutes, seven minutes thirty seconds, one minute fifty two seconds, fifty-six seconds, twenty-eight seconds, fourteen seconds, seven seconds, and three and a half seconds.
We connected the indicator LED to the first available output, so it flashes every 3.5 seconds when the timer is operating. The fourteenth output is connected to the other input of our AND gate. If you want the time period to be fifteen minutes rather than half an hour, then use output 13, and so on. For much more detail on the CD4060, see this month's Classroom.
After half an hour, the timer output goes high. If the output from the temperature sensor comparator is still high, the AND gate responds to both high inputs by sending its output high.
The output of the AND gate controls two transistors. One, fed via a 1kΩ limiting resistor, is a BC337 NPN transistor set up as an auxiliary output. This is for controlling external loads like relays or remote controls. On the board layout, the emitter is connected to ground, while two PCB pins are provided: One in the supply rail, the other in the collector of the transistor.
The other transistor is also fed by a 1kΩ limiting resistor. This one is a BC547 NPN transistor with the sole purpose of controlling the high-side switch BC327 PNP transistor. This transistor switches supply voltage to the NE555 oscillator. Having it controlled this way means the NE555 is completely off when not needed, though we could have had it running all the time and simply used a BC337 instead of the BC547 and placed it between the LED and piezo, and ground.
The NE555 itself is set for near-square wave oscillation and an LED and piezo buzzer connected in parallel between the output and ground, meaning they operate when the NE555 is in its source phase. The time period is roughly 0.7 seconds and is set by R18, R19, and C5. If you want to change these, the formula is:
t = 0.7 x (R18 + (2 x R19)) x C5
Remember to make sure all units are in base form. That is, Farads and Ohms, not microfarads, nanofarads, or kiloohms or megohms.
The Build:
The circuit is assembled on two soldered breadboards. There are a lot of PCB pins on these boards, which are indicated by the gold-coloured circles in the Fritzing. They are for all the indicator LEDs, adjustments, and switches. We used PCB pins because we found it easier to solder all the LEDs with their resistors, and the other items that were not mounting onboard, to fly leads first then panel mount them, trim their leads to length, and connect the leads to the board. You could choose to just solder wires straight to the board, and either solder the LEDs and such after mounting them, or just coil excess length.
3D Printed Enclosure
Additionally, we designed a 3D-printed case for our build. This is sized to fit under the holder for our workbench soldering station and so may not suit yours. It has front-panel holes for the indicator LEDs, reset button, and potentiometer, as well as spaces for a power socket, master power switch, and the trimpot at rear. We included a hole in the side for passing the wires for the IR LED, phototransistor, and NTC thermistor through. We decided against plugs and sockets for these. You could modify our design, make your own, or use a standard plastic project enclosure. You can download our working file or stl files from our website.
| Parts Required: | Schematic Id | Jaycar | Altronics | CORE ELECTRONICS |
|---|---|---|---|---|
| 1 x Pack of Breadboard Wire Links | PB8850 | P1014A | CE05631 | |
| 5 x 560Ω Resistors * | R1, R2, R8, R9, R17 | RR0566 | R7552 | - |
| 7 x 1kΩ Resistors * | R3, R6, R7, R14, R15, R16, R18 | RR0572 | R7558 | CE05092 |
| 1 x 2.7kΩ Resistor * | R10 | RR0582 | R7568 | - |
| 1 x 4.7kΩ Resistor * | R5 | RR0588 | R7574 | CE05092 |
| 1 x 10kΩ Resistor * | R13 | RR0596 | R7582 | CE05092 |
| 1 x 100kΩ Resistor * | R11 | RR0620 | R7606 | CE05092 |
| 1 x 150kΩ Resistor * | R19 | RR0624 | R7610 | - |
| 1 x 1MΩ Resistor * | R12 | RR0644 | R7630 | CE05092 |
| 1 x 4.7kΩ NTC Thermistor | R4 | RN3438 | - | CE07552 |
| 1 x 10k 25-turn Trimpot * | VR1 | RT4650 | R2382A | CE05105 |
| 1 x 10kΩ 16mm SG Linear Potentiometer | VR2 | RP7510 | R2243 | COM-09939 |
| 1 x 100nF Ceramic Capacitor * | C2 | RC5360 | R2865 | COM-08375 |
| 1 x 100nF MKT Capacitor * | C4 | RM7125 | R3025B | - |
| 1 x 1μF MKT Capacitor * | C3 | RM7170 | R3037B | CE07450 |
| 2 x 10μF Electrolytic Capactors * | C1, C5 | RE6066 | R5065 | CE05274 |
| 1 x BC327 PNP Transistor # | Q6 | ZT2110 | Z1030 | - |
| 1 x BC337 NPN Transistor # | Q4 | ZT2115 | Z1035 | - |
| 1 x BC547 NPN Transistor # | Q5 | ZT2152 | Z1040 | - |
| 2 x BC557 PNP Transistors # | Q2, Q3 | ZT2164 | Z1055 | - |
| 1 x NPN 5mm Phototransistor | Q1 | ZD1950 | Z1613 | ADA2831 |
| 1 x NE555 | IC4 | ZL3555 | Z2755 | COM-16473 |
| 1 x LM393 | IC1 | ZL3393 | Z2558 | CE07536 |
| 1 x CD4060 | IC2 | ZC4060 | Z4060 | CE07533 |
| 1 x CD4081 | IC3 | ZC4081 | Z4011 | CE07553 |
| 4 x Green Diffused 3mm LEDs | LED1, LED3, LED4, LED5 | ZD0120 | Z0701 | - |
| 1 x Red High-Brightness LED * | LED6 | ZD0156 | Z0862C | CE05103 |
| 1 x Infrared LED | LED2 | ZD1945 | Z0880A | COM-09469 |
| 1 x Piezo Buzzer | AB3462 | S6109 | ADA160 | |
| 1 x Pushbutton Switch | SW1 | SP0700 | S1084 | ADA1439 |
| 1 x Small Rocker Switch | SK0975 | S3202 | CE07554 | |
| 1 x 2.1mm DC Panel Mount Socket | PS0516 | P0622 | ADA610 | |
| 1 x Potentiometer Knob | HK7770 | H6030 | ADA2047 | |
| 2 x Solder Breadboards | HP9570 | H0701 | CE07066 | |
| 1m Cat5e Stranded-Core Cable | WB2020 | W2752 | - | |
| 32 x PCB Pins | HP1250 | H0804A | - | |
| 1 x 12V DC Plugpack | MP3147 | M9273B | - |
* Quantity shown, may be sold in packs.
# These transistors are the ones used in the project, but any other transistor with the same specifications and pin-out can be used.
ASSEMBLY
Solder the LEDs with their resistors, trimpot, potentiometer, NTC thermistor, and pushbutton switch to wires. We used pairs from a stranded-core Cat5e cable, because the twisted pairs give really neat wiring and there are four colours to help. We chose to use the white/coloured stripe core for positive, and the solid colour for negative.
Heatshrink your connections too. The IR LED is included here, as is the phototransistor. Both will need reasonably long leads, as does the NTC thermistor. We also coloured all over the phototransistor with black permanent marker, which in our case eliminated the ambient-light triggering issue.
The phototransistor itself is a little counter-intuitive. It is an NPN transistor, but the short leg is the collector and the long leg the emitter. We followed the convention we were using for the rest of the circuit: The collector, which we connect to supply, had the positive wire colour of white/coloured stripe connected to the short leg, while the long leg got the negative solid colour.
In all cases, we found it helpful to put masking tape labels on each item. This is particularly true of the CD4060 indicator, which has a 2.7kΩ resistor rather than the 560Ω resistor on all the other LEDs. Also, remember that there is a 100nF ceramic capacitor across the pushbutton terminals.
Add the onboard components as shown in the Fritzing and photos. Note the presence of wire links. There are no links or joins under the board, every connection is visible from above. We use the same wire links used on the solderless breadboards, meaning wire colour shows length. Be very careful with spacing.
At the far end of the main board, we had to bend wires to join the rails outside the board profile. That's the yellow wires in the photo. This is because the soldered versions are just shorter than the solderless cousins, and we need every row.
Also note that we had to bend the legs of the 1μF MKT capacitor inwards, so it takes up two rows and not three. There are also two vertical resistors around the CD4060, one is a 10kΩ between the negative rail and pin 12, and the other is a 100kΩ between pin 10 and the capacitor/resistor common row.
Finally, place the PCB pins if you are using them, or wires if you are not, for the off-board components. Be very careful of the PCB pin placement, as they are a little hard to see in the images. There are no PCB pins for the wires that join the power rails of the two boards, but you could add them if desired.
There are pins for the wire connecting the AND gate output to the two resistors for the output control transistors on the second board. There are two PCB pins that have no connections in this build as-is: They are the ones near Q4, the auxiliary output transistor and they are described earlier.
Mount the boards in the enclosure. We used double sided tape because there is little room in these boards for drilling mounting holes. Once you're happy with this, you can mount the front and rear panel items.
The switch snaps in, the power socket and potentiometer have a thread and nut, and the trimpot and LEDs glue in place. We used hot melt glue, but don't let it heat too much: 3D printed plastic does not enjoy very hot glue!
The last, and most tedious, step is to route the fly leads, trim them to length, strip the ends, and solder them to their relevant pins. Do this very carefully. With so many wires, it's easy to mix up a wire with another or just get the positive and negative to the wrong pics, such as with the phototransistor. We also used a label maker to print small labels for our panel indicators. You can do the same, or use the clear shipping or address labels that feed through your home printer, or just write on them. Or, you could rely on your amazing memory but we have no chance of that!
Also in this step, you'll need to add power wires from the socket and switch at the back, to the appropriate place on the main board. The master switch is not in the schematic and is in fact optional. If you use it, just wire the positive of the socket straight to the switch, and then the other terminal becomes your power supply positive connection. The buzzer can be glued over its hole now too.
With the wires for the IR LED, phototransistor, and NTC thermistor hanging out the side, close the lid of your box and place the soldering iron holder on top. Now you can route the wires for these sensors and glue the items in place. On our holder, the IR LED and phototransistor has to mount horizontally, as shown. We had wanted to mount the phototransistor facing the bench and the IR LED facing the roof but it cannot be done on this holder. Hot melt glue turned out to be good enough, though once we're happy with placement we'll likely use a more permanent (and less heat-sensitive) glue.
The same goes for the NTC thermistor, which was glued over the side of the frame so that it sits a reasonable distance above the barrel of the iron. This arrangement and the appropriate distance will vary between irons and holders.
OPERATING IT
We furnished the design with an indicator LED nearly everywhere we could, to make visual monitoring and fault-finding easier. There is an indicator for power, temperature sensor high (iron hot), IR sensor high (iron out of holder), and timer operating. There is also a high-brightness red LED in the front panel as the visual alert when the alarm goes off.
Now all that remains is to plug in a 12V plugpack to the socket, and turn on the power.
With the iron OFF and NOT IN the holder, adjust the trimpot at the back until the front-panel LED for the IR beam lights. Now place the iron in the holder, breaking the beam. The indicator LED should extinguish.
Adjust the front-panel potentiometer until the LED indicator for the temperature sensor is extinguished. Next, turn the iron on and wait for it to heat up. The temperature sensor indicator LED should light up at some point during the heating process. You may have to adjust the potentiometer until you see this LED light. When it does, turn the iron off. As it cools, you should see the LED extinguish.
While the iron is sitting in the holder, you should see the LED indicator for the timer flashing on and off every three and a half seconds or so. When the iron is cool, remove it from the holder, and the timer LED should go out. This will indicate that the IR sensor and its comparator is resetting the timer every time the iron is removed from the holder.
WHERE TO FROM HERE?
There are three main directions from here in this project. You can adjust timings, either of the overall timer, or the alarm oscillator. The overall timer can be adjusted by choosing a different output or by altering the timing components, or both. See this month's Classroom for the maths on how to choose components or outputs.
The oscillator is a standard astable 555 operation, and for makers who aren't already familiar with that, the information is everywhere. You could also make a more complex circuit, such as a 555-based saw-tooth wave siren, driving a speaker instead of the piezo buzzer.
Mounting enclosures and arrangements are the next direction. You might want to design alternative methods of holding the sensors in position, or even a whole new soldering iron holder factoring these in. That’s up to you if you have those skills, but any 3D printing will have to be tolerant of the temperatures involved, at least if the sensor mounts get close to the heat source.
Finally, the NPN output from the secondary board. We envisage this to be used to drive other circuitry, such as a relay for a buzzer, light, or both in another location. Maybe you can put one in the lunchroom at work or in the kitchen at home for when you've walked well away from your work area and left your soldering iron on.
However, it opens up the possibility of adding an IoT or SIM-based alert to send your phone a message if you're really left the area. You can also use a relay to trigger the button of a remote controlled power point that will turn off power to the whole station should contacting you remotely not be a practical or suitable option.
Hopefully this alarm can save you a few new iron tips, and in extreme cases, maybe even expensive fire damage. There are probably many other ways of doing what we have done here, and many other ways to design a similar approach. The design will not be perfect for everyone because of the sheer variety of circumstances but even so, we'd love to see what people come up with.