Have A Blast!

Historically themed electrical generator actuator

Daniel Koch

Issue 79, February 2024

Styled on the historical blasting machine that many of us know well from movies and cartoons, this project uses an internal generator to make electricity, and use it to send a controlled digital high signal to a microcontroller. It can be used for anything from turning on a lighting display, as a prop in a movie or as an objective in a game like paintball or Nerf tournaments, or to just relive your Looney Tunes childhood! If the blasting machine is not your thing, the principles used and some of the things we discovered are useful in other projects, too.

We will first look at the historical basis for the project, and compare it to our movie and cartoon inspiration. There are some significant differences in both design and operation between history and the media portrayal thereof. We will have a look at the somewhat curious technology used in the real deal, compare it to what is used now, then introduce our design and what it can be used for.


We will see why as we summarise how the real version works. Having said that, if someone doing the wrong thing or wishing harm can get their hands on explosives and unsophisticated detonators together, it is unlikely they will be building a device like this to use them. Our project does not do anything a pack of AA batteries or a 12V car battery would not do in those rare cases where a detonator does not require a properly designed modern blasting machine.

We also feel compelled to point out something else. This project should not be associated with war. While the popular culture/media representation of blasting machines that our project has been designed to look like has been used in some Westerns, it is generally not a war movie device (with some notable exceptions) and this style of blasting machine was less common in real war, too.

Most military explosives of the era were triggered with alternatives like smokeless fuse, timed chemical fuse, or friction igniters, and other technologies which were all more portable in the field.

Regardless, it would be tone deaf of us to not acknowledge the world at large. There are two main conflicts in the media at the moment but there are dozens of other conflict zones around the world, and there have been for decades, which do not make the mainstream media. All of these affect civilian populations and in some cases, are targeted at certain populations in race wars or ethnic cleansing operations. They are ongoing daily and have been for a long time. We sometimes hear about the domestic terrorist group Boko Haram or the situation in Myanmar, but generally we just do not hear about the many more like them.

Many of these conflicts involve the use of improvised explosive devices as weapons of war or terror but again, the type of blasting machine this project is inspired by is not related to the systems used to detonate such devices. Those are initiated either by real and modern commercial or military blasting machines, or simple battery packs. Some would say that now is not the time for a project like this, but we suggest the only difference between now and any other time is that those two conflict zones that will come to mind for many people have the media's attention.

Many of the others before that, or even now, were and are just as serious but simply didn't or don't generate the media interest. We hope the association with this project in people's minds tends toward Looney Tunes cartoons and fireworks displays as we intended rather than real destruction and human suffering.


Sometimes, you will notice the word 'dynamo' referring to an electric motor being used in reverse to generate electricity. It is less common today, partly because very few dynamo-powered devices are around now, being a bit of a craze in times gone by. Other terms encountered are 'generator', 'alternator' and 'magneto'. The word 'generator' is both a technical and a general term.

Typically, the word 'generator' is used in general to mean any device that produces electricity from rotational energy, while technically referring to a device that produces alternating current, which is also what 'alternator' means. However, most alternators are used in cars where the AC output is immediately rectified into DC. Generators, on the other hand, usually maintain an AC output. A dynamo produces DC right from the outputs, with no rectifier needed because of the commutator inside which changes the polarity of the output to maintain pulses of DC.

The problem is that there are regional and historical variations on these definitions and so you will always need to use context to determine which device you are reading about. We are going to use the term 'generator' as a catch-all term because we are going to experiment with configurations that could meet the criteria of dynamo, generator, and alternator.

Further, you may see the term 'magneto'. This is less common today, it is a design where the rotating part (the rotor) is the permanent magnet, and the outputs are DC pulses of current. Some magnetos had transformers inbuilt so that the DC pulses were quite a high voltage, hence the association between magnetos and older ignition circuits for internal combustion engines.


Our primary inspiration for this project comes from the Looney Tunes classic duel between Road Runner and Wile E Coyote. There are quite a few scenes throughout the series' history where Wile E Coyote uses a plunger blasting machine to try to blow up Road Runner, cause a landslide, block him in a tunnel, or otherwise facilitate the Coyote's next meal.

They also feature in some westerns, both accurately and well out of historical context (some films were set earlier than the blasting machine's invention or at least commonplace adoption), mainly in the context of mining explosives commandeered by the villain in some way. There is a spectacular (by the standards of the day) scene in Lawrence of Arabia (1962), in which a blasting machine is used by the protagonist to set off explosives to derail an enemy train.Plenty of more modern war films (and more besides) feature the almost ubiquitous M18 Claymore, a United States of America-originated manually-controlled mine for ambush and defensive emplacement. These are triggered by a 'clacker', which is a hand-cranked blasting machine used in the palm of one hand to initiate an explosive. These are a far more modern device than the ones we are basing our project on but they still feature a hand-cranked generator and send a burst of current to an electric detonator.

There are other examples, too, like a scene in the recent Oppenheimer biopic where a twist blasting machine is used to fire one of the test shots. You can hear the generator continue to spin after the handle stops. Beyond many film and television depictions, sometimes blasting machines, particularly the plunger style, have been the inspiration for props to begin fireworks shows (quite often unconnected, the real trigger being an advanced control console), light large Christmas trees, or even fire confetti cannons. This is far more the kind of use we have in mind for our project.


Before the invention of the electric igniter and blasting machine, all explosive work, both military and civilian, was controlled by pyrotechnic fuse. This was some sort of cord saturated or coated with, or a tube filled with, a flammable mixture that burned at a controlled rate. In fact, early fuse was not so controlled and even today, cheap fuse is considered inconsistent. However, the idea is that the burning fuse would both form a time delay to allow safe escape, and get the fire to where it was needed inside the body of the detonator with later explosives or inside the charge when blasting was done with bagged gunpowder.

It was some time around the late 1870s that electric igniters were invented, and the blastic machine soon followed. Battery technology of the time was bulky and heavy, and rather low on capacity for a given size. Using an electrical generator inside a portable box and operated by human power was a much more reliable and viable way of initiating electric detonators.

Designs varied, with push-down T-handles, twist handles, and winders that needed multiple rotations, all in common use. Some had gearboxes, too, so that the generator spun multiple times for a single travel or rotation of the actuator, whether it was a plunger, twister, or winder.

As detonators evolved, so did blasting machines. Electric initiators became more than a simple electric match touching off a charge of gunpowder. A shock is required to detonate high explosives, and that includes dynamite, one of the first high explosives. The electric section of the detonator has a fine wire in a pyrotechnic charge. The wire heats up with current, and initiates the pyrotechnic charge. This then initiates a primary explosive (a very sensitive low explosive), which then sets off a small amount of a less stable high explosive to provide the shock required to detonate the main charge. As these designs evolved, the electric side of the system got safer by needing a higher voltage at a reasonable current to set it off. This meant static electricity or stray currents from, say, radio induction, were less likely to set off the detonator.

To keep up, blasting machines evolved. Models with geared motors driving output transformers and those with capacitor banks became commonplace.Some of these are the plunger type, but many more are winder type, needing multiple twists of a rotor or winder to generate the required charge. Note that many plunger and winder machines had a ratchet engagement system so that turning the handle the wrong way or pulling the plunger upwards to set up the shot would not drive the generator and fire the charge.

Fast forward to today, and blastic machines are battery powered, very sophisticated devices with safety systems built in. Many have wireless capability, so charges can be wired to a controller then a remote taken far away for initiation. Many feature inverters for high voltages and most have output sequencing, where banks of charges can be activated one after another with programmed delays to suit a specific type of rock in a quarry, for example. These are all a far cry from the blasting machine Wile E Coyote uses to try to kill Roadrunner. Importantly, modern electric blasting caps, initiators, or igniters cannot be actuated with the kind of low voltages our project will output and even if there are some which could be, in that case our project will not do anything you cannot do with a cordless drill battery or a set of AAs.

It is worth noting that there are plenty of other designs and operating methods around too, and the ones we have described are only the ones we found good schematics or disassembled photos of.

As an interesting side note, we have a link to a Tom Scott video at the end in the 'Reading and Resources' section, detailing what High Explosives and Low Explosives actually are, because it has nothing to do with. Many people think this is how explosive the products are, but in fact that is far from the truth and in many cases, high explosives are safer to be around than low explosives.


Spinning a generator with a handle, be it a plunger or a rotational handle of some form, generated a charge over time. However, it is desirable to dump this charge suddenly into the detonator, mainly so a reasonable current is achieved. That means some method of storing an electrical charge was needed. However, blasting machines as we are in them came at a time in history when the field of electronics was in its infancy.

One of the most elusive details we discovered in researching the history of blasting machines is the method of adding storage capacity to them. Many of these devices predate modern battery technology, with batteries of the time being large, heavy, and non-rechargeable. The capacitor as we know it today dates from around 1900 in its earliest commercially-produced form.

Some blasting machines used Leyden jars, early capacitors which had foil on the inside and outside of a jar, which stored a charge. Others used capacitor precursors, such as layers of glass plates with metal foil in between. These had to be manufactured for the purpose, as they were not commercial, off-the-shelf products in the earlier days of blasting machines.

Of course capacitors became available after around 1900 or so but they were still small and low on current capability even if they could store a nominally large charge. Blasting machines of the plunger or winder style were made well past this date, so it is possible to find museum or even traded collectable historic examples with capacitors as the storage medium. However, we also found some with early forms of rechargeable battery, using a chemistry that probably didn’t have a great shelf life, but as a form of capacitor, was adequate.

More often however, at least in our research, flywheels were used. The plunger or winder would spin up a flywheel, either directly or through gears, which was either connected to the generator element or was the heavy generator element itself. This works because, while unconnected, the generator spins freely. This is one of the things that gets free energy believers excited. However, once a load is connected, current can flow from the generator, and when it does, a magnetic field is created in the windings of the generator, which opposes the magnetic field that created it.

This is true whether the rotating element is the windings or The magnets, as both designs exist. So, the flywheel can spin and store energy until the moment the load is connected to the generator. Versions with a manual switch generally had a drive mechanism that disengaged at the end of its travel, allowing the flywheel to spin freely. For example, the teeth of a winder gear would end just before the end of the rotation, so the pinion was not engaged with anything after the winder had reached its stop.

Many blasting machines had a switch involved to connect the load after a charge had been built up. In plunger styles, this was often a contact under the end of the plunger: The rack spun up the generator, either storing the charge in some form of capacitor or kinetically in a flywheel, then the switch was activated as the plunger reached the end of its travel, connecting the load and dumping the current suddenly. In other systems, this may have been a separate, manually-activated switch like in the machines which used a winder that rotated several complete rotations. In twist-style machines, where the rotating handle went through less than a complete rotation, a style of limit switch like the plunger description was often used.


Initially, we thought of using a DC motor as our generator element, driven by a rack on the plunger engaging with a pinion gear on the motor. However, some initial bench testing revealed that the 6V motor that we had on hand rarely produced above 1V when spun with fingers. Unfortunately the drill motor is inducing a lot of noise, particularly in the blue trace for channel 2.

This is through the breadboard, acting as an antenna, not through the shielded probe leads. The breadboard rails also exhibit significant capacitance, 56pF worth measured with an LCR meter! Even with the shaft in the chuck of a drill, we were able to achieve just under 4V as a peak. This was never going to be enough for what we wanted.

We may get more from a 12V motor, but we did not have one right there and then, and it was looking like it would be a big effort for minimal gain to fit a big enough one. The bench test consisted of hooking the motor coils to an oscilloscope directly, then through a bridge rectifier to compare the difference.

While we did not have a big 6V or 12V motor on hand, what we did have was a NEMA17 stepper motor. The stepper motor is wired very differently to a DC motor and in general works much better as a generator. It has multiple coils per pole, and two poles that can be controlled individually in normal use.

Configured as a generator, the wiring diagram for the motor was first used to make sure we knew which wire was which, then the coils were connected in phase. Then, the motor was spun by hand, with much better results. As for the DC motor, the oscilloscope was connected straight across one coil (yellow trace) then after a bridge rectifier (blue trace). The second image shows the two poles connected to their own probes, while the third shows both feeding into one power rail through their own bridge rectifiers.

The advantage in using a DC motor, from our perspective, is that diodes in a bridge configuration could be used so that no current reaches the load as the plunger is pulled upwards: The motor spins each way, so pulling the plunger upward would energise a load connected with no rectifier. This is only a problem for the very simplest version, however, and we eventually abandoned that anyway.

That very simple version just used a small DC-DC converter module with a 5V output and a high enough input tolerance to cope with any spikes from the motor. Because the stepper produces an AC signal, we lose this ability in using a stepper but as noted, in every other case it does not matter, as a switch is involved at the end of the down stroke.

We had several directions in mind for the project. One was as a digital 5V output, which could activate a microcontroller project such as a huge Christmas lights display, or anything else you want to ceremoniously and dramatically start. This would require a smooth 5V output. Another direction was for the launch of model rockets. Again, this is largely ceremonial and dramatic, as real rocket launch consoles for model rocketry are usually quite tame in comparison.

This would require a decent charge to be built up, then dumped at once. The electric matches commonly used in model rocketry require around 2A to fire reliably but more is considered desirable in the model rocket community. They are less voltage-dependent than current-dependent, but voltage helps get the current to where it is needed! Finally, we envisage a stand-alone device for use as a game objective, such as at Nerf events or paintball. Pushing the plunger would activate a siren or light or both so that the referee and other players are in no doubt that the game has ended.


All of these ideas have different requirements, but they can be built in together. Firstly, capacitors are required to store the charge. There is no use trying to make (and more importantly, balance) a flywheel when modern capacitors will do the job for us so easily. We therefore set about testing the stepper with more detail to see what the highest voltage involved would be, to determine what voltage rating of capacitor we would need. That in turn meant making the drive mechanism that we were going to use, so we shelved the electronics for a bit and switched to the mechanics.

Our initial prototype drive mechanism consisted of a 3D printed rack, engaging with a matching pinion, in a housing that mounted the NEMA17 motor. The pinion had a flat to match the shaft of the stepper motor, and the rack slid in a guide. The immediate problem we encountered was that the tolerances were not what we had calculated, either through our own error or printing tolerance, or both. The rack was a bit hard to drive in and out, and was 'jumpy'. Therefore, we set about reworking the whole system and including the use of another gear at the same time.

The improved mechanism makes use of an adjustable guide for the rack, which has spring compression applied to maintain contact with the pinion. The stepper motor housing is also now on a slider, which also has spring compression applied. This ensures smooth engagement at the correct pressure between the pinion on the stepper motor and the outer section of the large gear, which is the only fixed element.

The gear itself is a 1:4 step up, so that one rotation of the smaller section results in four rotations of the motor. This means the motor spins four times more than the first version for the same length of rack. Finally, the mount has been divided into two parts to facilitate assembly and fitting of the springs.

With an on-bench test confirming this setup would be at least viable enough to refine, we delved into the electronics. That same bench testing revealed that the stepper, in the mount as designed, outputs peaks of 32V. We could use an LM7805 linear regulator to give us 5V out, but there are two problems: Linear regulators dissipate all overhead as heat, meaning that extra voltage is useless.

In addition, linear regulators have a minimum overhead, under which they lose regulation. For the LM7805, the minimum input voltage is 7.5V to maintain a stable 5V output. Under 7.5V, the output does not necessarily decline under 5V: Because regulation is lost, it can actually increase, far enough above 5V to be a risk to some 5V logic.

Instead, we opted for a DC-DC step up/ step down converter. These can provide a stable output voltage from their maximum, down to around 1V or so for some models. This way, the maximum practical amount of stored charge is converted to usable output. The lower the load, the longer this will last.

The challenge is that the maximum voltage input for this device is 30V for the $45 one, or 22V for the $11.50 version. That means if our stepper does produce more than 30V or 22V, then we need a way to limit it. That will probably end up being a Zener circuit of some sort, but we will approach that later. The 32V we got was an absolute peak with a very firm push -we feel it would be better to simply add an extra gear or add some friction so the plunger cannot be pushed at the higher speed.


The prototype design consists of a NEMA17 stepper motor driven by a rack and pinion through a 1:4 step-up gear. The outputs of both poles of the stepper have their own bridge rectifiers, the outputs of which are fed to a 1000μF 50V Electrolytic capacitor. This was unnecessarily high on the voltage but at this point, we had no idea how much voltage and current the stepper would produce. We had been holding the motor and spinning it by hand or at most, rolling the rack over the pinion on the bench. Across the capacitor was a 10kΩ resistor, both to bleed the current between tests and also to ensure at least a light load.

To house the motor and gears, we designed a 3D-printed structure with adjustability. It consists of: Gear half, which has a bracket to hold the gear, as well as a spring-loaded tensioner for the rack; the gear itself; the rack tensioner insert; the motor housing half with sled track for the motor housing and spring holder; and the motor housing with sled base. These are described in more depth in the '3D-Printed Parts' section.

We found some scrap timber, which was a rough base made for science demonstration (hence the odd looking structure in the images), and cut a hole in it for the plunger rack. Over this, we screwed down the 3D print that holds the gear, with the opening positioned over the drilled hole in the wood. We manipulated the tensioner insert into place and used pliers to add the springs. The gear rotates on a 6mm aluminium shaft, which is a by-the-metre product from a local hardware store. We cut a length off and used glue to hold it into the bracket, with the gear spinning free.

The base is in two halves because there is no way to add the spring tension otherwise, and that will make more sense once we can show it assembled. The springs were installed on the pegs on the motor housing, which was slid on its sled into the track. With the other ends of the springs over the pegs on the spring holder, the motor housing was pushed to the end and the whole section screwed down firmly. Lastly, the pinion gear was installed on the stepper motor shaft, which has a 25% flat on it to hold the gear firmly.


We did not have a handle cut yet for the rack, but it was still usable. We clamped the test rig to the workbench and inserted the rack to mesh with the smaller section of the gear. We checked to make sure the larger section was meshing with the pinion on the motor, and hooked up the oscilloscope to measure the voltage across the capacitor. We also hooked up a multimeter set for current, between one motor pole and its bridge rectifier. This is where things fell apart brutally for us. We pushed hard on the plunger, and the voltage across the capacitor rose as expected. However, we barely saw any reading on the multimeter. We checked settings, and we had set to AC, which was correct for the output of the stepper. We tried again, and then dropped the resolution, and the reading was still low.

We tried by measuring voltage drop using an oscilloscope and a 10Ω resistor, and confirmed that the motor was outputting around 4 to 10mA! That is devastatingly small for a motor which is rated to 1.7A at 12V when used as a motor. While we have little experience with generators, the fact that some bicycle dynamos appear to be just a regular 12V DC motor in a geared housing and can run small bike lights, we thought we would get more current than 10mA! So, the unit was never going to produce enough current to fire a model rocket motor or sound a buzzer.

Then, we realised something, and this is the reason we chose to detail our failings here. No matter the experience, we can all miss things. The capacitor we were using had a 10kΩ resistor across it for discharging. However, we realised that the way this capacitor and resistor combination loads the rather cheap multimeter we happened to be using was falsifying the numbers: It was not reading accurately. The oscilloscope measurements we got were also low because of a mistake in how we connected the circuit with the capacitor in parallel.

We stripped out the capacitor, hooked up a better multimeter, and tried again with a 100Ω resistor as a load. Now, we were seeing peaks of 200mA or so, although they were brief and hard to read so the tens and units places could not be read. We hooked up the same 100Ω resistor with the oscilloscope probes across it, and then by measuring the voltage loaded (unlike the 10kΩ resistor we used for our previous voltage measurements which placed minimal load), we calculated that the current to peak at about 300mA and average around 240mA.


The 240mA average is ok, but still not what we really wanted for the launching of model rockets and the running of a siren. However it is realistic for most things. So, we proceeded with the original plan, just at a smaller current capacity. Therefore, we chose the $12 1A Pololu DC-Dc converter rather than the $43 2A version. There are even lower current versions too, but the maximum input voltage is also lower. The input voltage of 22V max for the 1A version needs attention, but the generator is capable, at extremes, to produce more than the 30V max of the more expensive unit, so anyone building this with the bigger converter needs to do something about max voltage.

Our solution was to use the LM317 linear regulator. This is a rugged device with lots of built-in protections for itself and a max input voltage of 40V. This will be set, with a resistor value according to the datasheet, for a 19V output. The LM317 has a drop-out voltage of 2.5V, meaning if the input voltage is 2.5V or less greater than the output voltage, regulation is lost. This is ok, because the Pololu DC-DC converter that forms our second stage has an input maximum of 22V. If regulation on the LM317 is lost at 21.5V, we're still safe. We said earlier that linear regulators are wasteful, but we are only using it to cut the top off the voltage available. From 21.5V down to 2V across the capacitor, the DC-DC converter will be outputting a steady 5V with nothing lost on the regulator.


When we initially thought we were going to get a max of 10mA or so from the stepper motor, we felt we could salvage the project as-is. Our idea of using the Pololu DC-DC converter units was gone, but we could still make a project that can give a digital high signal to an external microcontroller (or any other logic circuit), provided the impedance is not too low. We can do this easily with batteries and a microswitch at the end of the plunger travel of course, but we wanted to keep our original criteria of no batteries or external power. Therefore, we had a revised plan. The schematic shows a very simple Zener-based regulator circuit capable of outputting a 40mA maximum signal at 30V (based on the Zener's current-limiting resistor), down to 6.8mA at 5.1V. Using Zeners with wide input voltage ranges has its challenges, but there was no other practical, simple, economical way to do what we wanted to do. While this rules out the other things we wanted to do, like having enough energy to sound a buzzer, it does achieve our primary goal of activating a microcontroller.

We included this circuit and discussion because we feel there are people who only want a very simple digital controller and may not want to go to the expense of buying the DC-DC converter and having it posted.


The capacitor value chosen has an impact on how the circuit works because of the relationship between voltage and current in a coil. The stepper motor run as a generator produces its maximum voltage unloaded, open-circuit. It produces its maximum current with a total load, and a short circuit. All capacitors have some form of internal resistance, and this forms a load on its own. Ideally, we want a big enough capacitor to keep that maximum voltage under the maximum of our regulator. We also want a big enough capacity to store all the power we can generate, but to create a decent load, the capacitor chosen will be bigger than would be needed for storage when that is considered alone. For example, we measured a maximum value of 32V from our stepper with a 1000µF 63V standard electrolytic capacitor, but a value of only 18V with a 4700µF 50V version. We retained the LM317 in our design in case of unforeseen factors that might increase the voltage above 22V, like someone slamming the plunger down rather than pushing it.


At some point in our early research process, we had bought one of Jaycar's dynamo modules. This is aplastic frame with some gears, a crank handle, and what appears to be a 4.5V or 6V small motor attached. We decided to see how it compared to our stepper motor version, with some open- and closed-circuit tests. We did not use a capacitor, because the test would be unfair: The stepper is in its plunger rig, while the dynamo has a continuously-rotating winder and could therefore charge a capacitor differently. We used an oscilloscope for open-circuit voltage measurements, and the better multimeter we had switched to earlier for nearly short-circuit current measurements, with the only load being the multimeter's current shunt.

The stepper motor achieved a short-circuit current of 425mA AC peak, with an open-circuit voltage of 70V AC. The dynamo achieved a short-circuit current of 220mA AC peak with a value of 120mA being more constant, with an open-circuit voltage of 5V AC with short spikes of noise up to around 20V which were only a hundredth of a cycle or so wide. This confirmed our choice of the stepper motor as our generator.


Parts Required:Jaycar
1 x Small piece of VeroboardHP9540
2m 7.5A Hookup WireWH3046
1m Light Duty Hookup WireWH3011
1 x 620Ω ResistorRR0567
1 x 9.1kΩ ResistorRR0595
1 x 1MΩ ResistorRR0644
1 x 4700µF 50V Electrolytic CapacitorRE6245
2 x Bridge RectifiersZR1320
1 x LM317ZV1615
1 x 5V 1A Pololu Step Up/Step Down ConverterCore Electronics POLOLU-4083
2 x Black Binding PostsPT0461
5 x PCB PinsHP1250
4 x PCB Pin SocketsHP1260
1 x Limit SwitchSM1040
4 x springs for tensioning 3D printed partsHP0638

We decided to mount the electronics, what little there is of them, on Veroboard. This enables solid connections for the components but also facilitates header connections for easy connection of parts as we install them. So, the board had on it two bridge rectifiers, one big capacitor, the LM317 regulator and its resistor, and space to glue on the Pololu regulator.

There were PCB pins for the limit switch that will connect the Pololu regulator to the circuit when the plunger reaches the end of its travel, and for the motor to attach. There are also pins to attach the wires from the binding posts on top once the unit is complete, and also one pin so that, if needed, the regulator section can be bypassed and the capacitor charge dumped as-is to the output terminals by the limit switch. Overall, it is not complex, so we have a layout diagram to show the board arrangement. There are no track cuts, but there are some wire links and some component legs that stretch a long way.


The 3D printed parts for this project are centred around the drive mechanism. Because of space constraints, we have the main assembly in parts, so that the spring loaded portions can be assembled then the halves brought together. Designing them such that the springs could be installed once assembled would mean a much bigger print. We used three shells/walls, and four top and bottom layers, printed at 0.3mm resolution except for the tensioner, at 0.1mm and ten top and bottom layers.

This half of the housing holds the gear and its shaft, plus the guide for the plunger rack with a tensioner to push it onto the gear. We printed it as shown in the screenshot. Support is needed.

This half of the housing contains the track for the moving motor housing, and a bracket to hold two springs to push against it. Support is needed, and we used linear support to make it easier to remove from under the tracks.

The larger gear has four times the teeth on the larger circumference than on its smaller circumference. The pinion gets fitted to the motor shaft, and is the same as the smaller part of the big gear, but with a 25% flat in the shaft hole.

The tensioner has two holes in the rear face, the one on the bed in the image, for springs. These holes need support and whether or not you need a raft or prim for that depends on your printer. Ours, a Flashforge Guider 2S, printed well with no raft or brim, with the support directly on the build plate.

The motor housing has countersunk holes for M3 screws for the motor, and a sled to engage the track in the base. It also has pins to accept the springs used to push it against the gear. We printed it as shown, using support for the countersunk holes, the opening, and the interior in general.

The limit switch housing goes below the plunger travel. We have bevelled the top to guide the plunger in, and set the limit switch at a slight angle so that the lever engages well with the plunger without placing undue stress on the switch. We needed support to print it, because of the pins for the limit switch.

The plunger rack is printed flat on its back as shown. Support will be needed for the hole in one end, which is designed to take an M3 heat insert, for attaching a handle.







Front and Rear Panel





Side Panels





Top and Base Panels





The box to house everything is relatively straightforward in assembly, thanks to the utilitarian nature of the inspiration examples. While woodworking of the time could be taken to a high art with intricate joinery that did not even require nails and glue, such craft was wasted money on an item like this. Consequently, we have chosen basic pine, with cut blued steel tacks to join the pieces. Nails of the time were similar to this or were copper nails that were easier to manufacture.

We started by cutting the side and top panels. The Front and back panels are made from 140mm wide pine, which necessitated a rework of the 3D print compared to the prototype. The depth is determined by the plunger travel length, and the width by a convenient size of pine. All panels are 11mm thick. The cutting list has the exact sizes.

Before nailing the panels, we first stained them with a Walnut stain, then hit the wood with a pair of old socks, one inside the other, filled with a variety of heavy objects: Nuts and bolts, which have hard edges, and fishing sinkers, for soft edges. This gives an aged look after the final treatment. We also used random tools like a screwdriver, sandpaper, and hammer to give more marks.

Following this, a sealer is applied, and when dry, Black Japan stain was rubbed with a cloth all over the panels, being allowed to settle into all the wear marks in varying degrees. When dry, this was coated in matte-finish varnish.

The side panels were assembled around the base plate, with the short sides butt-jointed to the ends of the long sides, maximising the internal width. Clamps were used to hold the back and one side to the base, then nails were driven in. Then, clamps again were used to hold the front panel while it was nailed in place.

Before the final side was added, the lid needed to be built. This houses the generator. Accordingly, the 3D printed parts were screwed on, then the opening for the plunger was marked. Then, the parts were removed and the hole for the plunger cut with drilling and filing. Two holes were drilled for binding posts. Then, the 3D printed parts were reattached and two black binding posts installed. While modern plastic binding posts are not period-correct, they look as close as we can get to some of the black or brown Bakelite connectors that were used when metal ones were not. These had fly leads attached for later connection to the circuit board.

The lid was placed home and clamped before being nailed into position. The plunger was installed and the 3D-print for the limit switch was positioned and glued into place. Once this was in, the circuit board with the rectifiers, regulators and capacitors was glued to an out-of-the-way spot on one of the side panels where it is clear of the plunger path. The wires from the motor were plugged onto the header, as was the limit switch.

All that was left to do now for the assembly of the main box was to place and nail the last side, enclosing the unit, and to apply some more Black Japan and varnish to the nails to blend them and age them. However, the handle still had to be made.

The handle was made by slicing off a piece of 18mm Tasmanian Oak dowel to the right length (which is really a preference based on the size of hands), with a 3.5mm hole drilled into the centre. One side of this hole was countersunk, while on the other side, a flat spot was filed for the rack to sit. Before the rack was attached, the handle went through the same treatment as the box sides for colour and ageing.

We used a soldering iron to place an M3 threaded insert into the hole in one end of the plunger rack. Then, the handle was screwed onto the rack with a countersunk M3x25mm screw. We wanted to give a better finish for the rack, to make it look like old metal, but the reality was any finish would wear off the high-friction sections. We kept ours plain black for now, and later on when we find the right filament, we will print it in aged brass or bronze or something like that.

NOTE: For attaching the top plate, which houses the motor, we used 4Gx25mm brass screws so that we can remove this in the future. Screws were not common in the period, being far more expensive than nails in the earlier history of blasting machines (although common later when some machines even had cast metal cases). However, brass can be aged to look very old quite easily and when covered with Black Japan, will bend in just fine. Interestingly, some wooden blasting machines did use screws for assembly, although they were slotted/flat blade types reflecting the era.


With the unit assembled, it was time for testing. We hooked an oscilloscope to the output terminals, and pushed down hard on the plunger. Sure enough, at the end of its travel, we were suddenly greeted with a jump to 5V, which stayed constant for almost exactly ten seconds. This was definitely good enough to use as our digital switch, as intended.

The next test was to hook up a 5V to 15V piezo buzzer. This unit will be at its quietest at 5V but it was still worthwhile as a test. To our surprise, it sounded for eight and a half seconds, much longer than we expected given the charge lasted ten seconds with only the oscilloscope (and the regulator, of course) for a load. This means that it can be used as a game objective as-is!


This machine is far from perfect. Ideally we would like to have factored in a ratchet or clutch system so the generator can spin freely after the plunger reaches the end of its travel. As-is, however, the NEMA17 is not really big enough to have the inertia for a sudden stop to be a problem. There were also some lessons to be relearned as we described while testing the motor outputs, namely use quality equipment and always look at the basics! They can be easy to forget sometimes. Such a system is not that easy to make strongly, reliably, and compactly in 3D-printed PLA at this scale.

Overall, the design worked out usably but we are disappointed in the lower current and output, although we really should have expected this. Getting around 250mA consistently and 300mA to 350mA at peak for the time we got it to charge the capacitor is not bad considering the inefficiency of using a motor backwards, and the 12V NEMA17 we used had a max input of 1.7A when used as a motor. We would like to make something that can generate and store enough current to run, say, a car alarm siren for a few seconds. This would be perfect as a game objective, like in a youth group Nerf event, or even for paintball or laser tag. However, being able to run a 5V piezo for a bit over eight seconds is a great start.

In terms of physical construction, we would make a bigger box if we were to build this again, and put more time into a gear mechanism that centred the plunger. As it is, the tensioner was too strong and kept pushing the plunger sideways. We needed to make a new 3D printed part which guided the plunger from on top of the box, and cut some coils from the tensioner springs.


To continue developing this project, we would like to rebuild it with a bigger generator. As you can see in Our Own Devices' and Big YouTube videos about blasting machines, which feature units opened up, the generator is much, much bigger. While we do not need the more than 90V that one of these units produced, a bigger motor would gain us more current, likely enough to run a siren for a few seconds.

Additionally, we would like to pursue a dedicated generator of some sort, possibly a push bike dynamo or something like that, which can have a flywheel attached and a ratchet or clutch to disengage the drive gears to keep spinning when the plunger reaches the end of its travel. Further improvements would be more gear stages as well, to increase the rotation of the motor/generator for the same length of plunger travel.

Another direction we may take the build, and possibly publish as a mini project, is to build a rotary-handle version. This stands a far better chance of being able to launch model rockets which use pyrotechnic engines and electric match igniters drawing around 2A or more for a fraction of a second. While having much less of a Looney Tunes feel, they are still historically realistic and would be a viable battery-free way of launching model rockets, giving a user a better feeling of being part of the launch. Such a machine would feature a pushbutton switch to dump the accumulated charge.