.png)
This is the authoritative, illustrated, construction procedure for the Magnet Mobil Mk II. The complete parts list is included below so the guide can be used on its own. For information about the project history, design reasoning, and physics, read the blog post.
Parts and Tools
The quantities and dimensions below describe the configuration that successfully drove the prototype. One single electromagnet is installed: the approximately 460-turn, 4 Ω winding specified below.
The supplier links are searches rather than recommendations for individual listings. Listings often disappear, while the search terms and specifications remain useful.
Printed parts
Available for download at Thingiverse
Part | Qty | File | Notes |
Chassis base | 1 | stl/print_base.stl | Includes the bearing support, axle mounts, and electromagnet holder |
Control arm | 1 | stl/print_arm.stl | Carries the reed switch near the flywheel |
Electromagnet bobbin | 1 | stl/print_bobbin.stl | Wind with approximately 0.3 mm enamelled copper wire |
Flywheel | 1 | stl/print_flywheel.stl | Standard CAD setting produces six magnet pockets; twelve pockets are optional |
Road wheels | 4 | stl/print_roadwheels.stl | 25 mm diameter; one wheel has a drive groove |
Pulley cones | 2 | stl/print_pulley_cones.stl | Fit to the flywheel shaft to form a small V pulley |
Mechanical parts
Part | Qty | Specification | Supplier search | Notes |
Flywheel shaft | 1 | 1.5 mm smooth steel rod, approximately 60 mm long | Also carries the pulley cones | |
Road-wheel axles | 2 | 1.5 mm smooth steel rod, approximately 65 mm long each | Check the finished chassis and wheel spacing before cutting | |
Flywheel magnets | 6 standard; 12 optional | Nominal 8 mm × 3 mm neodymium disc magnets; verify fit in the printed pockets | Fit one magnet in each pocket of the standard six-pocket flywheel; populate alternate pockets if using six magnets in a twelve-pocket flywheel | |
Drive belt | 1 | Simple rubber band, approximately 15 cm total circumference when unstretched | Use only enough tension to prevent slipping without adding excessive axle friction | |
Magnet adhesive | As needed | Flexible, fast-setting UHU glue, as used on the successful prototype | Its flexible bond suits the rotating flywheel; use sparingly to avoid imbalance |
Electromagnet parts
Part | Qty | Specification | Supplier search | Notes |
Enamelled copper wire | Approx. 20 m | Approximately 0.3 mm diameter copper magnet wire. The enamel is important, as it acts as an insulator. | Wind until the bobbin is full; approximately 4 Ω and an estimated 460 turns describe the successful winding, but are guidance rather than fixed limits | |
Electromagnet core | 1 | 8 mm diameter soft iron or low-carbon steel, approximately 30 mm long | Avoid stainless steel unless its magnetic response has been confirmed |
Electrical parts
Part | Qty | Specification | Supplier search | Notes |
Reed switch | 1 | Normally open glass capsule, approximately 14 mm long × 2.2 mm diameter | Switches the coil current directly; verify its contact-current rating | |
Power switch | 1 | Simple SPST on/off switch | Disconnects the battery when the model is not running | |
PCB | 1 | Order from the newest Gerber archive in gerber/ | Current file: Gerber_MagnetMobil_PCB_MagnetMobil_2026-08-10.zip; the circuit has no resistor, transistor, or MOSFET | |
Two-pin connector pair | 1 | Keyed JST pair preferred; another suitable two-pin connector may be substituted | A correctly wired keyed connector prevents accidental polarity reversal; clearly mark and verify the polarity of any substitute | |
Battery holder | 1 | Two-AA holder wired in series | Provides approximately 3 V with two alkaline cells | |
AA cells | 2 | Matched cells of the same type and state of charge | Two alkaline cells provide approximately 3 V in series; two NiMH cells provide approximately 2.4 V nominally | |
Flyback diode | 1 | 1N4148 | Fit across the coil with its striped cathode toward the positive supply side; pulse-current suitability still requires validation | |
Hook-up wire | As needed | Flexible small-gauge wire | Used for the battery, coil, and reed-switch connections | |
Heat-shrink tubing | As needed | Small-diameter tubing sized to fit the battery leads. Alternatively: Electrical tape. | Insulates the soldered connection between the battery holder's negative lead and the connector's negative wire |
Tools and consumables
Item | Notes |
3D printer | PLA or PETG is a reasonable starting point |
Soldering iron and solder | Required for PCB, coil, switch, and reed-switch connections |
Flush cutters and small pliers | Used for component leads and wire preparation |
Sandpaper | Used to remove enamel and clean tight printed fits |
Clear adhesive tape | Separates coil layers and secures the completed winding |
Double-sided soft foam-core tape | Two small strips attach the reed switch to the control arm while leaving most of its glass body free; additional tape attaches the PCB to the base |
Hot-glue gun and glue | Attaches the battery holder and power switch to the base |
Multimeter | Used for coil-resistance, polarity, and reed-switch tests |
Safety glasses | Recommended when cutting wire or steel shafts |
Variable-speed power drill | Optional aid for winding the electromagnet; use only at a slow speed |
6–8 mm machine screw and washer | Optional winding mandrel; the head or washer must retain the bobbin securely |
1. Print the Parts
Download the 3D print files from our Thinkgiverse pages. Print the required STL files in stl/, or regenerate them from the SCAD files:
stl/print_base.stl
stl/print_arm.stl
stl/print_bobbin.stl
stl/print_flywheel.stl
stl/print_roadwheels.stl
stl/print_pulley_cones.stl
Suggested starting print settings:
Material: PLA or PETG
Layer height: 0.12 mm, as used for the successful prototype
Perimeters/walls: 3
Infill: 20-35%
Supports: avoid unless your slicer identifies a local issue
Check all axle holes and bores after printing. Ream the rotating fits lightly if needed so the flywheel and road wheels turn freely; even modest binding at these points can prevent the motor from running reliably. Avoid enlarging holes that are intended to remain a push fit.
The SCAD model also provides dedicated *_clearance settings for these rotating fits:
road_wheel_axle_clearance controls the road-wheel axle bores through the chassis flanges.
flywheel_bearing_tube_clearance controls the inside diameter of the flywheel bearing tube.
Increase the relevant value if the printed bore is too tight, or reduce it if the bore is too wide. Then regenerate the affected STL and print it again. This is useful when changing printer, material, layer settings, or extrusion calibration, all of which can alter the finished bore size.
2. Wind the Electromagnet
Print stl/print_bobbin.stl and wind the coil directly onto the bobbin. This can be done by hand, although a slow-running power drill makes the long winding process easier. Start with approximately 20 m of 0.3 mm enamelled copper wire. A simple resistance calculation suggests about 16.4 m in the finished winding, while the remainder allows for leads, winding waste, and wire variation.
The bobbin dimensions in the current CAD are:
Overall length: 18 mm
Flange diameter: 16 mm
Barrel diameter: 9.6 mm
Winding length: 14 mm
Core bore: approximately 9 mm for an 8 mm core plus clearance
Optional drill-assisted winding
Thread a 6–8 mm diameter machine screw through the bobbin bore. Use a screw with a head wide enough to retain the bobbin, or add a washer under the head so the bobbin cannot slide off. Clamp the threaded end securely in the drill chuck with the bobbin held against the chuck, but do not tighten the arrangement so strongly that the printed flanges are crushed.
Set the drill to its slowest practical speed. This works best with two people: one person operates the drill and can stop it immediately, while the other guides the wire back and forth across the bobbin with light, even tension. Keep fingers, loose clothing, hair, and the supply spool clear of the rotating chuck and screw. Do not attempt to correct a misplaced turn while the drill is moving; stop first and reposition the wire by hand.
The first layer is comparatively easy to guide because the red enamelled wire is clearly visible against the white bobbin. Before winding the second layer, cover the first with one neat layer of clear adhesive tape such as Scotch tape. Add another tape layer between subsequent wire layers. These intermediate layers hold the completed turns in place and create a smoother, more visible surface for guiding the next layer. Keep every tape layer inside the flanges and do not allow folds or overlaps to build up unevenly.
Wind the wire evenly, leaving enough free lead to reach the PCB, and stop when the winding space between the bobbin flanges is full. Keep the finished winding inside the flanges. The enamel insulates the wire, so a multimeter cannot continuously measure the winding by probing its outer surface. Do not scrape intermediate turns merely to check progress, as this creates damaged insulation within the coil.
After winding, scrape or sand the enamel completely from both leads. Short the multimeter probes together and note their resistance, then measure the finished coil and correct for the probe resistance. The successful winding measured approximately 4 Ω and is estimated at roughly 460 turns, but 4 Ω is guidance, not a fixed acceptance value. A higher resistance is acceptable. Install this as the model's single electromagnet.
After checking the finished winding, cover it with one neat layer of clear adhesive tape (such as Scotch tape). The tape keeps the outer turns from loosening or uncoiling during assembly. Keep it inside the bobbin flanges and do not cover the core bore.
Fit an approximately 30 mm long, 8 mm diameter ferromagnetic core through the bobbin. Soft iron or low-carbon steel with high magnetic permeability is preferable to hardened steel; do not assume that an unidentified stainless- steel rod will work well. The core material has a large effect on the strength of the electromagnet, so record the source and material of any successful core.
3. Assemble and Mount the Circuit
Order the PCB using the newest Gerber archive in gerber/. At the time this guide was written, the newest archive is: Gerber_MagnetMobil_PCB_MagnetMobil_2026-08-10.zip
Use the circuit diagram while assembling the board. The two-AA series battery pack, separate power switch, coil, and normally open reed switch form a single series path. The 1N4148 flyback diode is connected across the coil, with its striped cathode toward the positive supply side. There is no resistor, switching transistor, or MOSFET in the standard circuit; the reed switch carries the pulsed coil current directly.
The N.O. label in the diagram means normally open. With no magnet near the reed switch, its contacts are separated and no current flows through the coil. A passing flywheel magnet closes the contacts and energises the coil.
Fit the flyback diode and power connector
Install the 1N4148 flyback diode at D1 before adding the wires. Bend its leads to fit the holes without pulling on the glass body, place it close to the PCB, and align its striped cathode with the polarity shown in the circuit diagram. Solder both pads, inspect the joints, and trim the excess leads. This diode provides a path for the coil current when the reed switch opens.
The prototype uses a keyed two-pin JST connector at VCC so the two-AA holder can be disconnected. A keyed JST connector is preferred because, once wired correctly, it prevents the plug from being reconnected with reversed polarity. If JST connectors are difficult to source, any suitable two-pin connector can be used instead. Mark its positive and negative sides clearly and verify the polarity with a multimeter before connecting it to the PCB.
Fit the connector square to the board and follow the + and − markings; a keyed connector prevents accidental reversal only if its mating lead has been wired to the same polarity. The REED and L1 pads remain empty until their leads have been prepared and routed on the chassis.
Wire the battery holder and power switch
The battery holder's negative lead connects directly to the negative lead of the mating JST plug. Cut both wires to suit the chassis, slide heat-shrink over one lead before joining them, and twist the stripped conductors together. Keep the two positive leads separate at this stage because the power switch will be inserted between them.
Solder the negative splice, inspect it for complete wetting and mechanical strength, then cover the joint with heat-shrink so it cannot short against the switch or PCB. Support the wire while soldering rather than using the hot joint itself as a mechanical anchor.
Mount the toggle switch in the chassis and wire it in series with the positive battery lead. If the chosen switch has three terminals, use a multimeter to identify the pair that closes in the ON position. The photograph also shows the two-AA holder and control arm fitted to the base; these parts should be secure but must not obstruct the flywheel, axles, or later wiring.
Check the following before continuing:
Two-AA battery-pack polarity and series wiring
1N4148 orientation, with its striped cathode toward the positive supply side
Separate power-switch connection
4. Prepare the Reed Switch
Cut two leads approximately 20 cm long. This provides enough wire to run from the end of the white control arm to the PCB without being stretched and leaves some extra length for positioning and final routing. The prototype uses thin enamelled copper wire, which is light enough not to resist movement of the arm; fine flexible insulated wire can also be used.
Remove approximately 10 mm of enamel or insulation from one end of each lead. Fine sandpaper works well for enamelled wire: turn the wire while sanding until bright copper is visible all around. The exposed length allows the wire to be wrapped around the reed switch's metal lead before soldering.
Wrap one stripped wire end around each reed-switch lead, then solder the wrapped joint. The mechanical wrap holds the wire in place while soldering. Keep the soldering time brief and avoid bending a lead sharply where it enters the glass capsule; the glass-to-metal seal is fragile. The reed switch is not polarised, so either wire can later connect to either reed pad.
Before mounting the reed switch, connect a multimeter in continuity or low- resistance mode across the two exposed solder joints at the reed-switch end. Bring a loose magnet near the glass capsule and confirm that the resistance falls when the contacts close, then rises again when the magnet is removed. This checks the reed switch and both wrapped solder joints without requiring the far ends of the wires to be stripped yet.
Apply two small strips of double-sided soft foam-core tape to the top of the arm, one to the left and one to the right of the reed switch. Arrange the strips so they retain the switch near its ends while leaving most of the glass capsule free. Do not place one continuous strip beneath the complete glass body.
The tape is the permanent mounting method. Position the switch only lightly at this stage so its sideways offset can still be corrected after the flywheel has been installed.
5. Prepare the Flywheel and Mechanical Chassis
The flywheel can be generated with either six or twelve magnet pockets. The pocket envelope in scad/assembly.scad is approximately 8.5 mm x 8.5 mm x 2.8 mm, which should be checked against the magnets actually used.
The standard build uses six magnets, one in each pocket of the six-pocket flywheel. This keeps the flywheel balanced and matches the six-pulse spacing of the original Magnet Mobil. The photographed successful prototype uses the optional twelve-magnet configuration.
The current scad/assembly.scad setting has flywheel_extra_magnet_cradles = false, which generates the standard six-pocket rotor. Set this option to true to generate twelve pockets. A twelve-pocket wheel can still be loaded with only six magnets by populating alternate pockets, so pocket count and installed magnet count are related but separate choices.
Before gluing magnets:
Confirm every magnet sits below or flush with the flywheel surface.
Mark magnet polarity so all pockets are loaded consistently.
Test-fit the flywheel on the 1.5 mm shaft.
Spin the flywheel by hand and check for rubbing or wobble.
The successful prototype used flexible, fast-setting UHU glue. Use only a small amount; excess glue near the rim can unbalance the flywheel.
Fit the following parts:
1.5 mm flywheel shaft through the printed bearing tube
Flywheel on the left side of the bearing
Drive pulley cones on the shaft
1.5 mm road-wheel axles through the chassis flanges
Four printed 25 mm road wheels
Rubber-band drive belt left off for the initial mechanical and powered tests
Electromagnet bobbin in the printed holder
White control arm, with the normally open reed switch held on its cradle by the double-sided soft foam-core tape
Set the electromagnet core as close as practical to the magnet path without allowing contact anywhere in a full revolution. The mechanism should move freely before power is applied. Fix mechanical drag first; extra electrical power will not compensate for a binding flywheel or misaligned wheel.
6. Position and Connect the Reed Switch Without Power
Start with the complete control arm approximately 120° clockwise from the electromagnet, viewed from the flywheel side of the model. Place the glass capsule on the tape with a small sideways offset in its cradle on the white arm. If it is centred incorrectly, it may not close at all or may respond to only some magnets. Both positions remain adjustable: establish reliable switching without power first, then fine-tune the arm angle during the powered test.
Position the reed switch as low as practical, so it is as close as possible to the path of the neodymium flywheel magnets. The smaller gap exposes the switch to a stronger magnetic field and helps its contacts close reliably. It must not touch the flywheel or any magnet at any point: rotate the flywheel through a complete revolution and leave enough clearance for minor wobble.
Route the fine leads along the arm with enough slack at the pivot, and keep them clear of the flywheel rim. With the battery still disconnected, turn the flywheel slowly and watch the meter. Every installed magnet should close the switch; an inconsistent reading usually means the capsule needs more sideways offset or a small change in arm angle.
Once the switch closes reliably for every installed magnet, press the two supported areas firmly onto the tape without pressing on the centre of the glass capsule.
When the position is confirmed, cut the leads to a practical length, remove their insulation at the ends, and solder them to the PCB without disturbing the capsule's tested offset.
Solder the two prepared ends into the PCB pads marked REED. Provide light strain relief along the arm so movement or accidental handling is not transferred to the glass capsule, and recheck continuity through a full flywheel revolution after soldering.
Solder the two leeds of the electromagnet into the PCB pads marked L1.
Attach the PCB to the printed base with double-sided soft foam-core tape. Use small amounts of hot glue to attach the two-AA battery holder and power switch to the base. Keep tape and glue clear of connectors, solder joints, battery terminals, and moving parts.
7. First Power Test
Install two matched AA cells in the series holder for an approximately 3 V supply. This test assumes the successful approximately 4 Ω coil. Do not substitute a lower-resistance winding without first recalculating the expected current and verifying that the reed switch, diode, wiring, and cells can safely handle it.
Start with:
Flywheel spinning freely by hand
Drive belt removed, so the pulse motor is tested without drivetrain load
Reed switch already shown to close for all six magnets during the unpowered multimeter test
Electromagnet core close to the flywheel rim with a small air gap
Power switch off until all wiring and polarity checks are complete
Give the flywheel a gentle starting turn and switch on the power switch. The motor should pulse as each magnet approaches the electromagnet. If it only kicks once or twitches, switch off and rotate the white control arm by a small amount before trying again. If the coil pulls against the intended direction of travel, check the coil connections and magnet polarity. Correct timing and friction problems mechanically, and use only the specified two-AA series supply with the 4 Ω coil.
Once the unloaded flywheel runs reliably, switch off and fit a rubber band with an unstretched circumference of approximately 15 cm between the flywheel pulley and the grooved 25 mm road wheel. Use only enough tension to prevent slipping and check that the belt runs in line without rubbing the chassis.
With the cells removed again, check the final integration before the driving test: the coil leads are secured at L1, the reed-switch leads at REED, the battery holder is connected at VCC with the separate switch in its positive lead, and the rubber band runs around both the shaft pulley and the driven road wheel. Route every wire clear of the flywheel, road wheels, and belt.
With the power still off, make a final front-side inspection. Confirm that the electromagnet core has a small air gap throughout a full flywheel revolution, the belt remains seated and aligned, all four road wheels turn freely, and no wire can enter a moving part.
Inspect the reverse side as well. Check that every installed magnet is secure and flush, the flywheel rim does not rub the frame or electromagnet, and both road-wheel axles remain square and free-running. Rotate the drivetrain by hand for several revolutions before placing the model on the floor.
Place the model on a smooth, level surface, give the flywheel a gentle starting turn, and switch on. This 4 Ω, approximately 3 V configuration has successfully driven the complete vehicle.
Successful movement does not establish long-term electrical reliability. The actual pulse current, loaded battery voltage, coil temperature rise, safe runtime, reed-contact durability, and 1N4148 suitability remain to be measured. Stop the test if the coil, reed switch, diode, wiring, or cells become warm.
8. Troubleshooting
Check diode and battery polarity before applying power. The flyback diode must be reverse-biased during normal operation: its striped cathode faces the positive supply side. If either the diode is installed backwards or the battery input polarity is reversed—but not both—the diode becomes forward- biased when the reed switch closes. It then effectively short-circuits the battery through the diode and reed-switch contacts instead of sending current through the coil. Switch off and remove the cells immediately if the diode, reed switch, wiring, or batteries warm up, then correct both polarities before testing again. Do not deliberately reverse both connections as a workaround.
Work through problems in this order:
With the power off and belt removed, spin the flywheel by hand. It must coast freely without scraping or binding. Check the shaft, bearing, hub clearance, flywheel wobble, proud magnets, and electromagnet-core gap. A small magnetic detent near each magnet is normal.
If the flywheel is mechanically free but the powered motor only twitches, rotate the white control arm in small increments. Its angle determines when the reed switch energises the coil as a magnet approaches the core.
If the reed switch does not close reliably—or closes for only some magnets—move it lower, closer to the magnet path, while preserving clearance through a full revolution. Also adjust the glass capsule's sideways offset. Use a multimeter while turning the flywheel slowly and confirm that every installed magnet closes the contact before tuning the control-arm angle.
If there is still no magnetic pull:
Check that the coil wire ends are stripped and electrically connected. YOu may have accidentally soldered in the wire with the insulation part.
Measure the coil and confirm continuity and a plausible corrected resistance. The successful coil measured approximately 4 Ω, but a higher result is acceptable.
Check battery polarity and voltage.
Confirm the reed switch closes when a magnet passes.
Confirm the PCB switches current through the coil.
If the vehicle runs on the stand but not on the table:
Reduce road-wheel friction.
Check belt tension.
Make sure the driven wheel groove aligns with the drive pulley.
Try a smoother surface.
9. What the Earlier 2 Ω Prototype Taught Us
Before the successful electromagnet was wound, a first Mk II coil was made from the same approximately 0.3 mm enamelled copper wire. Its turns were not counted, but it measured approximately 2 Ω and is estimated at roughly 230 turns. Powered by one AA cell at approximately 1.5 V, it could run the unloaded flywheel but did not propel the complete vehicle reliably through the belt drive. It was removed and replaced by the 4 Ω coil; the two electromagnets are alternative prototypes, not parts used together.
The comparison is instructive:
Prototype | Estimated turns | Measured resistance | Supply | Theoretical DC current | Approximate ampere-turns | Maximum continuous-resistance heating |
Earlier unloaded-flywheel coil | 230 | 2 Ω | 1.5 V | 0.75 A | 173 A-turns | 1.1 W |
Successful driving coil | 460 | 4 Ω | 3 V | 0.75 A | 345 A-turns | 2.25 W |
The 2 Ω version needed less wire—about 8.2 m by the same resistance estimate—a single cell, and less winding time. Its lower turn count and inductance may also allow current to rise faster during a short pulse. Its disadvantage was the smaller magnetomotive force: at the same simplified peak-current estimate, it provided only about half the ampere-turns, and in practice it did not drive the loaded chassis.
The 4 Ω version needs roughly twice the wire, a two-cell holder, and more winding effort. Its higher inductance can slow the current rise, and at 3 V its worst-case continuous-resistance heating is higher. It may also store more magnetic energy for the flyback diode to handle. In return, doubling both the turn count and supply voltage keeps the theoretical current near 0.75 A while approximately doubling the ampere-turns—and this is the configuration that demonstrably drives the Magnet Mobil.
Simply adding turns without raising the supply voltage would not produce the same result. At 1.5 V, the 4 Ω coil would draw a theoretical 0.375 A, largely cancelling the ampere-turn benefit of the additional turns. All these current, heating, and ampere-turn figures are simplified theoretical values; the actual pulses still depend on inductance, battery sag, contact resistance, switch timing, and duty cycle.
A working, open-source pulse motor that makes electromagnetism, timing, and mechanical design visible, buildable, and open to experimentation.
TL;DR
The Magnet Mobil is a small vehicle driven by a visible electromagnetic pulse motor: flywheel magnets operate a reed switch, the switch pulses a hand-wound electromagnet, and a rubber band transfers the rotation to a road wheel.
It works and can be recreated in the classroom or at home. The illustrated build guide includes the complete parts list, circuit diagram, printable parts, coil-winding method, assembly sequence, testing procedure, and troubleshooting advice.
The parametric OpenSCAD files and other design files are provided in an open-source spirit. Makers can modify the geometry, experiment with the magnetic and mechanical design, or wire the simple circuit without the PCB if the specified parts are unavailable.
The prototype demonstrably drives, making it a useful hands-on STEM project.
1. Motivation & Background
I owned a YPS Magnet Mobil in 1980, and I loved it. What stayed with me was not speed or complexity, but the opposite: it made a motor from remarkably few parts and let me see how everything worked. A large red flywheel, six magnets, an electromagnet, a reed switch, a battery, and a rubber belt were enough to move the little vehicle across a table. There was no sealed motor and no hidden gearbox. The mechanism itself was the attraction.
The Mk II now drives successfully, and it still exposes the same simple mechanism that made the original memorable. Rebuilding it became both a return to a childhood object and an opportunity to pass on this gadget to the next generation of tinkerers.
A visible pulse motor
The Magnet Mobil is a small battery-powered model vehicle built around a visible electromagnetic pulse motor. Permanent magnets are fitted around the rim of its flywheel. As each magnet approaches, a switch briefly connects the battery to an electromagnet and gives the flywheel another pull. The flywheel's inertia carries it between pulses, while a small pulley and rubber belt transfer the rotation to one of the road wheels. The magnets control and transmit the force; the battery remains the source of the vehicle's energy. There is a great animation further below - feel free to scroll down if you want to see it now.
The magnetic switch is a small glass reed switch, visible on the white control arm above the flywheel. Inside the capsule, two flexible ferromagnetic contacts are normally separated. A passing magnet draws them together and closes the circuit; they spring apart again as the magnet moves away. This creates a position-sensitive switch without a mechanical lever or an electronic controller. It is sometimes loosely called a reed relay, although technically a reed relay includes its own operating coil; the Magnet Mobil uses a standalone reed switch operated directly by the flywheel magnets.
That simplicity is why I wanted to recreate the model. Its operation can be watched, adjusted, and understood one pulse at a time. In the process it brings together magnetic fields, electric current, inductance, switching, stored energy, inertia, friction, mechanical transmission, measurement, CAD, 3D printing, PCB fabrication, and the patient process of testing a design that almost works.
What was YPS?
For an Australian reader, some context is useful: YPS was never a familiar fixture at the local newsagent. In German-speaking Europe, however, it became part of the childhood landscape for a generation.
YPS was a German comic magazine built around a wonderfully direct promise: every issue came with a Gimmick. The magazine and its extra were sealed together in a clear plastic bag, so the contraption on the cover was also the thing readers took home to build, grow, test, or play with.
These Gimmicks were more than novelty giveaways. Many were inexpensive, self-assembly STEM kits: simple microscopes and radios, model aircraft, solar balloons, optical instruments, mechanical devices, and other experiments that put a scientific principle into a child's hands. They were not precision laboratory equipment, but that was part of their value. The low-cost parts made observation, construction, and experimentation accessible without requiring a specialist kit or workshop.
Other issues included toys, practical-joke devices, survival gear, craft projects, and tiny brine shrimp marketed as Urzeitkrebse—“prehistoric crabs.” Comics and illustrated instructions gave each object a story, explained how to assemble or use it, and encouraged the reader to try something rather than merely read about it.
The Magnet Mobil appeared in 1980 as YPS Extra No. 5, early in the more ambitious YPS Extra line. The specialist YPS Fanpage catalogue identifies the issue and places the first fifteen Extras between early 1980 and September 1982. Half toy, half experiment, it was designed to make a technical idea memorable by putting it directly into the reader's hands.
The motor on display
The original mechanism placed its six-magnet flywheel at the centre of the design, with the coil, switch, battery, and belt arranged around it.
That visibility is what makes the design worth revisiting. A conventional DC motor would be smaller, faster, and more reliable—but also far less revealing. Here, the wheel is not merely connected to the motor. It is part of the motor.
2. Recreating the Magnet Mobil
My goal was to make a functional and optically faithful, but not identical replica. I wanted to preserve the original idea while using the tools now available to make the model reproducible, adjustable, and open to further experimentation.
The original kit was made from moulded parts, and the tight cost constraints of a mass-market magazine extra. The Mk II can take a different path. Affordable 3D printers now make complex custom parts practical, while small PCBs are easy to design and manufacture.
The current Magnet Mobil Mk II therefore keeps the operating principle but rebuilds the structure around open, parametric CAD:
Element | Magnet Mobil Mk II design |
Flywheel | 80 mm diameter, normally fitted with six magnets, expandable to 12 magnets |
Chassis | Support-free printed base |
Timing | Pivoting control arm with a reed-switch cradle |
Electromagnet | 18 mm printed bobbin with a 14 mm winding area; successful coil approximately 460 turns and 4 Ω |
Transmission | Two printed pulley cones, a rubber-band belt, and 25 mm road wheels |
Electronics | Fabricated PCB from supplied Gerbers |
The model is written in OpenSCAD, with printable STL exports included for builders who do not want to edit the geometry. The main dimensions are grouped near the top of scad/assembly.scad, so builders can modify the SCAD files and produce a larger or smaller Magnet Mobil. Changing the flywheel_diameter parameter resizes the rotor and automatically adjusts the bearing height, electromagnet position, chassis length, and control-arm length around it. This is parametric adaptation rather than uniform scaling: dimensions tied to real hardware, including the magnets, shafts, and electromagnet bobbin, remain fixed unless their own parameters are changed. Within those practical limits, a new size can therefore be explored without repositioning every surrounding part by hand.
The construction details now live in the dedicated build guide, with quantities and component specifications in the parts list. Keeping those instructions in one place makes them easier to revise as the prototype is measured and refined. What follows here is the reasoning behind the design and the changes that made the Mk II work.
Six magnets, configurable to twelve
The standard Mk II uses six magnets. This produces a balanced flywheel and recreates the pulse spacing of the original Magnet Mobil: one possible electromagnetic impulse every 60 degrees. By setting flywheel_extra_magnet_cradles = true; creates as 12-magnet version of the flywheel. Filling them doubles the number of possible impulses to one every 30 degrees.
More frequent impulses can make the torque delivery smoother and give the motor more opportunities to start from rest, but twelve magnets do not automatically produce twice the useful power. They also double the number of reed-switch and coil events at a given flywheel speed and can increase average battery current and magnetic cogging. The better configuration depends on the coil, air gap, timing, mechanical drag, and load.
The completed 4 Ω driving prototype in the images shows the six-magnet arrangement. Twelve magnets provide a useful comparison of pulse frequency, current demand, cogging, and load behaviour.
Recreating the electromagnet
The original YPS Extra did not ask the reader to design or wind an electromagnet. The coil and its ferromagnetic core were supplied together as an assembled part, ready to install in the white plastic holder. That removed a large source of variation: the wire gauge, turn count, winding resistance, core material, and core geometry had already been chosen by the manufacturer.
For the Mk II prototype, that component no longer exists as an off-the-shelf part. We have to reproduce its function from a printed bobbin, enamelled copper wire, and a suitable ferromagnetic core. This makes the electromagnet one of the most consequential parts of the rebuild. Two models printed from identical STLs can behave very differently if their home-wound coils or cores differ.
The practical target was therefore not a visually exact copy of the original coil, but a replacement whose wire diameter, winding resistance, estimated turns, and supply voltage could be documented.
Core material is a first-order design choice. Soft iron or suitable low-carbon steel is a better starting point than an unidentified bolt or hardened steel: the aim is high permeability with low coercivity, so the core strengthens the field while the coil is on but retains as little magnetism as practical after the pulse. Some stainless steels are only weakly magnetic, while alloy and heat treatment can make apparently similar steel rods behave differently. This is one reason two models printed from identical files can perform very differently.
The successful driving prototype uses a hand-wound coil made from approximately 0.3 mm enamelled copper wire. It is estimated at roughly 460 turns, measures approximately 4 Ω, and is powered by two AA cells in series at approximately 3 V. With this electromagnet, the vehicle drives on its 25 mm road wheels using a rubber band as the belt.
The circuit
The Magnet Mobil Mk II uses a deliberately simple circuit. The two-cell battery pack, power switch, electromagnet coil, and normally open reed switch form a single series path. The flyback diode is the only component connected in parallel: it sits across the coil, with its striped cathode on the positive-supply side.
The N.O. label means normally open. With no magnet nearby, the reed contacts are separated and no current flows through the coil; “normally” refers to this resting state without a magnetic field, not to the position of the power switch. As a flywheel magnet closes the reed switch, current flows through L1 and creates the electromagnetic pulse. When the reed opens again, the collapsing magnetic field tries to keep the coil current flowing. D1 then conducts briefly around the coil, limiting the reverse-voltage spike that would otherwise arc across the reed contacts. It does not limit the normal coil current.
The simplicity is part of the appeal, but it also means the reed-switch contact rating, the flyback diode, the coil resistance, and the supply configuration all matter directly. There is no transistor or MOSFET isolating the reed switch from the pulsed coil current.
3. Operation, Physics & Results
One Push at a Time
The Magnet Mobil is a pulse motor. Instead of driving the flywheel continuously, it gives the wheel a short magnetic pull (or push) at the right moment and then lets inertia carry it forward.
The cycle is simple:
A permanent magnet on the flywheel approaches the trigger position.
Its magnetic field closes a reed switch mounted on the adjustable arm.
The switching circuit briefly energises the electromagnet.
The resulting magnetic force accelerates the flywheel.
The magnet moves past the trigger, the switch opens, and the coil turns off.
The flywheel coasts until the next magnet begins the cycle again.
Click the Play button in the animation below. The highlighted magnet at the reed switch is the trigger; because the six magnets are evenly spaced, another magnet is aligned with the electromagnet and receives the pull.
Once the wheel is running, a small pulley on its shaft drives a rubber belt. The belt turns one road wheel and converts the flywheel's rotation into the slow movement of the vehicle.
The sequence is easy to describe. Making it work well is the interesting part. The pulse must arrive at the right angle, the coil must be strong enough, the flywheel must spin freely, and the belt must transfer power without adding too much drag. A few millimetres of switch movement can separate a motor that runs from one that merely twitches.
Why reed-switch position matters
The switch position controls two separate things. Its sideways offset determines whether the passing magnets close the contacts reliably at all, while the angle of the complete white control arm determines when the electromagnet is energised. A poorly placed capsule may respond to only some magnets; a correct capsule position with poor timing may produce nothing more than a twitch.
That visible relationship between position and timing is central to the model. Rotating the arm changes the point at which electrical energy becomes a magnetic impulse, making an otherwise abstract timing problem directly observable. The successful prototype uses an offset capsule and a control-arm angle chosen through testing. The measurement and adjustment procedure, including the initial angle, is documented in the build guide.
Why the Core and Air Gap Matter
A coil does not create useful force from current alone. Its turns and current provide the magnetomotive force that drives magnetic flux, but the route available to that flux determines how much field reaches the flywheel magnet. It is helpful to picture this route as a magnetic circuit. Its opposition to flux is called reluctance, the magnetic counterpart of electrical resistance.
Air carries magnetic flux poorly compared with a soft iron or low-carbon steel core. Placing a suitable core inside the coil therefore strengthens and concentrates the magnetic field at the end facing the flywheel. The material matters: two metal rods of the same size can produce noticeably different pulling forces.
The air gap between that core face and the passing magnets is usually the largest source of reluctance in the circuit—even though it is physically the shortest part. Reducing the gap can therefore increase the flux density and the attractive force substantially. In a simplified, unsaturated magnetic circuit, the field rises roughly as the gap becomes shorter, while magnetic pressure depends on the square of the field. The real Magnet Mobil is less tidy: flux spreads at the core face, the permanent magnet is part of the circuit, and the core eventually approaches magnetic saturation. The design goal is therefore the smallest mechanically safe gap, not zero clearance. Flywheel wobble, shaft play, print distortion, or a slightly proud magnet can otherwise turn a strong magnetic design into a source of friction or contact.
There are three useful limits to remember. First, once the core begins to saturate, extra coil current produces much less additional flux and mostly more heat. Second, a better core and smaller gap raise the coil's inductance. That can slow the rise of current during a very short pulse. Third, the permanent flywheel magnets attract the ferromagnetic core even when the coil is off. This creates a detent, or cogging, torque as each magnet approaches and leaves the core. A narrower gap strengthens that unpowered attraction as well as the useful powered pulse. Core material, gap, turn count, resistance, supply voltage, and switch timing therefore form one system rather than independent upgrades.
That interdependence makes the Magnet Mobil a fantastic opportunity for hands-on experimentation. Change one variable at a time—core material, air gap, turn count, supply voltage, or switch position—and observe how the motor's starting behaviour, speed, cogging, current, and temperature respond. The effects are immediate and visible, turning abstract ideas about magnetic circuits, inductance, and mechanical load into measurements that can be made at the workbench.
Driving the Magnet Mobil
Getting the flywheel to spin is only the first milestone. Driving the Magnet Mobil across a surface requires enough torque to overcome the belt load, bearing friction, wheel friction, and the inertia of the complete vehicle.
With approximately 460 turns, 4 Ω measured resistance, and two AA cells supplying approximately 3 V, the Magnet Mobil drives successfully. The demonstrated chassis retains the existing 25 mm road wheels and uses an ordinary rubber band between the flywheel pulley and the grooved driven wheel.
An open-ended STEM platform
The interactions that make the model demanding are also what make it interesting. Moving the angle of the control arm changes the timing. Changing the coil changes the magnetic field. Narrowing the air gap strengthens both the useful pulse and the unpowered cogging force. Tightening the belt may reduce slip while adding friction. Every adjustment has an effect that can be seen, heard, or measured.
For a classroom, those effects can become experiments: map speed against switch angle, compare magnet counts, measure coil current, test different core materials, or record how long the flywheel coasts with the belt connected and disconnected. The model makes the usual diagrams of fields, current, inertia, and friction tangible.
The most useful way to approach Magnet Mobil Mk II remains in the spirit of the original: not as a product to consume, but as a small machine that rewards curiosity.
I invite teachers, students, and hobbyists to recreate the Magnet Mobil and adapt it to the tools and components they have available. The CAD files, printable parts, circuit information, and build documentation are provided in an open-source spirit so that the design can be studied, changed, and improved. Not having every specified part need not stop a build; it can be an invitation to creative tinkering. The circuit, for example, is simple enough to wire by hand without a PCB, while the parametric OpenSCAD model can be adjusted for different dimensions or locally available hardware. Build one, measure what it does, and share what you learn.
Classroom experiences, smarter tools for teachers, and professional learning on demand
September is a great time to explore some of the latest resources and ideas from DTI. Our three recent releases span Mathematics, Cybersecurity, AI and Robotics, offering new ways for students to explore, experiment and learn by doing.
That same philosophy extends beyond the screen. With B4 Spark, students can get hands-on with the inner workings of a computer, while our Curriculum Search Engine and Digital Transformation Tool help teachers and schools turn ideas and insights into action. Together with our on-demand professional learning resources, there’s plenty to explore in this edition.
🔔 In This Issue
📈 CurveLab – Explore functions, derivatives and curve sketching
⚔️ Cyber Command – Red Team vs Blue Team cybersecurity
🤖 Capture the Flag with AI and Robotics – Meet Scout the Robot
📊 Digital Transformation Tool – From insight to action
🔍 Curriculum Search Engine – Start with an idea, find where it fits
🎓 Professional Learning – Learn when it suits you
🧰 Computing You Can Touch
📈 CurveLab
CurveLab is our new free mathematics environment for exploring functions, graphs and derivatives.
Students can investigate intercepts, turning points, minima, maxima, concavity and points of inflection, while comparing the graph of f(x) with its derivatives. A built-in derivative checker also lets students test their own calculations and receive immediate feedback.
It works particularly well as a companion to classroom teaching: predict and calculate first, then use CurveLab to investigate and check.
⚔️ Cyber Command
What happens after students have learned the basics of Python?
Cyber Command provides one answer.
Two students enter the Battle Arena as Red Team and Blue Team, applying Python skills to password attacks, phishing and cyber defence before moving into a strategic incident-response battle.
Rather than completing another sequence of individual coding exercises, students must code, compete, react and make decisions under pressure.
🤖 Meet Scout: AI and Robotics
Our completely refurbished Capture the Flag with AI and Robotics course gives younger students a very tangible introduction to artificial intelligence.
Students meet Scout, a virtual robot, observe how its colour sensor provides input to its AI, and see how the resulting decisions control its motors. When Scout’s driving AI breaks down, students gradually rebuild it themselves.
The course combines robotics, AI, systems thinking and problem solving in an engaging mission story—and is now available in English, German, French, Spanish and Chinese.
📊 How Digitally Capable Is Your School?
The Digital Transformation Tool (DTT) takes a whole of school approach: rather than focusing on individual lessons, it helps schools understand and strengthen their overall digital capabilities.
Schools can assess areas including leadership, learning, managing and teaching, explore their results, identify priorities and quick wins, and use the integrated AI Advisor to turn those insights into concrete actions.
It’s deliberately more than a static survey and report. The aim is to help answer the question that comes afterwards: “What should we do next?”
🔍 Start With Your Teaching Idea
Teachers don’t usually start planning with a curriculum code. They start with an idea. Perhaps you want students to analyse patterns in weather data, investigate sustainability, build a robot—or explore cybersecurity.
Enter that idea into our Curriculum Search Engine, and it searches the Australian Curriculum V9 for relevant connections across learning areas, capabilities and priorities. Its meaning-aware search means you can describe what you want to teach naturally rather than trying to guess the terminology used by the curriculum.
The integrated AI Advisor can then help turn those curriculum connections into lesson ideas.
🧐 Teacher Professional Development
Do AI-related PD at times that suit you best:
The DTI Classroom offers an interactive, self-paced AI workshop. It includes everything teachers need to get started with AI teaching. 👉 Check it out
You can subscribe to our AI video channel to watch AI webinars at your convenience. 👉 Check it out and scroll to the bottom of the page.
🧰 Computing You Can Touch
With so much of Digital Technologies now happening on a screen, B4 Spark offers something refreshingly different: students can explore how a computer actually works by building and programming one themselves.
Through hands-on activities, students investigate binary, memory, instructions, algorithms and the fundamental operation of a computer processor. Rather than treating the computer as a black box, B4 Spark makes its inner workings visible and tangible.
It’s a great way to bring physical, hands-on computing back into the classroom—and a useful complement to coding, AI and cybersecurity activities on MyComputerBrain.
Explore at 👉 Spark Home.
We remain committed to supporting you with practical tools, inspiring PD, and curriculum-aligned resources. Whether you’re exploring AI, upskilling in Digital Systems, or searching for the perfect content, we’re here for you.
Warm Regards,
Dr. Karsten Schulz
