A magnetorquer is a spacecraft actuator that generates a controlled magnetic field to interact with the magnetic field of the planet it orbits, producing a torque that rotates the satellite. With no moving parts, no propellant, and power requirements measured in single-digit watts, magnetorquers have become standard equipment on small satellites and CubeSats. They are deceptively simple devices, but the engineering choices behind them, their operational limitations, and the creative ways mission designers work around those limitations make them far more interesting than the basic concept suggests.
How a Magnetorquer Produces Torque
The underlying physics is something you can demonstrate with a compass needle. A loop of wire carrying an electric current creates a magnetic dipole, just like a small bar magnet. Place that dipole inside an external magnetic field and it experiences a torque trying to align it with the field. On a spacecraft orbiting Earth, the external field is the planet’s own geomagnetic field, which is strong enough in low Earth orbit to be useful for attitude control but weak enough that you need a carefully designed coil to produce a meaningful response.
The strength of the torque depends on two things: the magnetic dipole moment the coil generates and the strength and direction of the external field at the satellite’s location. The dipole moment, in turn, depends on the number of turns in the coil, the area each turn encloses, and the current flowing through it. Mission designers balance these parameters against mass, volume, and power budgets. A satellite typically carries three magnetorquers mounted along perpendicular axes so it can produce torque in any direction the local field geometry allows.
Three Physical Forms
Magnetorquers come in three main flavors, each suited to different spacecraft sizes and mission profiles:
- Torque rods: A coil wound around a core of high-permeability material like permalloy. The core amplifies the magnetic moment by a factor of several hundred compared to an air-core coil of the same size, making torque rods the strongest option per unit volume. They are common on larger small satellites and some CubeSats where the mass and volume penalty is acceptable.
- Air-core coils: Wire wound into a frame without a ferromagnetic core. Lighter than torque rods and free of the nonlinear hysteresis effects that cores introduce, but they need more turns or more current to reach the same dipole moment. They are sometimes wound around the outside of a CubeSat’s structure to maximize the enclosed area.
- PCB-embedded coils: Copper traces etched directly onto a printed circuit board in a spiral pattern, sometimes spanning multiple layers of a multilayer PCB. These produce the weakest dipole moment per unit area, but they add almost no mass, require no separate mounting hardware, and can be integrated into boards that already serve other functions.
The choice among these depends heavily on what the satellite needs to accomplish. A CubeSat that only needs to detumble after deployment and maintain a rough orientation might get by with PCB coils alone, while a larger satellite performing precise Earth observation will likely need torque rods or a hybrid approach combining magnetorquers with other actuators.
PCB-Integrated Designs and Miniaturization
The push toward ever-smaller satellites has driven serious engineering effort into making magnetorquers that are essentially free in terms of mass and volume. PCB-integrated magnetorquers embed the coil traces directly into the circuit board’s copper layers, turning the satellite’s own electronics boards into attitude-control hardware. Research comparing theoretical coil models against electromagnetic simulations of these embedded designs has shown strong agreement, with the theory slightly overestimating coil resistance by about five percent but accurately predicting the relationship between resistance and the magnetic moment produced.1arXiv. Efficiency-optimized design of PCB-integrated magnetorquers for CubeSats That kind of predictability matters because designers need to know in advance exactly how much torque they can expect from a given board layout.
For larger CubeSat form factors like 2U and 4U, researchers have demonstrated reconfigurable planar coils that can switch between different winding configurations depending on whether the satellite needs maximum torque for fast detumbling or lower power draw for steady-state pointing. Thermal analysis is a key part of these designs because the coils are buried inside a multilayer PCB stack; dumping too much current through them can heat the board and affect nearby electronics. Studies comparing these embedded designs against commercial off-the-shelf magnetorquers have found that purpose-built planar coils can outperform the commercial alternatives.2Computers and Electrical Engineering. PCB-integrated embedded planar magnetorquers for small satellites intelligent detumbling
At the extreme end of miniaturization, magnetorquers have been designed for PocketQube satellites, which are roughly a quarter the size of a 1U CubeSat. Even at that scale, PCB coils can function on a regulated 3.3-volt supply. Testing of two different PocketQube coil topologies showed that one design drew roughly twice the current of the other at the same bus voltage, which translated into about double the instantaneous power consumption and a much shorter capacitor discharge time of around four seconds versus ten.3ResearchGate. Printed Circuit Magnetorquer for a PocketQube Satellite That kind of trade-off between torque authority and power budget becomes acute on satellites that weigh a few hundred grams and generate only a few watts from their solar cells.
What Magnetorquers Actually Do on a Mission
The most common job for a magnetorquer is detumbling. When a satellite is released from its launch vehicle, it is almost always spinning or tumbling in an uncontrolled way. The first priority is to stop that rotation so the satellite can deploy its solar panels toward the Sun and point its antennas at ground stations. Magnetorquers are ideal for this because the detumbling algorithm is simple, the hardware is reliable, and the process does not consume any propellant. A basic detumbling controller measures the change in the local magnetic field and drives current through the coils in a direction that opposes the spin. Within a few orbits, the tumbling rate drops to something manageable.
The second major role is momentum dumping, also called desaturation. Many satellites use reaction wheels as their primary pointing actuators because wheels can produce torque about any axis at any time, unlike magnetorquers, which are constrained by the local magnetic field direction. But reaction wheels accumulate angular momentum over time from external disturbance torques like atmospheric drag and solar radiation pressure. Eventually a wheel reaches its maximum spin rate and can no longer absorb more momentum. Magnetorquers bleed off that excess momentum by working against the geomagnetic field, bringing the wheel speeds back down to their operating range. This desaturation function is so important that even satellites with sophisticated reaction-wheel systems almost always carry magnetorquers as well.
A third and increasingly common application is deorbiting. As space debris becomes a growing concern, regulations now encourage or require satellite operators to deorbit their spacecraft within a set number of years after mission end. One approach uses a deployable drag sail to increase atmospheric drag and pull the satellite down faster. Keeping that sail oriented correctly requires active attitude control, and a scheme using magnetorquers together with a bias momentum wheel and an onboard model of the geomagnetic field can keep the sail perpendicular to the satellite’s velocity using nothing but magnetometers as sensors.4Journal of Guidance, Control, and Dynamics. Magnetic Attitude Control Scheme for Drag Sail Deorbiting with Limited Sensors This is appealing because by the end of a mission, some sensors and actuators may have degraded, and a minimal control system that relies on just magnetometers and magnetorquers has a good chance of still working.
The Geomagnetic Field Problem
Magnetorquers have a fundamental limitation baked into their physics: they can only produce torque perpendicular to the local magnetic field vector. If you want to rotate your satellite about an axis that happens to be aligned with the field at that moment, you are out of luck until the satellite moves along its orbit and the field direction changes. In low Earth orbit, where the field changes constantly as the satellite circles the globe, this is manageable over time. The satellite simply waits for a more favorable field orientation, or the control algorithm exploits the field variation over a fraction of an orbit to build up the needed rotation.
At higher altitudes, the geomagnetic field weakens rapidly, roughly as the cube of the distance from Earth’s center. By the time you reach geostationary orbit at about 36,000 kilometers altitude, the field is too weak for magnetorquers to be practical as primary actuators. They remain useful in low Earth orbit, roughly up to about 2,000 kilometers, where the field is strong enough to generate meaningful torque with reasonable coil sizes and currents.
The dependency on a planetary magnetic field also limits where magnetorquers can be used at all. Mars, for instance, does not generate a global magnetic field of its own, though interactions between the solar wind and the Martian atmosphere create a weak induced magnetosphere. A study examining satellite swarm concepts around Mars noted that while this induced field could in principle provide some torque for desaturation of reaction wheels, the variability and weakness of the field would make magnetorquers unreliable there.5Physica Scripta. Review and viability of a Dyson Swarm as a form of Dyson Sphere For missions beyond Earth orbit, thrusters or other non-magnetic actuators are typically the only viable options for managing angular momentum.
Magnetic Cleanliness and Sensor Interference
One of the trickiest practical challenges with magnetorquers is that they create strong local magnetic fields that can blind or bias the satellite’s own magnetometers. This matters because those same magnetometers are often the sensors the attitude-control system depends on to measure the local geomagnetic field and figure out which way the satellite is pointing. If the magnetorquer’s field leaks into the magnetometer reading, the satellite’s knowledge of its own orientation degrades.
The standard mitigation is time-sharing: the control computer switches off the magnetorquers briefly, takes a magnetometer reading during that quiet interval, then switches the torquers back on. This works but introduces gaps in the control authority. More sophisticated approaches try to model and subtract the magnetorquer’s contribution to the magnetometer reading, which requires good calibration of both instruments. Data from the GRACE satellite mission, for example, showed that accounting for magnetic disturbances produced by the spacecraft’s magnetorquers was a necessary step in processing the onboard magnetometer measurements.6Earth, Planets and Space. Magnetometer data from the GRACE satellite duo
The scale of the interference can be substantial. Calibration work on a CubeSat magnetometer revealed that the maximum expected magnetic field at the sensor location, set largely by its proximity to the spacecraft’s magnetorquers, reached about 150 microtesla.7Geoscientific Instrumentation, Methods and Data Systems. Verification and calibration of a commercial anisotropic magnetoresistive magnetometer by multivariate non-linear regression For context, Earth’s magnetic field at the surface is roughly 25 to 65 microtesla depending on latitude. So the magnetorquer can create a local field several times stronger than the geomagnetic field the satellite is trying to measure. Testing also found that hysteresis effects, where the magnetometer retains a slight offset after exposure to a strong field, contributed measurement errors on the order of tens of nanotesla. The researchers concluded that maintaining magnetic cleanliness near the magnetometer, keeping ferromagnetic components and current-carrying traces well away from the sensor, was essential for meeting their 100-nanotesla accuracy requirement.8Geoscientific Instrumentation, Methods and Data Systems. Verification and calibration of a commercial anisotropic magnetoresistive magnetometer by multivariate non-linear regression
Passive Magnetic Systems
Not every mission needs active magnetorquers with coils and control electronics. Some small satellites use a purely passive magnetic attitude control system consisting of permanent magnets and hysteresis strips. A permanent magnet onboard aligns the satellite with the local geomagnetic field, like a compass needle, while hysteresis strips (thin pieces of magnetically soft material) dissipate rotational energy as the satellite oscillates around the field-aligned orientation. Over time, the oscillations damp out and the satellite settles into a stable alignment with its long axis pointing along the field lines.
The AstroBio CubeSat, a 3U science mission, used exactly this approach: permanent magnets and hysteresis strips stabilized the spacecraft with its longitudinal axis along the geomagnetic field vector, with no active electronics in the attitude-control loop at all.9Aerospace. Characterization and Testing of the Passive Magnetic Attitude Control System for the 3U AstroBio CubeSat The tradeoff is obvious: passive systems are extremely simple and cannot fail in the way active electronics can, but they provide no ability to point the satellite in any direction other than along the field. For missions where “roughly aligned with the geomagnetic field” is good enough, like certain technology demonstrations or atmospheric-science experiments, passive magnetic stabilization is an elegant solution. For missions requiring specific pointing at ground targets or other spacecraft, it is not an option.
Why Magnetorquers Persist Despite Their Limits
Reaction wheels can point a satellite at almost any target. Cold-gas thrusters can produce torque regardless of the local magnetic environment. Electric propulsion systems can handle both orbit and attitude adjustments. So why do magnetorquers remain nearly universal on small satellites? The answer comes down to a combination of factors that no competitor matches simultaneously. They consume no propellant, so they never run out. They have no moving parts, so they rarely fail mechanically. They draw modest power. They are light. And for the specific jobs they do well, detumbling, momentum dumping, and coarse attitude maintenance, they are hard to beat on a cost-per-function basis.
Even on larger, more capable spacecraft that rely primarily on reaction wheels or control-moment gyroscopes, magnetorquers typically remain onboard as the system that keeps the wheels from saturating. Losing that desaturation capability would mean either carrying extra propellant for thruster-based desaturation or accepting a shorter mission life. For missions in low Earth orbit, magnetorquers fill that role with zero consumables and minimal complexity. The ongoing miniaturization of PCB-embedded designs means the mass and volume penalty keeps shrinking, making it increasingly hard to justify leaving them off a spacecraft even if the primary attitude control comes from other hardware.
Thermal Considerations in Embedded Designs
When magnetorquer coils are etched into a printed circuit board rather than wound as separate hardware, thermal management becomes a distinct design concern. The copper traces that form the coil carry enough current to generate resistive heating, and because they are sandwiched between layers of fiberglass-epoxy laminate with relatively poor thermal conductivity, the heat does not dissipate as readily as it would from a standalone coil exposed to the spacecraft’s thermal environment. This is especially relevant during detumbling, when the magnetorquers may run at high duty cycles for extended periods to bring a tumbling satellite under control as quickly as possible.
The thermal evaluation of embedded planar magnetorquers for 2U and 4U CubeSats specifically examined whether the coils could sustain stable temperature profiles under realistic operating conditions.10Computers and Electrical Engineering. PCB-integrated embedded planar magnetorquers for small satellites intelligent detumbling The concern is not just that the coil itself might overheat, but that excessive temperatures could damage solder joints, degrade nearby integrated circuits, or alter the electrical characteristics of the traces themselves, changing the coil’s resistance and therefore its magnetic moment in ways the control software does not expect. Designers address this by limiting duty cycles, distributing coil traces across multiple PCB layers to spread the thermal load, and in some cases incorporating thermal vias that help channel heat toward external radiating surfaces.
At the PocketQube scale, where total spacecraft power budgets might be measured in milliwatts, the thermal issue is less about overheating and more about energy management. The factor-of-two difference in power consumption between two PocketQube coil designs translates directly into how long the satellite’s energy-storage capacitor can sustain a torquing pulse and how quickly that capacitor recharges from the solar cells. A four-second discharge window versus a ten-second one is not just a thermal or electrical question; it determines how much angular impulse the satellite can deliver per activation and therefore how quickly it can detumble or correct its orientation.