Functional description¶
A complete FluxGrip-series electropermanent magnet (EPM) consists of an at least one magnetizable magnetic assembly and a controller module. The controller module can magnetize (turn on) and demagnetize or relax (turn off) the magnetic assemblies connected to it. Both states are stable, allowing the magnetic assemblies to remain in either state indefinitely when powered down or even disconnected from the controller; power is only required for state transitions.
The FluxGrip product family adopts the following prefix nomenclature for its parts:
FGC — a FluxGrip controller unit (e.g., FGC1) containing everything but the magnetic assembly.
FGX — a FluxGrip magnetic assembly (e.g., FGX1S) driven by a controller unit.
FG — a fully integrated unit containing both the controller and the driven magnetic assembly.
FluxGrip solution architecture¶
To switch the connected magnetic assembly between the on and off states, the device consumes a small amount of energy from the power supply input. The device is designed to consume a fixed input current, which implies that the power consumption increases with the supply voltage, and therefore, the time it takes to transition between the states is inversely proportional to the supply voltage.
Tip
The magnetization and relaxation processes follow a non-linear time profile, with most of the transition effect occurring rapidly relative to the total transition time.
When the device is powered on but does not transition between the on and off states — or, in other words, resides in the idle state — some insignificant idle power is consumed by the built-in electronics and its communication interfaces. It is possible to power down the device completely when switching is not needed to conserve energy further, and turn it on ad-hoc whenever a state transition is needed.
The force that is required to pull the payload away from the magnetic assembly — the holding force — depends not only on the performance of the magnet itself, but also on the qualities of the payload. Ferromagnetic materials with high permeability (\(\mu_r > 1000\)) and high saturation polarization (\(J_\text{sat} > 2 \ T\)), such as iron, low-carbon steels, or specialized magnetic alloys like permendur, are recommended.
Another factor affecting the holding force is the proximity of the payload to the magnet pole shoes. For optimal performance, the payload should be in direct contact with the magnet pole shoes across its entire area. Imperfect contact, foreign objects, dirt, dust, or metallic shavings between the magnet and the payload may impair the performance of the magnet and may even cause damage if any hard impurities are forced into the polymer composite body of the magnetic assembly.
The standard equation for the force between two nearby magnetized surfaces applies:
with the magnetic flux density in the circuit \(B \ [\text{tesla}]\), contact area \(A \ [\text{meter}^2]\), and the permeability of free space \(\mu_0 = 1.25663706127(20)\times{}10^{−6} \ [\frac{\text{henry}}{\text{meter}}]\). The area is defined by the magnet pole geometry and thus fixed; \(\mu_0\) is likewise fixed assuming the customer resides in our universe; the only remaining variable is the magnetic flux density \(B\), which is defined by the unchanging properties of the magnet and the highly variable reluctance of the rest of the magnetic circuit: the payload and the airgap separating it from the magnet. It can be seen that halving the reluctance of the magnetic circuit by using a better payload material and/or reducing the airgap will quadruple the holding force.
A spurious partial magnetization or relaxation may occur if the magnetic assembly of the device is exposed to a strong external magnetic field. For example, placing a strong permanent magnet directly against the pole shoes of the device may cause it to partially acquire the magnetization of the external magnet. This effect will be undone at the next magnetization or relaxation cycle.
Attention
The magnetic assembly is only able to retain full magnetization as long as it is held in contact with the payload.
The magnetic assembly is only able to retain full magnetization as long as it is held in contact with the payload. As soon as the payload is removed, the intrinsic demagnetizing field of the magnetic assembly will cause partial relaxation. This means that once the payload is re-applied, the holding force will be lower than it was before its removal. To restore the full holding force, a new magnetization cycle is required; this can be achieved by sending a forced magnetization command to the device as explained below, or, alternatively, by simply cycling the magnet off and then back on again.
The device accepts commands via the communication interfaces in the form of the desired state: OFF, ON, and FORCE, whose handling is described in the following state transition table.
Command |
State |
Action |
|---|---|---|
OFF |
Off |
|
OFF |
On |
Demagnetize |
ON |
Off |
Magnetize |
ON |
On |
|
FORCE |
(any) |
Magnetize once |
Having received a FORCE command, the device will execute one magnetization cycle regardless of the current state and ignore further FORCE commands until any other command is received.
Fault states¶
Self-diagnostic capabilities enable the device to detect and report malfunctions. Upon detection of a problem, the device will enter a latched fault state and will not respond to magnetization commands until the fault is cleared, or the device is restarted.
Fault state reporting and clearing vary by communication interface; please refer to the relevant chapters for the specifics.
Failure modes detectable by the built-in diagnostics include malfunctions of the high-voltage power stage electronics, built-in sensor suite, and invalid configuration parameters. The latter fault mode should be addressed by correcting the configuration parameters (e.g., via the Cyphal interface) and then restarting the device.
A deep brownout of the power supply may cause the device to fail to magnetize or demagnetize the magnet, which may also trigger a fault state.
Registers¶
Registers are named values that store the configuration parameters and provide insights into specific states and events. They are accessible via the Cyphal interface.
Configuration parameters are not committed to the non-volatile memory until the device is commanded to restart. One implication of this is that the device will not retain changes to the configuration parameters if powered off without a prior explicit restart.
Some registers are not documented on purpose. These are not intended for production use and should not be relied upon; their name, type, contents, and semantics may change arbitrarily between minor releases with no regard for compatibility.