How does electrical termination control damping in a bilateral electromechanical transducer?

One coupling function drives torque and back EMF

The paper's central claim is that a bilateral electromechanical transducer is not two separate devices glued together. Starting from a magnetic pendulum, it defines the flux linkage as a function of rotation angle and takes its derivative with respect to angle to obtain a single electromechanical coupling coefficient. The torque produced by current and the back electromotive force produced by motion both follow from that same coefficient, so energy conversion in the two directions is described by one function rather than by two independent laws.

The paper traces this to the magnetic coenergy, whose mixed partial derivatives with respect to current and angle are equal. That equality is what forces the torque-per-current and the flux-change-per-angle to be the same quantity. The opposing sign of the back EMF in the electrical balance is attributed to the chosen port orientation, consistent with Lenz's law.

Radial-field geometry makes the coupling constant

In the uniform-field magnetic pendulum, the coupling coefficient is proportional to the sine of the rotation angle, so it vanishes at the stable equilibrium. The paper contrasts this with the d'Arsonval movement, where curved pole pieces and a concentric soft-iron core keep the field radial across the coil's operating range. Because the radial field stays normal to the surface swept by each active side throughout the rotation, the angle-dependent factor disappears and the coupling coefficient becomes a constant, written as two times the number of turns times the radius times the axial length times the field magnitude.

The paper is explicit that this is not an approximation. The sine factor is absent because the geometry removes it, not because a small-angle expansion was applied. The cost is that a constant coupling gives an electromagnetic torque independent of angle, so it supplies no restoring action; a separate suspension or hairspring must provide the restoring torque, and the static balance then sets the steady deflection as the product of coupling and current divided by the spring constant.

Damping enters through the electrical port

Once the current is no longer imposed independently of motion, the changing flux linkage adds a back EMF to the electrical loop. Neglecting the coil's electrical transient for the moment, the current becomes slaved to the instantaneous mechanical state, and substituting that current into the torque relation splits the torque into a static-current part and a motion-induced part. The motion-induced part is a damping torque whose coefficient equals the square of the coupling divided by the total resistance seen by the coil.

The paper shows that this contribution can only remove mechanical energy from the coil, whatever coupling a design produces. For the uniform-field coil, however, the damping coefficient vanishes quadratically as the angle approaches zero, so electromagnetic damping does not even appear in the linearized dynamics about the stable equilibrium. For the radial-field coil, the constant coupling makes the damping coefficient constant as well, including at zero angle. The paper describes this as the puzzle resolved by design rather than by approximation.

Critical damping becomes a resistance choice

Real galvanometer movements also dissipate energy through mechanisms intrinsic to their construction, which the paper represents at the level of the mechanical equation as an equivalent intrinsic damping coefficient. Adding that to the electromagnetic term gives an ordinary damped harmonic oscillator. Critical damping then requires the sum of intrinsic and electromagnetic damping to equal twice the square root of the product of moment of inertia and spring constant.

Solving that condition for the external resistance gives the external critical damping resistance. The paper notes two conditions attached to this expression: it assumes the intrinsic damping is below twice the square root of the inertia-spring product, and it yields a physically realizable positive external resistance only when the coupling term exceeds the coil resistance. Otherwise critical damping cannot be obtained by adding a passive series resistance. When intrinsic damping is zero, the entire damping requirement is supplied through the electrical port.

Same coil, same deflection, different transients

The numerical example compares two electrical driving conditions on one galvanometer, both producing the same 30-degree steady deflection. Under an ideal current source, the electromagnetic damping vanishes and the response reduces to a second-order mechanical system with a damping ratio of about 0.15, which the paper calls strongly underdamped despite the intrinsic damping. The current amplitude needed is 50.0 microamperes.

Under a voltage source whose resistance is set equal to the external critical damping resistance, the coil current is determined dynamically by the applied voltage and by the motion-induced EMF, giving a third-order system in angle, angular velocity, and current. No electromagnetic damping term is added explicitly; its effect arises from the coupling and the termination. The voltage step needed for the same steady deflection is 8.46 millivolts, and the quasi-static theory predicts a damping ratio of one.

Time-scale separation justifies the reduced model

The paper's quantitative criterion for dropping the electrical transient is a separation between the electrical and mechanical time scales. In the example, the electrical relaxation time is 4.30 microseconds while the mechanical time scale is 0.229 seconds, a separation by a factor of about 53,300. Consistently with that criterion, the complete third-order response differs from the quasi-static critical response by at most 4.01 times ten to the minus four percent of the final deflection.

This is the paper's check that the reduced second-order description is not merely convenient but quantitatively adequate for the parameters used. The comparison is made against the full third-order dynamics rather than against a separate experiment, so the agreement speaks to the internal consistency of the model and its quasi-static reduction, not to a measured device.

The termination is part of the transducer

The paper's two-port representation identifies electromagnetic damping as an electrical impedance reflected into the mechanical domain. That framing is what makes the result portable beyond the galvanometer: the same moving-coil system, driven to the same equilibrium position, can show markedly different transient dynamics solely because its electrical boundary conditions changed. Nothing about the coil, the magnet, or the spring changes between the ideal-current and critical-termination limits; only the impedance of what drives the coil changes.

For measurement systems, the paper draws the consequence that a transducer and the circuit connected to it cannot always be treated as dynamically independent. Loading may affect not only the magnitude of a measured signal but also the dynamics through which that signal is produced. The paper presents the moving-coil system as an undergraduate setting in which familiar electromagnetism and oscillator dynamics lead to this broader picture of coupled systems.

What decides the damping is the total resistance

The practical priority this paper establishes is that the electromagnetic damping coefficient is set by the total resistance seen by the coil, source included, not by the coil alone. The two limits are the decision-relevant comparison: as the source resistance decreases toward a low-output-resistance voltage source, the total resistance falls and the electromagnetic damping rises; as the source resistance grows toward an ideal current source, the back EMF can no longer perturb the prescribed current and the electromagnetic damping disappears.

The threshold that matters for design is the external critical damping resistance, which combines the coupling, the inertia-spring product, the intrinsic damping, and the coil resistance. The paper's own caveat is that this threshold is reachable only when the coupling term exceeds the coil resistance and the intrinsic damping stays below twice the square root of the inertia-spring product. Within those conditions, the same coil can be made strongly underdamped or critically damped purely by choosing what drives it.

Sources used in this answer