Feedback traps are tools for trapping single charged objects in solution. They periodically measure an object’s position and apply a feedback force to counteract Brownian motion. The feedback force can be calculated as a gradient of a potential function, effectively creating a “virtual potential.” Its flexibility regarding the choice of form of the potential gives an opportunity to explore various fundamental questions in stochastic thermodynamics. Here, we review the theory behind feedback traps and apply it to measuring the average work required to erase a fraction of a bit of information. The results agree with predictions based on the nonequilibrium system entropy. With this example, we also show how a feedback trap can easily implement the complex erasure protocols required to reach ultimate thermodynamic limits.
Feedback traps can create arbitrary virtual potentials for exploring the dynamics of small Brownian particles. In a feedback trap, the particle position is measured periodically and, after each measurement, one applies the force that would be produced by the gradient of the “virtual potential,” at the particle location. Virtual potentials differ from real ones in that the feedback loop introduces dynamical effects not present in ordinary potentials. These dynamical effects are caused by small time scales associated with the feedback, including the delay between the measurement of a particle’s position and the feedback response, the feedback response that is applied for a finite update time, and the finite camera exposure from integrating motion. Here, we characterize the relevant experimental parameters and compare to theory the observed power spectra and variance for a particle in a virtual harmonic potential. We show that deviations from the dynamics expected of a continuous potential are measured by the ratio of these small time scales to the relaxation time scale of the virtual potential.