One of the challenges for nanomechanical devices is to understand the different sources of noise and dissipation that act on such systems. In thermal equilibrium, these noise sources all have the same temperature and are thus indistinguishable from each other. It has been proposed, however, that the noise from electrons scattering off impurities and boundaries in the nanoresonator can be driven above the bath noise by applying a voltage across the resonator. The force acting on the nanoresonator due to these collisions is predicted to be detectable with current technology.

Here we describe experiments to measure this noise. Using a radio-frequency single electron transistor to measure the effective temperature of a nanomechanical mode, we have found that a) the mode temperature increases linearly with current through the nanoresonator and b) the mode temperature follows the expected temperature of the electron gas due to Joule heating.

We have not been able to identify the associated damping, however. Experiments on aluminum based nanoresonators have failed to yield the expected increase in dissipation at the crossover between the superconducting and normal states. We are left to conclude that the nanomechanical mode is coupled to the electron gas, but it is unclear whether this coupling is direct or the result of an intermediate dissipative system that itself is heated by the electron gas.