 Hello, I'm Patrick Dills from the Reach Lab at the University of Wisconsin-Madison, and I'm here to present on the stability properties and rendering limitations of admittance-controlled haptic devices. Haptic force-reflecting devices must be able to achieve a wide range of impedances to successfully render a virtual environment. Admittance-type haptic devices regulate a forced-to-velocity relationship to render a virtual environment to a user. This controlled topology typically renders large impedances easily, but has trouble rendering small impedances. Stability behavior is inherent to an admittance control architecture and limits the range of stable impedances that a device can achieve. This work examines the stability and rendering limitations of admittance-type haptic devices under a wider set of virtual admittances than previously examined, and identifies factors affecting the stability and rendering performance. Specifically, we examine virtual mass, stiffness and damping, and combinations of virtual mass, stiffness and damping. Our model used to analyze stability and rendering limitations includes a virtual admittance, a position controller, and a human impedance model. A single degree of freedom admittance control loop can be represented in block diagram form, and in the case of high performance admittance-type haptic devices, the control loop can be simplified by neglecting the human's disturbance on the high bandwidth position control loop. Considering high frequency behavior of the resulting open loop transfer function results in approximate expression showing the impact of the human's impedance, the position control bandwidth, and time delay on the minimum pure stiffness and damping and admittance-controlled haptic device can stably render. Results show that increases in position control bandwidth increase the minimum stable stiffness or damping a device can stably render. Furthermore, increases in time delay generally increase the minimum damping and stiffness and admittance control loop can stably render. However, if the device's phase crossover frequency is below the position control bandwidth, then increases in time delay can decrease the minimum stable stiffness. We also examined the effects of position control bandwidth and time delay on combinations of all three virtual admittances, and were able to create a map showing the volume of unstable impedances and admittance-type haptic device characteristically demonstrates. We experimentally validated these results with a user study where we asked six participants to interact with a single degree of freedom admittance-based haptic device. Participants interacted with bilateral virtual environments composed of a pure stiffness, combinations of mass and damping, and finally combinations of mass stiffness and damping, where we decreased the displayed impedance until unstable behavior occurred. Each user's impedance was estimated via frequency domain methods, and the data was used to generate theoretical stability boundaries, which compare well with the experimental results from our user study. In summary, we identified the effect of the human's impedance position control bandwidth and time delay on the minimum stable stiffness and damping and combinations of mass stiffness and damping and admittance device can achieve. This was validated with the user study, and we hope our work will inform the design of admittance-type haptic devices in the future.