As embedded fixed-point signal processing becomes more common [1], it becomes increasingly important that engineers have the proper training and tools to tackle the difficult problems intrinsic to implementing algorithms in fixed-point. However, educational institutions and tool vendors have been slow to respond to this need. Engineers thus endure a long slow learning curve before they can be productive. In this paper, we draw on experience to suggest an approach to teaching fixed-point signal processing. We also introduce new tools that make this difficult subject more accessible to educators, students, and practicing engineers.