Rotary encoders, electronic components that convert rotational displacement into electrical signals, find applications in various settings, such as adjusting air conditioner temperature and orienting objects in virtual reality environments. Recently, tactile and auditory sensations during operation have become pivotal in assessing the value of switches and other electronic components as affective products. This paper proposes a method to construct an affective editor that automatically adjusts the physical parameters of a rotary encoder in response to a user’s request to modify the tactile sensation of the encoder. Two experiments were conducted to develop this method. In the first experiment, 17 participants evaluated the tactile sensation of rotary encoders using 30 adjective pairs across 29 different parameter settings, yielding four factors through factor analysis. In the second experiment, the stimuli presented increased to 188. Despite having only one participant, the obtained factors were essentially the same as those from the first experiment. Based on this data, an affective prediction model was constructed for each factor using an artificial neural network with inputs representing the physical parameter values. The parameter scan method revealed the relationship between the physical parameters and predicted values. Utilizing these relationships, an affective editor that generates physical parameter values based on affective demands was constructed. An evaluation experiment of the editor, conducted with six participants, demonstrated that, except for one factor, the impressions of the factors were successfully controlled as intended. This suggests the potential to control tactile sensations.