Character input in immersive environments is a non trivial task that has attracted much attention in recent years. This paper presents an evaluation of keyboard position, orientation and interaction together with the influence of visual interaction feedback in a controlled character input task with 27 participants in Augmented Reality (AR). It presents 5 different keyboard locations (3 bounded to the headset and 2 bounded to the non-dominant hand of the user) and 3 visual interaction feedback methods (finger raycast, fingertip glow and both combined). Objective (completion time, accuracy, Key per Minute (KPM)) and subjective (After Scenario Questionnaire (ASQ)) metrics are presented. Results showed that keyboard placement had an effect on accuracy, KPM metrics and subjective preference, with keyboard visualisation parallel and bounded to the headset position and orientation outperforming other keyboard locations.