In a recent work [Phys. Rev. E 91, 022808 (2015)] it was reported that placing an obstacle in front of a gate has a beneficial effect in the flow of sheep through it. Here, we extend such results by implementing three different obstacle positions. We have observed that the flow is improved in two cases, while it worsens in the other one; the last instance happens when the obstacle is too close to the door. In this situation, the outcomes suggest that clogging develops between the doorjamb and the obstacle, contrary to the cases when the obstacle is farther, in which case clogging always occurs at the very door. The effectiveness of the obstacle (a strategy put forward to alleviate clogging in emergency exits) is therefore quite sensitive to its location. In addition, the study of the temporal evolution of the flow rate as the test develops makes evident a steady behavior during the entire duration of the entrance. This result is at odds with recent findings in human evacuation tests where the flow rate varies over time, therefore challenging the fairness of straightforward comparisons between pedestrian behavior and animal experimental observations.