How are you doing the docking? (What movers are you using?) Often with the docking movers there’s ways (settings you can pass) to limit the allowed range of translation they make. Alternatively, if you already know that the helix is in the rough binding orientation (e.g. within a few Angstroms in translation and no more than several tens of degrees off on rotation), you may be able to skip the low resolution stage of the docking protocol, and just use the high resolution refinement step.
Another option is to define the binding site/orientation that you’re interested with constraints (restraints), and use those bias the docking protocol. This works well if you know of particular interactions, but are unsure of position/orientation.