The best advice I could give is to get really comfortable using the whiz-wheel (or Angriffsscheibe, the "attack disk") included in the NewUIs mod. A phenomenal handbook for this tool can be downloaded from this thread:
http://www.subsim.com/radioroom/showthread.php?t=126824
My basic procedure can vary quite a bit depending on whether I'm surfaced or submerged (is it a visual or a hydrophone contact?), the likelihood of encountering enemy destroyers or aircraft if I choose to run surfaced, and the weather/visibility, among others.
My first step is usually to request a dead reckoning estimate from my navigator (when playing with Real_Navigation) and then to plot a relative bearing to the target on the navigation map. Before I do any real calculation, I try think about likely shipping routes to estimate the most logical destination for the target and, as a corollary, the most logical course required to reach that predicted destination. I find this to be a key step, as it provides a rudimentary means of evaluating the quality of true course and AOB calculations later. Sometimes (more often than I'd like to admit
), I find that my course estimate doesn't make much sense (i.e. is it really likely that this guy is running at 12 knots directly toward a desolate stretch of the Scottish coastline, or is it more likely that I made a mistake?---usually, it's the latter). Of course, this is a lot easier if I can actually see the target through the scope or UZO.
After that, I will bring the boat onto a very general intercept course. Then, I use the handy AOB finder someone here uploaded a while ago (sorry, I just looked for the link but couldn't find it; maybe somebody else can?). I take three bearings at equal intervals and then use the AOB wheel to calculate the AOB at the first bearing, writing all this down as I go. I then use the in-game whiz-wheel, plugging in my true course, the relative bearing to the target, and the estimated AOB. The whiz-wheel will then tell me the target's true course, provided my information is accurate and he doesn't zig-zag. Finally, I use the attack course pointer on the whiz-wheel to position myself on a perpendicular intercept with the target's course.
After I've begun my intercept, I use the back side of the whiz-wheel to get the target's speed after taking two bearings and applying a "speed correction" to account for the fact that the u-boat is moving relative to the target (example #10 from the handbook). Usually, I also use the "crossing the wire" method with the u-jagd stopwatch for one last speed check before I send the eels on their way. I like to recheck things whenever I have a free moment, so I usually recompute the AOB/true course at some point with an alternative method. Example #5 ("determining angle on the bow of a target from its aspect ratio") from the handbook link posted above has worked well for me
Occasionally, I will also plot the bearings on a sheet of graph paper while I wait the few minutes between bearing measurements. If done to scale, this is a quick and accurate way to get all the information you need to intercept and set up a firing solution. I guess it's sort of like using a maneuvering board, but I'm a bit uncomfortable with the watch officer's accuracy in calling out the range to the target. It kind of feels like cheating, even though I am the one physically plotting everything and measuring it.
Anyway, sorry for the super long-winded response. I'm no real expert at any of this, having learned every bit of it from the amazing community here at Subsim. That said, there's nothing like the tension and exhilaration of closing with and torpedoing a target based only on your own calculations. Good hunting to you!