I've added a new Q to the FAQ:
How can I use memory techniques to learn equations?
Generally, I try to avoid "memorizing" equations. Again, the goal here is to learn efficiently by giving tangible meaning to what you're learning. In the case of equations, true understanding should be achievable, so memory techniques should generally take a backseat.
That said, I do use memory techniques for specific pieces of equations I find difficult to remember. So I'd recommend trying to identify the one or two tiny things that are tripping you up, and to encode those things specifically. For example, take this equation:
dc1/dt = Q/V1*(cin-c1) - PS/V1*(c1 - c2/lambda)
Let's say you're struggling to remember that the lambda is under the c2. In an appropriate area of your palace, you might imagine a swan (for 2) landing on and crushing a lamb (for lambda). Or maybe the lamb's simply raising both its front legs.
I'll also use memory techniques to better recall the steps needed to reach a final equation, assuming I've already understood it and still find that sequence difficult to remember (and still believe knowing that sequence is important).