I respectfully disagree with Seamus, mostly because of his rather audacious use of the word "need". I don't think a philosopher needs to know anything in the sciences—in some cases they ought to in some regards, but they don't need to. The first (ancient) philosophers of course started in exactly this way, and they got along fine. Even centuries later, when the sciences were moving full steam ahead, philosophers like Immanuel Kant* and Hume and many others proposed many great theories without the aid of science that are still accepted today, because they were founded on logic and reason; phenomenological experience being the only thing they needed to start.
Furthermore, in some ways scientific theories can be perniciously misleading. Science is, after all, a collection of constantly revised theories that go through rather large changes from time to time, called paradigm shifts. Basing your arguments on potentially false theories in science could have obvious potential drawbacks. There were points in history with concepts like the flat Earth, the Sun revolving around the Earth, Atlas holding up the sky, souls being located in the pineal gland, etc... To have based a philosophical theory on such things would have been disastrous.
That said, I think it is the duty of every philosopher to never stop learning more, and those who continue to seek knowledge will always be more well-rounded and more prepared than those who do not. As Seamus and others have suggested, if you are interested in a particular field of philosophy that relates directly to a science (say, the philosophy of quantum mechanics), obviously it behooves you to learn about said field from as many perspectives as possible. Outside of that, the bad philosophers and the good philosophers will always be divided simply by their ability to form rational conclusions**.
*I'm aware the Kant was a scientist and taught many classes in the sciences. But he also created many theories regarding the scope of human knowledge and the limits of the human mind when there was no formal understanding of psychology in his time.
**Note that I don't think one's ability to use first-order and predicate logic makes a difference in their ability to reason; that is, the ability to form logical conclusions and to tell when an argument is invalid is independent of being able to use the formal notation.