Here's a suggested simple circuit to drive a conventional modellers servo with a pulse at ~20mS intervals with a width (1-2mS) controlled by a light-dependent-resistor (LDR). Depending on the characteristics of the LDR (here modelled as a 33k resistor) R1 and R2 may be unnecessary.
Thanks for the fantastic replies everybody. With respect to the last poster, could you provide a brief explanation of how this circuit works? Just to make sure I understand what's going on correctly.
Sorry for the late reply; my forum settings have been screwed up.
U1 is a Schmitt inverter. When its output is high it drives current via D2, LDR2, R3 to charge up C1. When C1 voltage reaches the Schmitt input upper threshold U1 output goes low. C1 then discharges through R1, LDR1, D1. When C1 voltage has dropped to the Schmitt lower threshold U1 output goes high again....and so on. The result is a pulse train from U1 with a 'mark' determined partly by LDR2 and a 'space' determine partly by LDR1. (The resistance values shown in the schematic for LDR1 and LDR2 are purely examples for LTSpice modelling purposes). When U1 output is high it turns on Q1 to pull the servo input voltage low.