Is this a radio control servo?
If so, they generally run off of +5V and ground, and the control signal is a
1-2 msec positive pulse at anywhere from 20 to 50 Hz pulse repetition
frequency. 1 msec pulse width (nominal) is hard over one way, 2 msec is
hard over the other, 1.5 msec is midrange.
Bypass the servo supply leads with a cheap 250 uF electrolytic capacitor, if
you build your own control circuit. I played with a Royal Titan Standard
servo several years ago. It drew 250 mA in smooth motion, and upwards of
500 mA while buzzing. That 250 uF capacitor was empirically determined to
be "about right" to stop the buzzing. (What was happening was that the
servo was putting hash back into the supply leads, and screwing up my 555
control circuit. Adding the capacitor cured it.)
Best Wishes
Here's a web page that might help a little bit.
http://www.seattlerobotics.org/encoder/200009/Servos.html
NE544 was a Signetics part, long since discontinued. The circuit diagrams
in the page appear to be reproduced from an old Signetics databook.
Here's the datasheet for the M51550L.
http://www.mitsubishichips.com/data/datasheets/assps/assppdf/ds/M51660E.pdf
It doesn't say anything about the control signal waveforms.
Here's a discussion of digital proportional R/C systems.
http://www.rchelibase.com/radio/radio.pdf
This will probably cover what you need.
You might try writing a letter to Sanwa, telling them that you are a student
doing a project with one of their servos, and asking NICELY if they can send
you some literature. This may not be practical depending on how long it
takes mail to get to where you are. I don't know the .my domain; I'm
guessing Burma (Myanmar).