Mehmet,

You've received quite a bit of input in this thread. Consider the following before choosing your solution -

(1)Use of a quadrature encoder requires the special processing of the A and B output tracks to prevent incorrect accumulation of up or down pulses in the position counter. The Agilent component spec sheet can hep you sort out what that processing is.
(2)Once the A and B track processing is complete, the pulses must be accumulated in an up/down counter to convert from incremental plus direction format to absolute position format.
(3)Once the absolute position format is available, the micro must be able to access it and use it. Consider how you wish the micro to access it - parallel or serial. If parallel and more than eight bits are used to represent the position, then multiple accesses need to me made or a software register implementation is necessary to deal with under/overflow of the up/down counter. If serial, then when the micro accesses the position data the up/down counter must transfer its value to a shift register and the shift register used to transfer the position count into the micro.

With these things in mind, you can evaluate whether to use a commercially available chip or whether to "roll your own" using a CPLD.

This past week I had a quadrature encoder connected to a step motor. I watched the encoder A and B tracks on an oscilloscope and what an eye opener that was. At low speeds, motor mechanical resonance caused incorrect sequencing of the A and B tracks, but at high speeds the oscilloscope showed an ideal waveform. The same thing will happen using any type of torque producer.

Bruce