Novel error correction memory compression technique of a DDS using the equi-section division method
Kenichi Tajima, Kenji Kawakami · 2012
This paper presents a novel memory compression technique for error correction of a direct digital synthesizer (DDS) using the equi-section division method. In the phase-to-amplitude converter (PAC) of the DDS, the sinusoidal phase data of a one-quarter period is divided into 2Sequi-sections; then the equi-section type sine amplitude approximation (ESAA) is performed. An error correction memory is used to reduce error after the ESAA. In order to reduce the size of the error correction memory, we propose a phase adjustment technique to realize a shared memory. In this proposed technique, phase data which is equivalent to memory address is changed to optimum value in each section. To verify the proposed technique, we calculate the maximum spurious level at output of the PAC in the case of 8 sections. The maximum spurious level is −89.5 dBc. The proposed technique reduces the total size of memories to approximately 1/2Swithout affecting spurious levels.