Dear all

 A fractional-N pll with its reference frequency of 10Mhz and divide ratio of 10.3 will get a output frequency of 103Mhz. It's known that the fractiional spur will occur at 0.3x10Mhz offset the center frequency, but in real test spurs at 1M/2M/4M also exist. Could any body explain that why there are so many harmonic frequency appeared?

Thank you raja, I've read the paper.  It mentioned in Fig.11 that a 105Mhz reference with a fractional divide ratio of 255/1024 could have  
spurs at 410khz offset, with the author‘s equation I can't get a result of 410kHz，indeed 105M/410k is not a integer.