frequency multiplier spur

this question make me confuse: when I use a signal which was produced by DDS to multipliter, would the spur of signal be worse like the phasic nosie? 20logN? Is there any book refer this question ? I hope some guy could help me,thanks!

Depends on what the spur is caused by.

If the spur is a phase modulation sidetone, then its peak deviation will increase as a factor of N with the frequency multiplier, so yes 20 Log N increase in spur.

If the spur is AM based, and you have NO am to pm conversion in your multiplier, then the mutiplier will probably reduce the spur level, as it clips the envelope to be smoother.

Remember that a ssb sidetone is half AM and half PM!