I was speculating two oscillators with periods p and q could be composed (as long as there was no way for them to interact) to create an oscillator of period p*q/gcd(p,q) but wondering why large primes wouldn't be a problem.

I guess this is my answer.