You should read that link I put up. It required much more than this to suppress the noise from a Lumatek 1k.
The Fix
Attempts to tame the RF interference from these electronic ballasts is much more than wrapping just a few turns through a torroid. We were able to borrow a MicroMole 1000 ballast, 1000W HPS lamp, and set it up in the basement of W0IVJ. The RF interference on 40 meters (7 MHz) was a constant S9+25dB as measured using an outside dipole at a distance of about 40 ft. and a Flex-1000 SDR. We focused on 40 meters where the interference was the worst but all HF bands were obliterated. These measurements were made on Nov. 7, 2011.
We wound 10 turns of the input and output cables through very large #31 toroids (Fair-Rite #2631814002) which are 4"O.D. These common mode chokes (CM impedance measured >2000 ohms) reduced the interference 15dB to S9+10dB which was still terrible. Building a filter for the AC line which consisted of a good common mode choke on just the Line and Neutral wires and 4700pF 250VAC rated caps to the safety ground/chassis (still big toroid on output) reduced the interference to S8. Finally building an output filter similar to the AC input one reduced the interference to below the ambient level of S5. However the output filter ran quite warm and the capacitors on the output needed to be much higher voltage if the product label of 5kV is to be believed.
http://www.w0qe.com/RF_Interference/grow_light_electronic_ballasts.html