Hi Jim,
I think noobies cannot attach schematics here until they are "passed" by a moderator to avoid SPAM. So the OP attached his schematic over at RapidShare who takes all day to wake up and is full of ads.
Soulfire - get a lightmeter, there's really no other way to do it. You are working with a filament resistance and light spectrum that change at lower temperature. Crank down the juice until you get 70% of what it is at spec. Use this empirical data to set the power.