Hoping some of you guys can help me with a big problem. I am having a hard time passing a 500 candela test with an LED array pulsing at 40ppm utilizing the modified allard method. Any suggestions on optimum pulse widths or mathmatical gymnastics that could aid the situation?