Multiplying PAR W by 45 can't be right. I think SupraSPL uses a constant between 4.47 (5000K) and 4.66 (3000K) depending on colortemperature.
I keep seeing these calculations with PAR W and such for getting the umol/s values. Instead of going through all these calculations for efficiency and PAR W which are all depending on constants and then multiplying by some other constant, why not divide by a single constant?
If we take all the individual steps we get:
Efficiency=(lumen/W) / 325
par W = Total W * Efficiency
umol/s = PAR W * 4.66
To calculate umol/s in one formula we get:
umol/s = (Total W * (lumen/W / 325)) * 4.66
Which simplifies to:
umol/s = lumen / (325/4.66) =
lumen / 70
or for 5000K to:
umol/s = lumen / (325/4.47) =
lumen / 73
I took a list of umol figures that
SupraSPL posted:
Cree CXA3070 3K AB @ 52W = 2.11 umol/s/ dissipation W
Cree CXA3070 3K AB @ 25W = 2.46 umol/s/ dissipation W
Cree CXB2530 3K U2 @ 18W = 2.4 umol/s/ dissipation W
Cree CXB3070 3K AD @ 50W = 2.34 umol/s/ dissipation W
Cree CXB3070 3K AD @ 24W = 2.7 umol/s/ dissipation W
Cree CXB3590 3K CB @ 49W = 2.52 umol/s/ dissipation W
Cree CXB3590 3K CB @ 23W = 2.86 umol/s/ dissipation W
Cree CXA3590 5K DB @ 24W = 2.96 umol/s/ dissipation W
Looked up the lumen for those and arrived at a conversion factor of around 65. so i use:
umol/s = lumen / 65
The luminous flux values that Cree reports have a 7% tolerance, so it's not like it's all very exact anyway.