I think the lux is divided by 60.6 for umol.
This assumes the PAr is 400-700nm though, lux meters measure the bell shape at 555 nm.
If the source of light is standard, this should work okay.
Better than nothing.
The other thing is to tie the probe next to a PAr meter and go a measuring various light sources and distances and see how close things are to the 60.6 conversion factor(this is what I'd do).