I work in the Physical lab. I don't have much experience with optical items so another tech helped me get this done. We weren't able to test the whole range at the time, but tested enough points to get a line fit on the data and due to its small non-linearity I feel comfortable extrapolating beyond the calibrated range.
The standard we used is a calibrated bulb at 5500K that at 6.9 amps provides an output which at certain calibrated distances equates to a value in foot candle units. The light meter was mounted on a calibrated rail to ensure the proper distance from the bulb was applied. I found a simple conversion from foot candles to par and used that. We could have tested higher, but didn't want to tie up working equipment for too long. 100 par was good enough for me and as stated before I have confidence in my unit to remain linear beyond 100 par. This calibration beats standing outside at noon with your light meter any day of the week...
Here is the data I took.
| TI | STD |
| Par (mmol) | Calculated PAR |
| 2.4 | 2 |
| 5 | 4 |
| 8.8 | 7 |
| 17.6 | 14 |
| 25.2 | 20 |
| 50.6 | 40 |
| 75.5 | 60 |
| 100.6 | 80 |
| 114.7 | 91.2 |
| 125.4 | 100 |
I used these values to generate a slope and intercept.
| Slope | 0.795914548 |
| Intercept | -0.029186939 |
This can be applied to my unit to correct for the discrepancies(PAR Reading x Slope + Intercept). The worst error for the calculated outputs was 0.244% of reading at 40 par. I think that's pretty good for the price of this meter and decent enough for checking basic light levels like I am.