During last year’s GBE Symposium, several groups investigated how changes to lighting can affect experimental outcomes. My observations and recommendations on this topic were provided to MARSfarm and Fairchild, and are attached here: https://wxanalyst.com/20220512_nPAR_Zero_Sum.pdf.
My primary recommendation is to keep PAR constant while changing the spectral distribution within the 400nm to 700nm wavelength interval (which defines PAR).
My observation is that this “zero sum” change can be difficult to accomplish, since the irradiance output by each GBE light channel (RGBW) is affected by the other RGBW light settings.
Sorry for being slow to respond on this, but thought I would share some work I did with the GBE LED unit a while back. I was researching controlling the LED via a Raspberry Pi (PWM control), and ended up working with an LED amplifier (signal booster for long LED strips).
For PWM to work the lights have to have separate channels, but interestingly, all the LEDS feed back to a common ground wire, thus, as you described; changing the input for one channel impacts the current (Amps) for all - not a linear response. I had a PAR meter for a while and ran some experiments, the linked spreadsheet is the results.
[GBE PWM Testing]
(Loading Google Sheets)
Also, like your comments on safety.
@hmw Wow, so the red channel dropped from 2400 mA to 1290 mA. That definitely validates the observation @wxazygy made in his report about why lux is higher than PAR when everything is set to max. Really cool to see those spectral graphs - wish I had a spectrometer!
It looks like 1290mA is the equivalent of a knob setting at ~165 for red - that’s not very good. After we looked at the circuit and chip datasheets, we concluded that the power supply cannot carry enough current (4.5 amps) for all channels to be run at the max setting simultaneously.
Scott, I really like your idea of making “light intensity” an independent variable, so that the focus of the experiments would focus on the distribution of that PAR throughout the spectrum. We’ve seen teachers use your idea of an overall “light budget” with light intensity and photoperiod. This worked well in situations where light intensity and duration were possible to manipulate but spectrum was not.
