LED Lighting for orchidbox

Maintains a constant current, the actual value of which depends on the model. Voltage will vary depending on the load, but current will remain constant over a certain range.

You can change this somewhat (for dimming purposes) by either putting a resistor across the DIM connections, a voltage (1-10V DC), or a 10V pulse-width modulated signal.

 

Clusty, you are misunderstanding what carl said.

"D" model, which I linked, has the dimming capacity (you have to pay a bit more than the non-dimming version). It has Meanwell's 3-in-1 dimming function. There is a set of wires (DIM+ and DIM-), which are used to control the output constant current, and it can dim down to 10% or so (so 0.188-1.88A). But efficiency will suffer if you dim down too much.

For the dimming control, you can use 1-10VDC, 10V PWM, or variable resistor (this is why it is called 3-way). If you are controlling it from Arduino or something, you might want to use PWM. But for the manual control, variable resistor is the cheapest and easiest. Basically you connect 100k ohm variable resistor, and you can adjust the brightness by twisting the knob. When it is set to 100K ohm, it outputs 100% (1.88A). When you lower the value to 10K ohm, you get 10% (0.188A). For my 4x CXB3590 rig, I additionally use this feature for thermal protection. I attached normally open bimetal thermostat (something like this, but you might need to experiment with the cut-off temperature) to the heatsink and connect them in parallel to the variable resistor. So if CPU fan malfunction and heatsink overheat, it will reduce the current to 10%, and protect the COBs. But with the smaller Samsung strips in passive cooling, I think this heat protection is probably an overkill.

 

Sounds good, Clusty. What did you decide to go with?

I'm not completely sure what you mean by "spread" the color temp. Do you mean use several color temp? Or do you mean using a high CRI LEDs? From the perspective of PAR (photosynthetically active radiation), it probably doesn't matter too much. I usually go with 3000-4000K, but if there is a better deal with 5000K, I wouldn't hesitate using it. With CRI, it used to be lower CRI was better in terms of PAR efficiency, but more recently, the difference seems to be small, too.