Which bar lights emit the least heat?

[Samwell Seed Well]

Saw that, wanted to ask about it. That driver looks huge. and adds to price?

Ya, you can see i here in a photo of a 400 watt bar... those panels can't be more then 3-4 inches so not to big..

Proprietary power sources always tricky if and when the sources dries up or worse

Standardizes

 

[Markshomegrown]

Quite relevant to the above, here is a very informative video of somebody building an array of LED bulbs and doing precise measurements on the efficiency.

It's striking how much it helps to cut the diffusers off the bulbs.

He gets around 1.8 umol/J efficiency; really good in my opinion.

This is 1.8 umol/j with 6 bulbs, overlapping the light from each bulb will be more effective with a larger panel but this is the same with any led you buy, adding the reflectors makes them even more effective.

 

[Apostatize]

This is 1.8 umol/j with 6 bulbs, overlapping the light from each bulb will be more effective with a larger panel but this is the same with any led you buy, adding the reflectors makes them even more effective.

I see diy can ~get you there μmol/J-wise -- and it's tremendous what you've done -- but what's the PAR range look like? To what extent can you copy a spectrum out there in the market (apologies if you pasted the PAR range on an earlier comment)?

I'm not familiar with the arrays of bulbs available online. For example, does Osram make a far-red bulb?

 

[Markshomegrown]

I see diy can ~get you there μmol/J-wise -- and it's tremendous what you've done -- but what's the PAR range look like? To what extent can you copy a spectrum out there in the market (apologies if you pasted the PAR range on an earlier comment)?

I'm not familiar with the arrays of bulbs available online. For example, does Osram make a far-red bulb?

Thanks Apostatize, The Philips 13w 2700k look a very effective bulb, from this picture on their website,

So that's 24 x 2700k(13w) bulbs and 4 of the 4000k bulbs(12w) total of 360w
1.1m x 1.1m = 1.21m2, 360/1.21 = 297w per m2
across the canopy its between 38k(corners) and 45k max, so the average is about 42k


799 / 297 = 2.69 watts uMoles/j
not saying the DIY led is this effective but it could be closer to 2.5 (useable uMoles/j)
Lots of LEDs base their uMoles/j on the center being really high, 1500 par

lift the panel up, so the center reads 800, and the uMoles/j drops down, easy 30%

 

[ComputerSaysNo]

what's the PAR range look like? To what extent can you copy a spectrum out there in the market (apologies if you pasted the PAR range on an earlier comment)?

The guy from the video answers that question in another video: the exact spectrum does not matter much with regards to PAR, as long as you don't go too far into one extreme.

So you can use 2700K, 3000K, 4000K, 5000K lights and they'll all be very close in PAR.

The red lights added to modern grow lights make the light more efficient, but the LEDs are also more expensive than the white ones, so that's why they only use a few of them per board.

This was one of the questions that was nagging me, but he answered that very well, and it's all backed up by measurements.

So with the modern lights you spend (a lot) more money upfront, mainly due to the expensive red LEDs being added, but then you get more PAR efficiency, so you might make it up with additional yield, or respectively a lower power bill.

I think going with simple LED bulbs is extremely good to keep mother plants, for small veg areas, and cloning, simply because you just need adequate lighting and not more. Maybe for flowering a better light will pay off in the long run, especially with high energy costs (in my area energy is ~0.3€/kWh...)