Cree 1000W DE HPS Replacement Reference Design

[churchhaze]

but its for aquarium not plants

Spectrum in aquariums has more to do with how the guy running it wants colors to be rendered. (that's the only reason aquarium users use 10000k and actinic tubes. They think it looks better). The scoring system on that clearly has more to do with plant growth rather than usefulness in an aquarium.

 

[Abiqua]

exactly or at least cold white with high CRI 2700/3000k

sure i agree we shouldnt be trying to make super spectra

something like this?View attachment 3716251

or this?

View attachment 3716253


post up some proper graphs of sunlight please i wanna play around a bit more with the cold white idea

its tough to graph sunlight since it is a non static entity. I would say looking around the world if they have graphs might be quantitative? only representations because temp, wind, barometer, etc all "make" sunlight, too.....

 

[weed-whacker]

Spectrum in aquariums has more to do with how the guy running it wants colors to be rendered. (that's the only reason aquarium users use 10000k and actinic tubes. They think it looks better). The scoring system on that clearly has more to do with plant growth rather than usefulness in an aquarium.

sure just set it for fresh water not reef?

anyways its a fun tool, hope everyone has fun playing with it

 

[Sativied]

for a long time ive thought that we would be better off using cold white(5000-6500K ect) and reds

I like to use http://spectra.1023world.net/ to play with spectrums

here is 50% cold white with 20% 625nm and 30% 660nm
View attachment 3716200

Nice... Peak at 634 even and 660....Thanks, for the link too. Would be even better if it would produce a table with the ratios like the one in the cree guide.

That reference you showed appears to line up with the red (ypf) line in the graph in my first post, and does not account for the fact the short-wavelength photons carry more energy. See the weighted blue one instead. Which lines up with many (21st century) trials and as I pointed out across several threads there are plenty of reasons to avoid high blue.

High blue, and suboptimal spectrum at high intensities doesn't always have to be a problem in practice, and will certainly vary per species and cultivar too but there's no reason to aim for suboptimal either, especially with this cree design. I would try to close that gap between 634 and 660 a little with more of those 625. Or, if you can achieve something similar with lower K temp whites I would probably do that instead. Especially when that lowers the blue to a healthy minimum but also green to as little as possible, and add a little FR at the same time. I vegged stretchy sativa dom varieties for years and the whole blue for compact plants is taken too far by some. It's pretty much irrelevant after transition to flower anyway. Especially for the og kush and other afghanicas growers it should be a non-issue. We're not growing ornamentals after all.

If you want to use light on anything more than the orange-red peaks you simply need a good reason. Supplemental led (to daylight and even HPS), such as side lighting, is often solely for assimilation, nearly all red, sometimes pure red. Cree is also obviously not referring to taking pictures with their "photo red" leds. Adding anything else is steer/control light added for very specific tested and verified reasons and never because people thought the coolers white cobs are more efficient.

 

[weed-whacker]

Nice... Peak at 634 even and 660....Thanks, for the link too. Would be even better if it would produce a table with the ratios like the one in the cree guide.

That reference you showed appears to line up with the red (ypf) line in the graph in my first post, and does not account for the fact the short-wavelength photons carry more energy. See the weighted blue one instead. Which lines up with many (21st century) trials and as I pointed out across several threads there are plenty of reasons to avoid high blue.

High blue, and suboptimal spectrum at high intensities doesn't always have to be a problem in practice, and will certainly vary per species and cultivar too but there's no reason to aim for suboptimal either, especially with this cree design. I would try to close that gap between 634 and 660 a little with more of those 625. Or, if you can achieve something similar with lower K temp whites I would probably do that instead. Especially when that lowers the blue to a healthy minimum but also green to as little as possible, and add a little FR at the same time. I vegged stretchy sativa dom varieties for years and the whole blue for compact plants is taken too far by some. It's pretty much irrelevant after transition to flower anyway. Especially for the og kush and other afghanicas growers it should be a non-issue. We're not growing ornamentals after all.

If you want to use light on anything more than the orange-red peaks you simply need a good reason. Supplemental led (to daylight and even HPS), such as side lighting, is often solely for assimilation, nearly all red, sometimes pure red. Cree is also obviously not referring to taking pictures with their "photo red" leds. Adding anything else is steer/control light added for very specific tested and verified reasons and never because people thought the coolers white cobs are more efficient.

more pics less words bro

I used the data sheets and the amount of leds that cree used to calculate watts of 4000k and 660nm from their light and got this


compared to the one from cree



looks right to me


so are u suggesting to use less blue? so how about 2700K instead and 660 same ratios?




you likey?

 

[coolbreez1]

Nice... Peak at 634 even and 660....Thanks, for the link too. Would be even better if it would produce a table with the ratios like the one in the cree guide.

That reference you showed appears to line up with the red (ypf) line in the graph in my first post, and does not account for the fact the short-wavelength photons carry more energy. See the weighted blue one instead. Which lines up with many (21st century) trials and as I pointed out across several threads there are plenty of reasons to avoid high blue.

High blue, and suboptimal spectrum at high intensities doesn't always have to be a problem in practice, and will certainly vary per species and cultivar too but there's no reason to aim for suboptimal either, especially with this cree design. I would try to close that gap between 634 and 660 a little with more of those 625. Or, if you can achieve something similar with lower K temp whites I would probably do that instead. Especially when that lowers the blue to a healthy minimum but also green to as little as possible, and add a little FR at the same time. I vegged stretchy sativa dom varieties for years and the whole blue for compact plants is taken too far by some. It's pretty much irrelevant after transition to flower anyway. Especially for the og kush and other afghanicas growers it should be a non-issue. We're not growing ornamentals after all.

If you want to use light on anything more than the orange-red peaks you simply need a good reason. Supplemental led (to daylight and even HPS), such as side lighting, is often solely for assimilation, nearly all red, sometimes pure red. Cree is also obviously not referring to taking pictures with their "photo red" leds. Adding anything else is steer/control light added for very specific tested and verified reasons and never because people thought the coolers white cobs are more efficient.

I have a strain that showing "Sativa" traits, the amount it stretches at the beginning of flower seems to have very little to do with the color temperature or the light intensity. Light temperature and intensity seems to mostly have an effect on internodal spacing and leaf development during vegetative growth, but at the same time, my experience is the internodal spacing is also greatly effected by root ball and nutrient avalaility. . My best guess is that the stretch that happens at flower has more to do the potential range of bud size..................... This is why I need a bunch of damn sensors, to hard to realistically isolate variables without more reference information.

@weed-whacker I think that you are going to get much better results with the first of the two lights. I think the you have to consider that Emerson effect and I also think you need to consider that in most cases with the white COBs you are going to get a good 20% more photons out of a blue heavy light vs a red heavy one. This also effects the heat generation and cooling needs.

In fact the Vero 29 2700k-3000k is beat hand down by the 4000k because the 4000k simply puts out so much more light and actually his a bit more far red. The 2700k-3000k pretty much removes blue, increasing the relative amount of red without actually increasing the red. You are already adding in red, if you feel you need more relative red then simply add more red.

 

[Greengenes707]

 

[Sativied]

more pics less words bro

Not here to spoonfeed you pics, try to understand the meaning of those words and you can make pics yourself. Or better yet, combine those words with many other sources and make a more informed decision.

so are u suggesting to use less blue?
View attachment 3716328
you likey?

Obviously less blue equates to more of another color so there's a little more to it. Not entirely what I pictured with words but for flowering, I do "likey"-ish. It's a bit low on blue lol... How about some royal blue

Greengenes' looks much better than 3500k and I like it more than the Cree ref spectrum too. In that one I would prefer to remove some of the 510-560 region and spend that on 630-660 instead which would relative drop the blue a little and bump the red a little if possible/practical and not with a major efficiency loss.

Ok, fine, with a disclaimer... don't pretend this suggests I suggest the ideal spectrum for cannabis. This would come closer to the what I had in mind with my words.

144 xpe warm white (2700-3700k), 48 photo red

Want more blue, replace one photo red per engine with white..

Cri 95, R9 96


More pics lol


Like I said in the first post, either more red leds or warmer whites with higher red. It appears that the latter is an easy way to go about it. Anything lower than 4000K, but same W:R. Obviously something similar, and certainly similar enough, can be achieved with different variations, but you obviously want a somewhat even spread per engine and per the entire light. If you can get a lot more efficient red (630) than photo red I would use those too.

For example, per engine with 16 leds, 3 photo red, 2 red, 2 cool white, 5 warm white... Edit: make that a 12 led engine lol, you get the point, or not...

The advantage over using white only is obviously the more granular control than picking spectrum based on K temp alone, yet the more different colors you use the less the uniformity of the spectrum across the entire canopy will be compared to white and photo red alone.

 

[weed-whacker]

View attachment 3716444

whats the recipe for this?
I like that you have a lot more 620ish, why? thanks for sharing

 

[loftygoals]

Quite tempted to make a 1/4 size 130w version and test it out in a 2x2 tent. Possibly a project for the end of the year... PDF saved

 

[coolbreez1]

The performance difference between the 3000k-4000k has largely been eliminated for the new version of the Vero 29 that they just released the specs for. The 3000k is 155lm/w the 4000k is only 161lm/, this means the 3000k is likely the better way to go, augmented with more 660 red.

 

[JorgeGonzales]

The performance difference between the 3000k-4000k has largely been eliminated for the new version of the Vero 29 that they just released the specs for. The 3000k is 155lm/w the 4000k is only 161lm/, this means the 3000k is likely the better way to go, augmented with more 660 red.

I'd humbly suggest comparing their actual photon output instead of lumens. The difference is usually neglible between CCTs, even favoring the lower color temps.

There's still a lot to process in those new data sheets, and obvious typos in the lumens column doesn't help. They look promising, however.

 

[coolbreez1]

I'd humbly suggest comparing their actual photon output instead of lumens. The difference is usually neglible between CCTs, even favoring the lower color temps.

There's still a lot to process in those new data sheets, and obvious typos in the lumens column doesn't help. They look promising, however.

Even if this is true, according to the new sheet on the Vero 29s the peak red output on the 3000k is shifted 10-15nm up, when compared to the 4000k. This puts the peak red on the 3000k closer to 620nm vs 600nm on the 4000k. With only a 2-3% output difference I think this shifts things far more in favor of the 3000k.

 

[JorgeGonzales]

Even if this is true, according to the new sheet on the Vero 29s the peak red output on the 3000k is shifted 10-15nm up, when compared to the 4000k. This puts the peak red on the 3000k closer to 620nm vs 600nm on the 4000k. With only a 2-3% output difference I think this shifts things far more in favor of the 3000k.

It's true that comparing lumens across CCTs is pointless for our purposes. Can't speak to the rest, because I haven't looked yet!

 

[PurpleBuz]

Even if this is true, according to the new sheet on the Vero 29s the peak red output on the 3000k is shifted 10-15nm up, when compared to the 4000k. This puts the peak red on the 3000k closer to 620nm vs 600nm on the 4000k. With only a 2-3% output difference I think this shifts things far more in favor of the 3000k.

Having the peak closer to 600nm is actually advantageous when supplementing with 660nm deep reds, as it will fit the mcree curve better. That whole 580nm to 660nm ie yellow-orange-red deep red colors is the beef. I mean its all needed and in abundance imho.

 

[weed-whacker]

ok check this light recipe out

basic
50w of cold white(6500-8000k)+150w of warm white(3000k) + 20w of 660nm +15w of 625nm








or

complex


50w of cold white(6500-8000k)+150w of warm white(3000k) + 20w of 660nm +15w of 625nm + 5w of UV(2w 400nm+3w 410nm)+ 5w of 640nm+5w of 650nm








I know this is not too interesting for a lot of you but i feel a discussion of spectrum is interesting, plus after pages of trolling and flaming i figured we could get this thread back on track and all learn a bit more from eachother

 

[OLD MOTHER SATIVA]

good idea

so are Cree trying to" improve on de".. with just mono's?

[so many to have to connect for diy]

or is this adding new cree monos with white cobs?

 

[loftygoals]

Asked sinkpad for a quote on 50 units of the 1950 board.

"$3.75 ea and the Minimum order is 28 boards"

Actually reasonable... the LEDs and drivers are available. Heatsinks can be sorted easily. If the optics are reasonable too then this reference design could be a nice little project.

Edit: optics can be had for $4.99 each on orders over 36.

 

[Mr celsius]

Asked sinkpad for a quote on 50 units of the 1950 board.

"$3.75 ea and the Minimum order is 28 boards"

Actually reasonable... the LEDs and drivers are available. Heatsinks can be sorted easily. If the optics are reasonable too then this reference design could be a nice little project.

Edit: optics can be had for $4.99 each on orders over 36.

Thank you for looking into this. Maybe if you said the exact cost per board, we could get a group buy together?

 

[Greengenes707]

So the big thing that is left out...and actually I would bet my good nut that it has never happened by cree or anyone using this exact reference, so it's not technically "left out" of the paper, but where are the horticulture results from this...
You know...GROWING SOMETHING WITH IT...
Cannabis, tomatoes, peppers, lettuce, arugula, wasabi, mushrooms...anything.

Not saying it doesn't work...I've tested VERY similar spectrums....and it will work just fine. But is it the true horticulture reference...and if so how did they decide that. Based on what results? Who did them? When? Where? On What? How many times? What others were tested?
Or do they do as many light maker companies do....and just listen and make what they see rather than lead by superior example as a horticulture company would.