3500k cob specrum question

Nicholas Daniel

New Member
So after looking at spectrum graphs it looks like 3500k is the best all around temp. However looking at the charts of one in at full intensity I only get a small portion of the spectrum and I I run at low relative intensity I get the majority of the spectrum.

My question is how is relative intensity (y axis) measured. Is in watts or amps? Of the Vero 29 can handle 4200 miliamps and I use a driver that uses 1400 is that led running at 33% relative intensity?

Am I understanding this correctly?
 

CobKits

Well-Known Member
So after looking at spectrum graphs it looks like 3500k is the best all around temp.
depends what you want to do. thats def a good color to cover both veg and flower but a lot of people like 3000k for flower and 4000k for veg

However looking at the charts of one in at full intensity I only get a small portion of the spectrum and I I run at low relative intensity I get the majority of the spectrum.
not following you

My question is how is relative intensity (y axis) measured. Is in watts or amps?
its unitless, and relative. spectrum is almost independent of current (there is a tiny shift, likely irrelevant for our application)


Of the Vero 29 can handle 4200 miliamps and I use a driver that uses 1400 is that led running at 33% relative intensity?
it would put out about 40% of the light at 1400 mA vs 4200 mA. spectrum is almost the same between both currents
upload_2017-7-21_11-1-7.png
 
Last edited:

Rahz

Well-Known Member
Of the Vero 29 can handle 4200 miliamps and I use a driver that uses 1400 is that led running at 33% relative intensity?

Am I understanding this correctly?
The intensity curve isn't linear, that's part of the reason people use low relative currents. At 4200ma the Vero 29D (3000K 90CRI, electrical efficiency will be similar for other samples) will be about 35% efficient. At 1400ma = 46% efficient.

4200ma = 60 PAR watts
1400ma = 23 PAR watts
So about 38% relative intensity.

The C version is a much better chip for a few dollars more and can save money either up front or in electricity depending on how you use it.
 

Nicholas Daniel

New Member

Looking at this graph, the higher the relative spectral power (y-axis) the less usable red and blue spectrum is available. Using 3000k, 80CRI if I am at 70% relative spectral power, then I am not getting any blue spectrum and a smaller amount of red spectrum. Whereas down at 30%, I get blue spectrum and a much larger amount of red and even yellow spectrum.

I am going to be running 8 - bridgelux vero 29SE (3500k 80cri poke-in) cobs driven with 4 - meanwell HLG-185H-C1400B (200W/1400mA) so each cob will be driven at 100 watts each for a total of 800 watts. I know if I get 4 more cobs and run them all at 60-70 watts each ill still have my 800 watts but it will be more efficient giving more lumins per watt but how can I run this so that it is at the best relative spectral power?

Rahz you are saying that using a 1400mA driver will run the 4200mA chip at 38% relative intensity regardless of wattage used correct? I interpreted the above graph correclty right so that the wattage will effect efficiency and amperage effects spectral intensity
 
Last edited:

Rahz

Well-Known Member
Rahz you are saying that using a 1400mA driver will run the 4200mA chip at 38% relative intensity regardless of wattage used correct? I interpreted the above graph correclty right so that the wattage will effect efficiency and amperage effects spectral intensity
Running a chip with a 1400ma driver will result in a specific wattage being used. In this case 48.5 watts. The voltage at 1400ma is 34.6 so 34.6 * 1.4 = 48.44

volts*amps=watts.

The voltage changes slightly depending on the amperage used. You can get exact figures using the Bridgelux Product Simulator easily found with a search.

Anyway, At 1400ma the Vero29D is 46% efficient, so 48.5 * .46 = 22.3 PAR watts.

At 4200ma the Vero29D is 4.2 amps * 42.1 volts = 172.5 watts at 35% efficiency = 60.375 PAR watts.

So 22.3 PAR watts is 37-38% the intensity of 60.375 PAR watts.

I may have confused you using the term relative intensity. The graph you posted is a relative spectral distribution graph and won't show you electrical characteristics. It's just a curve that illustrates the spectrum, though it can be used to determine the amount of lumens in a PAR watt which is used to derive the efficiency figures I mentioned. The lumens per par watt for the 3000K 90CRI spectrum is 280, so if a cob provided 140 lumens per watt that would be 50% efficient. You can use the Bridgelux Product Simulator to see the lumens per watt for various Bridgelux products at whatever current you might use.
 

xX_BHMC_Xx

Well-Known Member
The graph you posted is a relative spectral distribution graph and won't show you electrical characteristics. It's just a curve that illustrates the spectrum, though it can be used to determine the amount of lumens in a PAR watt which is used to derive the efficiency figures I mentioned.
This. The graph posted above only shows the relative power of each nm compared to each other. That graph won't change (noticeably anyway) running the chips at higher or lower current.

Running chips at lower current is pretty much always more efficient than higher currents.
 

CobKits

Well-Known Member
Looking at this graph, the higher the relative spectral power (y-axis) the less usable red and blue spectrum is available.
its relative to total output of a color at a given measurement

notice how each color touches 100% somewhere
Using 3000k, 80CRI if I am at 70% relative spectral power, then I am not getting any blue spectrum and a smaller amount of red spectrum. Whereas down at 30%, I get blue spectrum and a much larger amount of red and even yellow spectrum.
regardless of current, "spectrum" aka the relative color distribution, is almost identical at different currents, i.e. the ratio of red to blue is the same.the graphs are misleading as when the chips convert light from the blue chips to reddish white light via the phosphor layer, its not perfectly efficient but its pretty high. in other words you sacrifice some photons to go from blue to red, but the red light is so much more readily absorbed that spectrum wins out over output in this case. which is why someone would select 3000k over 6500k for flowering, even thought the "efficiency" in "lumens/watt" is a lot less

plants dont care much for lumens, they want PAR and the red is their flavor of choice as its more absorb-able. whenever you see someone quoting "ppfd/W" or "umol/J" this is a great reference but still leaves room for variation in useful light to plants as the "PAR" reading equally weights all spectra from (usually)400-700nm and some spectra are better absorbed than others. So say 1000 ppfd centered from 580-730 nm would be more effective than the same ppfd centered from 285-455 nm. so "PAR" measurements by themselves are only one part of the picture, you also need to understand the spectrum you are applying. there are times you need to apply "specialty lighting" with more blue like veg and early flower to avoid stretch, or late in flowering to push trich production. but in general from a metabolic perspective with most LED spectra cannabis wants light around 2000-3500k depending on growth cycle. Thats the limit of my understanding, id love to hear some other people expand on this

@Growmau5
@Greengenes707
@Rahz
@Abiqua
@stardustsailor
@captainmorgan
@The Dawg
 
Last edited:

Rahz

Well-Known Member
Higher K low CRI to produce the shortest plants in veg. For most applications I think 4000K 70-80 CRI is about as high as the K needs to be though there will be some disagreement there. Some people like to use 5000 or 5600.

Although if you want a mother plant you can take some nice tall clones from then the mother clone chamber might benefit from a 3000K 90CRI spectrum.

A full cycle light, 3000K 90CRI wouldn't be a bad choice since there's less stretch in flower.

A veg only lamp where plants are bulking up for flower, and there was a need to keep the plants as short as possible, high K low CRI. I can't comment on the height difference between 4000/80 and 5000/70. I don't know if there would be a substantial difference but if headroom isn't an issue I'm not sure it would matter in which case I would choose 4000/80 to keep the orange levels up. 3000/80 or 3000/70 might even be appropriate.
 

Profound Bastard

Active Member
I think the proper way to do this is like @stardustsailor was doing here, with the traditional science literature-y breakdown of B/G/R at 400-499nm, 500-599nm, and 600-700nm.

https://www.rollitup.org/t/cxa-3000k-80-vs-93-cri-an-estimation.833171/

3000K, 4000K, 80, 90 it's all meaningless until you break down the SPD charts and look at how much power is where and at what ratio things are at. It varies too much between manufacturers to even talk about CRI or color temp. One makers 3000K is closer to 3500K, and another's should be labeled 2700K. A potential 50% swing in blue power. Anything not 90+ CRI is mostly green light.

Sunlight is roughly equal R and G, with 25% blue at midday...6500K or so. Not saying that should be a goal, but it's something to think about.

Maybe somebody could look at newer mid power/Vero/Citizen/Luminus stuff and see what's up in the 80 and 90CRI options. But again, discussing CRI and K without dissecting the spectrum only clouds the water, it's just not apples to apples.
 

Rahz

Well-Known Member
It's a pretty good general rule, though you are right that there are minor variations... though not as much as you suggest. If you look at the SPDs for Citizen, Cree, Bridgelux their 80/90 CRI samples show similar relative blue levels across the board. Certainly not 50% different. It's also wrong to suggest anything not 90+ CRI is mostly green. Most brands 80CRI has a peak at 595-610 while 90CRI will peak around 630.

4000K 80CRI will have more green than a higher CRI or lower K sample but will still be predominantly yellow/orange. @doz is setting up a test at the moment to compare the growth rate and yield of 4000/80 -vs- 3000/90.
 

Nicholas Daniel

New Member
Thanks for clearing that up guys. I was looking at relative spectral power as a variable that would adjust not simply a descriptor. Makes since now. what do you guys think is the best chip for an LED to be used veg and flower. Its the Vero 29 there is the 3000k and 3500k 80 CRI in the D chip, 3500k 80 CRI in the C chip or 3000k 90 CRI in the D chip. I dont kow the subtle differences in chip or how CRI affects color spectrum. which would be best for veg and flower or is the difference negligible? with the above chart the 3000k 90 CRI D chip looks like it would be the best but I dont know how it compares to the 3500k 80CRI with the C chip. what is the difference between the D and C chips?

Thanks again
 

Rahz

Well-Known Member
The C chip has more diodes and will provide more light at any particular wattage. They each have different voltage requirements so that will factor in to the driver choice.

As far as spectrum, we're still in the process of figuring out what is best for yield but there is some evidence 90 CRI reduces flowering time. 90 CRI does cause a bit more stretch in veg but also less stretch in flower so it wouldn't be a bad choice for full cycle or flower only.
 

Profound Bastard

Active Member
It's a pretty good general rule, though you are right that there are minor variations... though not as much as you suggest.
Truly not trying to be argumentative, but they are not minor. Just to be clear, the 50% difference I was referring to was the ratio of R:B, as you jump between CRI and manufacturers. Keeping nominal color temps the same, that you can easily hit a difference that big.

Pulling numbers out of my ass, a 3000K chip might have 12% to 16% blue between manufacturers at 80 or 90 CRI, with red percentages anywhere from 41-47% of the spectrum from 400-700nm. That's a pretty big difference. ~25% of blue as percentage of red, vs ~40%? An increase of ~60% give or take for two chips labeled "3000K".

Does cannabis care? No idea. But many publicly accessible experiments with LEDs concentrate on R:B ratio differences. My own experiences are that many plants, including cannabis, do well under well under high CRI lights that push the R:B closer to sunlight. I live in a cold climate and start seedlings inside for the garden, they are not all happy all the time under anything less than 3500K 90 CRI. It was much less tricky to avoid problems under fluorescents, but the high CRI stuff really does make healthier looking plants.
 

Rahz

Well-Known Member
I would need to see the SPDs digitized before I believed it but perhaps you are correct. I've broken down SPDs into RGB before but not across brands.

One thing I like to keep in mind is that the yield variation seems to be pretty small for different K samples so variances between one brand and another same K probably doesn't count for much... but that's no reason not to be exact about it. We both seem to be fans of lower K higher CRI so I don't guess we have much to argue about.
 

Abiqua

Well-Known Member
The end result, profiles of plant phytochemicals: is sorely missing. That would help where we want to go....

I find some strains like something closer to sunlight, sometimes. Others don't, sometimes.

I think plants have been shown pretty thoroughly around here, what they like as a combo of absorption irridation etc...the far left and far right of spectrum analysis, may hold the most "feeling" these days...merely a coincidence of all the other quacking going on in todays age :).....or is it...
 

Profound Bastard

Active Member
I find some strains like something closer to sunlight, sometimes. Others don't, sometimes.
I do wonder what decades of breeding under HPS have done, but I think there has to be limits to how far breeding can take us from millions of years under that giant fiery thing in the sky.

When LEDs show nutrient deficiencies I sometimes wonder if HPS nutrient requirements are less because they are not pushing the plants as hard.

Also, why do my outdoor plants never show signs of light stress despite summer sun above and beyond what my indoor plants ever see? More UV protection developing that doesn't trigger indoors? Epigenetics switching on different genes? But hey, LEDs are the future, we might not need blue sky 6500K sunlight 100% of the time, but I bet 4-5000K high CRI would be a nice compromise on averaged daylight, we just don't have lights that can do that yet without way too much green in there.
 

Abiqua

Well-Known Member
I do wonder what decades of breeding under HPS have done, but I think there has to be limits to how far breeding can take us from millions of years under that giant fiery thing in the sky.

When LEDs show nutrient deficiencies I sometimes wonder if HPS nutrient requirements are less because they are not pushing the plants as hard.

Also, why do my outdoor plants never show signs of light stress despite summer sun above and beyond what my indoor plants ever see? More UV protection developing that doesn't trigger indoors? Epigenetics switching on different genes? But hey, LEDs are the future, we might not need blue sky 6500K sunlight 100% of the time, but I bet 4-5000K high CRI would be a nice compromise on averaged daylight, we just don't have lights that can do that yet without way too much green in there.
Epigenetics, how did the pheno appear? yes, we can "guess" a ratio, but lots of answers.....and in short answering some of the questions individually, with a question....what about just breeding correctly to improve growth...What I mean is anywhere in Canada, sans Windsor, probably won't harvest a jungly sativa very well, especially outdoors and late, but oh we still try and try. How many wackadoodle hybrids exist with No thought into how they actually grow and what environment they may be exposed too....then we starting hitting it with light schemes and its double whammy of sorts....:peace:
 

captainmorgan

Well-Known Member
I do wonder what decades of breeding under HPS have done, but I think there has to be limits to how far breeding can take us from millions of years under that giant fiery thing in the sky.

When LEDs show nutrient deficiencies I sometimes wonder if HPS nutrient requirements are less because they are not pushing the plants as hard.

Also, why do my outdoor plants never show signs of light stress despite summer sun above and beyond what my indoor plants ever see? More UV protection developing that doesn't trigger indoors? Epigenetics switching on different genes? But hey, LEDs are the future, we might not need blue sky 6500K sunlight 100% of the time, but I bet 4-5000K high CRI would be a nice compromise on averaged daylight, we just don't have lights that can do that yet without way too much green in there.
You have to take into account sunrise and sunset outdoors with red shifts and lower intensities. Indoors it's blasted with whatever intensity your using the entire time, outdoors you would have to average the intensity over the entire day. And the intensity difference between July and September or Summer and Fall.
 
Last edited:

Nicholas Daniel

New Member
This has had alot of great information and it seems the 3000k at 90CRI would be recomended over the 3000k or 35000k 80 CRI. I happened upon this graph which seems to say that 80 CRI would be better because of higher amounts of reds (if im interperteing it right). What do you think about this graph?
 

Rahz

Well-Known Member
That graph and some testing someone did back in the day is the reason 3000/80 and 3500/80 were the go to spectrums, but phosphor technology has improved since then. The loss for the additional red shift isn't as great as it was. But ultimately it's yield results that matter.... pick two spectrums and use them both and tell us how it works out. Build two lamps that you're happy with on paper and you'll end up keeping them both ;)
 
Last edited:
Top