What's new
  • Seeds Mafia is running a TURBO contest with great prizes! You can check it here.

DIY leds Discussion Thread for all your how tos and doubts and anything related

Is DIY led worth it.

  • No idea never tried and it seems complicated.

    Votes: 3 8.3%
  • No, i tried it and it was just shit/i burnt down my house/im just a negative nelly about it

    Votes: 0 0.0%
  • No, its too expensive nowadays, can find cheaper than diy growlights

    Votes: 3 8.3%
  • No, it takes up too much time and work for the results it gives

    Votes: 3 8.3%
  • Yes! The time and effort it takes is what actually makes it enjoyable

    Votes: 3 8.3%
  • Yes, with my prices considerations and needs its actually cheaper than bought lights

    Votes: 2 5.6%
  • Yes, its actually safer with me doing the work since i know what im doing and can choose parts

    Votes: 2 5.6%
  • Yes, it means i can repair it myself if it breaks

    Votes: 6 16.7%
  • Yes, it means i can get a light that is perfect for my unique space and needs

    Votes: 8 22.2%
  • Yes, cause i cant get the results i want which i cannot find in any light on the market

    Votes: 1 2.8%
  • All of the above yes answers

    Votes: 9 25.0%
  • I dont know but im leaning yes

    Votes: 6 16.7%
  • I dont know but im leaning no

    Votes: 2 5.6%

  • Total voters
    36

Aristoned

Active member
@Rocket Soul @Neferhotep @Ttystikk

If what I am reading here is correct, the most beneficial ratio of 700nm to 660nm is 1:2. That does not necessarily mean for each 700nm diode we would require two 660nm diodes. We would have to see where we would operate the diodes and how many photons each are producing. Then, from there were can make our estimations based on calculations.

Therefore, if we were to find diodes in the 650nm & 670nm wavelengths specifically alongside 700nm diodes spread uniformly based on micro-moles of photons then we should see an astounding difference in our flowering cycles.

The incandescent fixtures may actually arrive today. I’m heading to the hardware store to purchase some supplies to finish setting up another plug-and-play driver. I swear, these XLG’s are amazing. They may not be as efficient as the HLG’s but I can swap them around like legos and I absolutely enjoy playing with toys that can kill me.

Cheers!

Oh, Digikey is also sending me (x3) Bridgelux Vero Gen 8 Vero 29C’s in 2700k and 97 CRI.

🤯

They should be here just before the first week of the weight cycle. This should get interesting with the tungsten filaments!

The 150W vegetation luminaire is also being tested currently. This thing is running 75W @ 24” and I’m seeing north of 400 ppfd.

🤣
 

Attachments

  • 29DCA6B4-271A-42B4-B324-53B74AEE1682.jpeg
    29DCA6B4-271A-42B4-B324-53B74AEE1682.jpeg
    2.6 MB · Views: 9
Last edited:

Rocket Soul

Well-known member
D
From what i remember reading about the emmerson effect (though it wasnt the actual papers, it was summed up in some horti article, and some from other places):
- the effect happens when you combine over and under 700nm red light. The effect is strongest when these wavelengths are around 50nm from eachother, the strongest "couple being 680 and 730nm.
But...
Its mostly a low light phenomena, the effect is almost negligible above 150ppfd. As such im not sure its the best avenue to get best yield even though far red in itself can have a positive effect on flowering response and yield. Ive seen some indications on far red being very productive when combine with strong red and blue, one of these vids from Bugbees coworkers but i wouldnt swear by it.
Theres another way of thinking about it; that emmerson activation could lead to triggering interesting genes for flower quality. Ive tested 680+730 and got very nice "volatile" terps, some called it sulfur compounds im not sure. But it was really a clean test cause i also added a munch of other reds and uva, and sadly im not attached to the grow where i did this anymore so its hard to know if it was the nm or wattage; we added a total of 120w of targeted reds and uvs per tray which is likely a bit too much.

On 660+700 emmerson effect: i dont know what nm he tested in your paper or if it was early or "final". Another thing that could give good effect using these 2 nm is that they are both inside the par range while 730 is outside.
In any case i encourage anyone to go do their own practical research (though this phrase seems to have been patented by a lot of questionable people, 😂).
 

Aristoned

Active member
From what i remember reading about the emmerson effect (though it wasnt the actual papers, it was summed up in some horti article, and some from other places):
- the effect happens when you combine over and under 700nm red light. The effect is strongest when these wavelengths are around 50nm from eachother, the strongest "couple being 680 and 730nm.
But...
Its mostly a low light phenomena, the effect is almost negligible above 150ppfd. As such im not sure its the best avenue to get best yield even though far red in itself can have a positive effect on flowering response and yield. Ive seen some indications on far red being very productive when combine with strong red and blue, one of these vids from Bugbees coworkers but i wouldnt swear by it.
Theres another way of thinking about it; that emmerson activation could lead to triggering interesting genes for flower quality. Ive tested 680+730 and got very nice "volatile" terps, some called it sulfur compounds im not sure. But it was really a clean test cause i also added a munch of other reds and uva, and sadly im not attached to the grow where i did this anymore so its hard to know if it was the nm or wattage; we added a total of 120w of targeted reds and uvs per tray which is likely a bit too much.

On 660+700 emmerson effect: i dont know what nm he tested in your paper or if it was early or "final". Another thing that could give good effect using these 2 nm is that they are both inside the par range while 730 is outside.
In any case i encourage anyone to go do their own practical research (though this phrase seems to have been patented by a lot of questionable people, 😂).

When doing research on the internet I found the wavelength values to vary quite a bit. That’s what pointed me to finding papers submitted for the research, if these graphs are accurate, then it would make sense that a diode “tuned” to 660nm would produce photons on either side of that wavelength. This would make the 660nm diode ideal as we get both side in one package instead of having to use two in the same space.

The only change I could make is from 730nm to 700nm to see. For now, I will focus on what I have available and do side-by-side comparisons with the 2x4.

Also, I’m doing what I can to mimic the sun as I would not doubt these biological machines are fine tuned to the spectrum the sun produces.
 

Attachments

  • 5D50EC8E-6A80-4C22-97CE-12A5CE3F37D2.jpeg
    5D50EC8E-6A80-4C22-97CE-12A5CE3F37D2.jpeg
    163.6 KB · Views: 8

Rocket Soul

Well-known member
D
When doing research on the internet I found the wavelength values to vary quite a bit. That’s what pointed me to finding papers submitted for the research, if these graphs are accurate, then it would make sense that a diode “tuned” to 660nm would produce photons on either side of that wavelength. This would make the 660nm diode ideal as we get both side in one package instead of having to use two in the same space.

The only change I could make is from 730nm to 700nm to see. For now, I will focus on what I have available and do side-by-side comparisons with the 2x4.

Also, I’m doing what I can to mimic the sun as I would not doubt these biological machines are fine tuned to the spectrum the sun produces.
sounds like a good idea. Though one thing to note when going thru research papers: most effective photosynthesis is not exactly the same as most buds harvested. Theres a lot in the morphogenic response: how the plant grows and how it divert energy to different functions. Best of luck with your results and please keep on posting them :)
 

Ttystikk

Well-known member
Veteran
sounds like a good idea. Though one thing to note when going thru research papers: most effective photosynthesis is not exactly the same as most buds harvested. Theres a lot in the morphogenic response: how the plant grows and how it divert energy to different functions. Best of luck with your results and please keep on posting them :)
Ok so what wavelengths results in most buds harvested? This question gets to the heart of my own interest since I've always been results oriented.
 

Rocket Soul

Well-known member
D
Ok so what wavelengths results in most buds harvested? This question gets to the heart of my own interest since I've always been results oriented.
My best guess is that you want 630-640 to be higher than 660.
Edit: and possibly replace 660 with 680, if you have a decent supplier. But all this is debatable of course 🤷‍♂️
 

Ttystikk

Well-known member
Veteran
My best guess is that you want 630-640 to be higher than 660.
Edit: and possibly replace 660 with 680, if you have a decent supplier. But all this is debatable of course 🤷‍♂️
Another thing that tends to get lost in these discussions of the minutiae of spectrum timing; exactly how much difference does it make?

It's important to not only work out what spectrum is best along with optimal combinations but also to quantify the magnitude of gains.

For instance, if we're throwing all this effort into a gain of 1% then I'm not sure it's worthwhile. I think a reasonable threshold might be closer to 5%.

Finally, are we SURE we're holding some other *quantifiable* variable constant, such as wattage consumed vs weight/resin produced?

I see a lot of people just hitting the plants with more light and getting good gains. How much does spectrum really matter in the final analysis?
 

Rocket Soul

Well-known member
D
Another thing that tends to get lost in these discussions of the minutiae of spectrum timing; exactly how much difference does it make?

It's important to not only work out what spectrum is best along with optimal combinations but also to quantify the magnitude of gains.

For instance, if we're throwing all this effort into a gain of 1% then I'm not sure it's worthwhile. I think a reasonable threshold might be closer to 5%.

Finally, are we SURE we're holding some other *quantifiable* variable constant, such as wattage consumed vs weight/resin produced?

I see a lot of people just hitting the plants with more light and getting good gains. How much does spectrum really matter in the final analysis?
Screenshot_2025-03-16-21-29-21-72_e2d5b3f32b79de1d45acd1fad96fbb0f.jpg

Heres some quantified tests from the study i posted a few pages back. Each spectrum grew plants with the same ppfd at tops (wattage may have varied somewhat since the first two, 2peaks/narrow, are heavily red supplemented around 50% which alwasy means better ppf/w).
Depending if youre looking at low or high ppfd and what baseline you use its somewhere around 5 to almost 20%. Winner seems to be 2 peaks, followed by broad which has a little more fat tail on the far red side.
 

Aristoned

Active member
The UVA and IR bulbs are in right now. Next weekend I will switch to the 2200K/100CRI and put the UVA away until the last two weeks.

The IR is setup to wake the plants and put them to sleep, it also will run during the day to warm the tent as needed. The UVA will come on with the LED light-cycle.

Cheers!
 

Attachments

  • 949C4735-05E3-47DC-B352-2634920BA6A2.jpeg
    949C4735-05E3-47DC-B352-2634920BA6A2.jpeg
    3.2 MB · Views: 7
  • E95C6DB3-D130-416E-952B-DC3F48A326B1.jpeg
    E95C6DB3-D130-416E-952B-DC3F48A326B1.jpeg
    2.5 MB · Views: 6
  • 5C554B9A-B874-4C95-9E3D-E7762A59603D.jpeg
    5C554B9A-B874-4C95-9E3D-E7762A59603D.jpeg
    2.4 MB · Views: 7
  • 51AFED0F-6B41-4741-A3A0-AE4E9FD35F72.jpeg
    51AFED0F-6B41-4741-A3A0-AE4E9FD35F72.jpeg
    2.6 MB · Views: 7
Last edited:

Aristoned

Active member
This is the first few hours with the tungsten filaments, the IR is 🤩

Everything is standing tall and reaching for the sky.

My initial observation leads me to believe the IR being produced is directly warming the leaves in a way that radiant heat/convection cannot. From here on, my LED’s will have tungsten as supplimental lighting where applicable.

This is just crazy. I cannot wait to see the effects from the wake/sleep cycle. The temperatures are also VERY stable now. Between 84.1° and 85.1° F.

Ya know what’s even crazier? Gen 8 Vero’s in 2700K and 97 CRI, three of them. Going to see how well that works on this Kush.
 

Attachments

  • 353A80BB-4182-4A06-8255-4A981ACCAE5E.jpeg
    353A80BB-4182-4A06-8255-4A981ACCAE5E.jpeg
    4.2 MB · Views: 6
Last edited:

jonesfam7715

Well-known member
This is the first few hours with the tungsten filaments, the IR is 🤩

Everything is standing tall and reaching for the sky.

My initial observation leads me to believe the IR being produced is directly warming the leaves in a way that radiant heat/convection cannot. From here on, my LED’s will have tungsten as supplimental lighting where applicable.

This is just crazy. I cannot wait to see the effects from the wake/sleep cycle. The temperatures are also VERY stable now. Between 84.1° and 85.1° F.

Ya know what’s even crazier? Gen 8 Vero’s in 2700K and 97 CRI, three of them. Going to see how well that works on this Kush.
I've ran some of the cree cmu2287 cobs in the 3000k & 4000k 95+ cri z models. I like the high cri cobs. Mine have ran for years at this point never running lower than 80 watts and still work, but I am now replacing them with 90 cri replacements because $$, can't find a good deal on higher.

I would like to do a side by side comparison of my own, samsung lm301b strips 3000k 80cri + 660nm & 730nm reds vs osram 4000k 90cri strips with a wider red 660nm 730nm 850nm & 930nm and a few 395nm uva & 460nm to keep stretch down. Basically a generic vs a more complex spectrum, and focus more on quality/trichome development under microscope and local testing all the extra plant material is useless if the quality is trash, we need more well documented side by side tests, I recently lost access to a quality meter tho kinda putting a pinch on plans.
 

Drop That Sound

Well-known member
Veteran
The UVA and IR bulbs are in right now. Next weekend I will switch to the 2200K/100CRI and put the UVA away until the last two weeks.

The IR is setup to wake the plants and put them to sleep, it also will run during the day to warm the tent as needed. The UVA will come on with the LED light-cycle.

Cheers!
Now alls ya needs is a big mean lizard in there, to bask with the plants and keep guard.. Name him budzilla!

🦎
 

Drop That Sound

Well-known member
Veteran
I still haven't made up my mind yet on the supplemental low wattage halogen bulb array for my latest LED fixture...


1. Make the easiest circuit ever and use cheap 120v 25w G8 or G9 bulbs and matching ceramic sockets. Just deal with the extra AC line flicker (roughly a 10% loss of light that you don't really notice, but the plants def lose out on!), and use a cheap dimmer. Be able to use mini reptile 25w red heat and blue "uv" bulbs too!?

1742183075222.png
1742183190801.png
1742183770085.png


Or..
2. Go with lower voltage longer lasting 12 or 24v bulbs ( I like the beefier transversely mounted tungsten filament\coils), but then have to have to use a step down transformer or other AC electronic ballast along with a more sophisticated dimmer switch. OR, go ahead drive them babies just like LEDs with a nice clean common 12 or 24v DC power supply?


Hmm. What to do..
 

Ca++

Well-known member
I would look at R50 or R80 reflector lamps. Not just halogen. The standard incandescent can extend the LED spectrum smoothly. The halogen takes some of that non-visible radiation away, to make visible light with it. Giving an irregular spectrum.

I used halogen, and was happy, and I think D9 also did. However, I don't think it's the best idea, and a halogen reflector lamp tends to be dichroic. Dichroic mirrors (which I'm almost certainly spelling wrong) have a filtering effect, tuned to not reflect the heat radiation. I see you are not looking at reflector lamps, but are interested in efficiency. So you will end up needing something. Costs soon roll out of proportion, which is why many track and bedside lights use the R50/80/120 reflector lamps.
3000k+gls.jpg
iirc the smooth line is an incandescent, and the lock ness line the 3000K LED. The red line I plotted out, to show the combination. I obviously had some wattage in mind, but it's ages ago
edit: That might be the worst LED ever, if the colours are to be believed. I will leave the pic up though, as it kind of shows how the incandescent isn't a total waste of power, because while they make bad lights in the visible spectrum, the out of band is actually of interest




Edit2
1280px-Halogen_spectrum.svg.png

By Mark A. Peterson, professor of physics, Mt. Holyoke University
This is a halogen spectrum, but not 'the' halogen spectrum. Halogen isn't a specific gas, but rather a group of gasses, that are often mixed in these lamps. This one looks alright tbh. Nice 730 and 840 if your trying to send messages to the plant. However the incandescent might be better for smoke signals.
1920px-2200K_Spectrum.jpg
 
Last edited:

Latest posts

Latest posts

Top