Rider's first foray into LED strip lighting: Samsung H Series GEN3

[Rider509]

Thanks

I can't find any info on it, certainly something that can affect my final build decision, for now I have 10 of the 1' Q-series strips on the way for a little test build

Yes! I was hoping someone would give those a test run. Which driver will you use?

 

[PurpleGunRack]

Yes! I was hoping someone would give those a test run. Which driver will you use?

I'm getting some help from the guys over at LEDGardener, but I was thinking about having 9x strips in series on a HLG-60H-C700A HLG-60H-C700A MEAN WELL - Pwr sup.unit: switched-mode | TME - Electronic components

The build will be 3x 1000mm U channel with 3 strips on each channel, for a total build it would take about 20 of these 3 strip U channels.
I was thinking about getting the 1120mm strips. but in a 1200mmx1200mm tent with negative pressure it would be a pretty tight fit..

I have a mate who works in a metal factory, so I might persuade him to make some heatsinks with pre-drilled holes so the strips can be easily screwed onto the heatsink where the little rounded indents are, and thus not have to mess with thermal tape/paste.
I only needed the 9 strips for my build, but 10 strips are cheaper than 9, and it leaves me with one I can send to my mate to work with.

 

[LEDXpert]

@PurpleGunRack How did the it go with the Q series? Q series uses the new LM301B. I've designed LM301B in to a new design here at work but it didn't get picked up because the LM561C is like 25% less cost per LED and the upside is minimal. From a circuit board designer level it buys you more space to lay out a complicated design because they are smaller and has better thermal characteristics but for grow rooms we're not seeing 120-140F where it starts to matter. Downside to them is what they call "LES" Light Emmitting Surface. Which again doesn't matter here since we're not trying to blend even light behind a lens.

Samsung has come out with yet another module of H influx which is designed to replace designs where F Series (LM561C) would go with new LM301B. Might be interesting for you to check out to cut down on the number of strips you have to use. Again don't see a huge value proposition to spend that so much extra money for so little boost especially if you might benefit from some infrared existing from excess heat.

@Rider509 I'm thinking about jumping on and doing my first go here now that I've read up haha. I'll post when I do. I know you used 3500K on your modules, have you tried 3000K and 4000K before? Know anything about over stretching with 3000K? Thoughts on 4000K? I know it has less reds but it also has more total light because it is more efficient so plants may be getting close to the same amount of red just less on a relative basis. 2700K and 90CRI seem pointless to me since efficiencies are in the dumps and all it does it help the balance of colors be close to the sun but in doing so total par being way down.

 

[Rider509]

I use 5000K 80CRI H series to veg under and then flower under the 3500K 80CRI. I've not experimented with a mix of 3000 and 4000. 3500K is a proven winner. I'm also supplementing with UVA at 4% of total light wattage. Can't say total light output because UVA at 400nm or less is below the PAR range.

 

[PurpleGunRack]

It went pretty well, the money spend was well worth the lessons learned by doing the build, and got me a nice little supplemental lamp.
DIY Test Build with Samsung Q-series strips

Just ordered some single row F-series both 2' and 4' for the next adventure

If you have to build with 1' strips the Q-series is a good choice beacause they're about the same price as the H-series which has only 4 diodes more, unless of course you need 160+ strips which is the minimum order for the 1' F-series.

Know anything about over stretching with 3000K?

No worries, 3000K is awesome for veg with tight node distance, enormous leaves early on and no stretch issues
Sorry for the photobombing @Rider509 , but it's easier than words

DSC021226

• PurpleGunRack
• Nov 30, 2017

DSC0212311

• PurpleGunRack
• Nov 30, 2017

DSC021287

• PurpleGunRack
• Nov 30, 2017

DSC021545

• PurpleGunRack
• Dec 7, 2017

DSC021964

• PurpleGunRack
• Dec 15, 2017

DSC022239

• PurpleGunRack
• Dec 20, 2017

DSC022428

• PurpleGunRack
• Jan 5, 2018

DSC022447

• PurpleGunRack
• Jan 5, 2018

DSC023688

• PurpleGunRack
• Jan 18, 2018

DSC024626

• PurpleGunRack
• Feb 4, 2018

DSC02837.JPG

• PurpleGunRack
• Feb 27, 2018

I use 5000K 80CRI H series to veg under and then flower under the 3500K 80CRI. I've not experimented with a mix of 3000 and 4000. 3500K is a proven winner. I'm also supplementing with UVA at 4% of total light wattage. Can't say total light output because UVA at 400nm or less is below the PAR range.

How do you like the UVA, and how/when do you apply it?

I was thinking about ordering some UVA monos with my upcoming 730nm mono order

 

[Rider509]

The UVA comes into play during flowering. There are some killer articles that explain it better than I can, but in a nutshell you are stressing the plants and their response is to produce more trichomes as protection. You only need about 4 or 5% of the wattage of the white light. It's a well documented response, and not hard to implement.

 

[StoneyMemoirs]

AWESOME thread--whew!

So, help me out guys...what would you do? I'm going to be building out a 4x8 tent from scratch...stick with 3500s all around? Go some 3000, some 4000? Trying to keep costs to a minimum of course, so it'd be nice to get the 25+ price break...

Looking at these, at the moment. How many drivers, lights, etc you think I'll need to power up a 4x8 properly?

SI-B8U261560WW Samsung Semiconductor, Inc. | Optoelectronics | DigiKey

 

[Graytail]

For a 4x8, shoot for 1300 watts of Samsungs, about 40W/sqft. Design the layout to emphasize the walls and corners, with the boards or strips snugged up against them. The middle will take care of itself.

Drivers are pretty uniform in their own cost/W, so you could have more small ones or fewer large ones - price per watt will be similar. It'll be more important to match voltages.

I run exclusively 3000K in the bloom room and 4000K in veg. My own opinion is that spectrum is wayyyyy down on the list of important factors. Many other people use a blend. I find the reddish tone from the 3000K boards to be irritating in pics, but it's not noticeable in person. The 4000K spectrum looks white both ways.

And Arrow has better prices. Vendor Cost for Samsung SI-B8V261560WW

 

[StoneyMemoirs]

For a 4x8, shoot for 1300 watts of Samsungs, about 40W/sqft. Design the layout to emphasize the walls and corners, with the boards or strips snugged up against them. The middle will take care of itself.

Drivers are pretty uniform in their own cost/W, so you could have more small ones or fewer large ones - price per watt will be similar. It'll be more important to match voltages.

I run exclusively 3000K in the bloom room and 4000K in veg. My own opinion is that spectrum is wayyyyy down on the list of important factors. Many other people use a blend. I find the reddish tone from the 3000K boards to be irritating in pics, but it's not noticeable in person. The 4000K spectrum looks white both ways.

And Arrow has better prices. Vendor Cost for Samsung SI-B8V261560WW

Thanks Gray--a lot of this is still over my head. Want to make sure I don't mess it up!

 

[LEDXpert]

For a 4x8, shoot for 1300 watts of Samsungs, about 40W/sqft. Design the layout to emphasize the walls and corners, with the boards or strips snugged up against them. The middle will take care of itself.

Drivers are pretty uniform in their own cost/W, so you could have more small ones or fewer large ones - price per watt will be similar. It'll be more important to match voltages.

I run exclusively 3000K in the bloom room and 4000K in veg. My own opinion is that spectrum is wayyyyy down on the list of important factors. Many other people use a blend. I find the reddish tone from the 3000K boards to be irritating in pics, but it's not noticeable in person. The 4000K spectrum looks white both ways.

And Arrow has better prices. Vendor Cost for Samsung SI-B8V261560WW

I'm going to agree and disagree with Graytail. BTW I've done some of my own "testing" since my last post ha and have learned a lot more of the science beyond just LED.

Agree: About 600W in a 4x4 is approximately equivalent to a 1200W in terms of flower weight if you're using efficient chips and not junk. Although more is better. There are diminishing returns on what plant can take up. Make sure you know your par values and see if you need supplemental CO2.

So if you had 600W of LED power (when you calculate volts x amps on your modules) then with 10% inefficiency of your driver you're at about 660W system. Meanwell waterproof metal housed drivers tend to be 88-92% efficient depending on model, load, input voltage, output voltage etc. 90% is good average. Other drivers are typically 82-85% efficient.

Disagree: I wouldn't focus on edges and corners. As the light travels through the air it gets less intense because it is spreading out so you want your light tight and close to your plants. The more point sources you have the less hot it is in a given area so the tighter you can get to your plant. Also less obstruction of light from neighboring plants (better fill). You also want to get tight because a bare led module will be shooting light at near 180 degrees with 80%+ being in 120 degrees. It means you loose a lot of intensity quickly. You can look at beamspread of samsung led on data sheet. Then any bounce off an edge or a corner will suffer significant losses on the reflection point plus same distance issue of the bounce. This will also drive your plant to grow in to the wall. Reflection is good as a pickup of light but should not be relied on as a way to spread light evenly. If you brought a light all the way out to the edge 60 of the 120 degrees would just fire in to the wall. You want that light on a plant instead. I'd make something roughly 40-44" square. Wall will bounce 4-8" on sides and in the middle of each 4x4 where there is no wall the overlapping beams from each light will fill in the dead space.

If you're trying to fill in mid sections of popcorn buds add side lights instead.

Agree: Spectrum of white light isn't super crucial as long as you're 4000K or below. Because white LED light is made with blue light at about 4000K and below there is enough red shift in the phosphor pigment to give you red for flower. 3000K would trigger chlorophyll receptors better but you typically loose efficiency. If you take a look at the spectrum of HPS it is in the orange and green area. Almost zero chlorophyll production wavelengths that fall in blue and red and that is the "standard". Which will soon be replaced by LED, mark my words. I've helped a couple legal home grow people select components and they have done both spectrums with good results.

Agree: Arrow has good stock and better pricing. Company I work for uses Arrow so didn't want to disclose that since everyone was on Digikey but now that its out Arrow is good with free overnight shipping. DigiKey though is first to get new H Influx boards which I would recommend because of efficiency and they moved back to aluminum pcb which will help with thermals a lot.

Agree: Driver pricing doesn't get much better at any point. I'd stay under 100W and look for something called "class 2 power supply". In theory it is safe to touch the positive and minus of the output ternminals together because it is power limited. This will help with risk of fire and shock. (you have a fire source and lots of liquid). It will spread heat round to break up wattages and allow you to replace things. Also if you get a bad supply with say a 600W supply it will suck trying to get that replaced, less of a deal with smaller wattages. It does happen.

 

[StoneyMemoirs]

I'm going to agree and disagree with Graytail. BTW I've done some of my own "testing" since my last post ha and have learned a lot more of the science beyond just LED.

Agree: About 600W in a 4x4 is approximately equivalent to a 1200W in terms of flower weight if you're using efficient chips and not junk. Although more is better. There are diminishing returns on what plant can take up. Make sure you know your par values and see if you need supplemental CO2.

So if you had 600W of LED power (when you calculate volts x amps on your modules) then with 10% inefficiency of your driver you're at about 660W system. Meanwell waterproof metal housed drivers tend to be 88-92% efficient depending on model, load, input voltage, output voltage etc. 90% is good average. Other drivers are typically 82-85% efficient.

Disagree: I wouldn't focus on edges and corners. As the light travels through the air it gets less intense because it is spreading out so you want your light tight and close to your plants. The more point sources you have the less hot it is in a given area so the tighter you can get to your plant. Also less obstruction of light from neighboring plants (better fill). You also want to get tight because a bare led module will be shooting light at near 180 degrees with 80%+ being in 120 degrees. It means you loose a lot of intensity quickly. You can look at beamspread of samsung led on data sheet. Then any bounce off an edge or a corner will suffer significant losses on the reflection point plus same distance issue of the bounce. This will also drive your plant to grow in to the wall. Reflection is good as a pickup of light but should not be relied on as a way to spread light evenly. If you brought a light all the way out to the edge 60 of the 120 degrees would just fire in to the wall. You want that light on a plant instead. I'd make something roughly 40-44" square. Wall will bounce 4-8" on sides and in the middle of each 4x4 where there is no wall the overlapping beams from each light will fill in the dead space.

If you're trying to fill in mid sections of popcorn buds add side lights instead.

Agree: Spectrum of white light isn't super crucial as long as you're 4000K or below. Because white LED light is made with blue light at about 4000K and below there is enough red shift in the phosphor pigment to give you red for flower. 3000K would trigger chlorophyll receptors better but you typically loose efficiency. If you take a look at the spectrum of HPS it is in the orange and green area. Almost zero chlorophyll production wavelengths that fall in blue and red and that is the "standard". Which will soon be replaced by LED, mark my words. I've helped a couple legal home grow people select components and they have done both spectrums with good results.

Agree: Arrow has good stock and better pricing. Company I work for uses Arrow so didn't want to disclose that since everyone was on Digikey but now that its out Arrow is good with free overnight shipping. DigiKey though is first to get new H Influx boards which I would recommend because of efficiency and they moved back to aluminum pcb which will help with thermals a lot.

Agree: Driver pricing doesn't get much better at any point. I'd stay under 100W and look for something called "class 2 power supply". In theory it is safe to touch the positive and minus of the output ternminals together because it is power limited. This will help with risk of fire and shock. (you have a fire source and lots of liquid). It will spread heat round to break up wattages and allow you to replace things. Also if you get a bad supply with say a 600W supply it will suck trying to get that replaced, less of a deal with smaller wattages. It does happen.

Cheers and thanks for the input!

 

[Graytail]

Disagree: I wouldn't focus on edges and corners. As the light travels through the air it gets less intense because it is spreading out so you want your light tight and close to your plants. The more point sources you have the less hot it is in a given area so the tighter you can get to your plant. Also less obstruction of light from neighboring plants (better fill). You also want to get tight because a bare led module will be shooting light at near 180 degrees with 80%+ being in 120 degrees. It means you loose a lot of intensity quickly. You can look at beamspread of samsung led on data sheet. Then any bounce off an edge or a corner will suffer significant losses on the reflection point plus same distance issue of the bounce. This will also drive your plant to grow in to the wall. Reflection is good as a pickup of light but should not be relied on as a way to spread light evenly. If you brought a light all the way out to the edge 60 of the 120 degrees would just fire in to the wall. You want that light on a plant instead. I'd make something roughly 40-44" square. Wall will bounce 4-8" on sides and in the middle of each 4x4 where there is no wall the overlapping beams from each light will fill in the dead space.

If you're trying to fill in mid sections of popcorn buds add side lights instead.

Some of this is misguided, I think. What you don't consider is the Inverse Square Law. I went over it with Rider near the beginning of this thread. The measurements we've seen show that intensity diminishes by roughly 60% at twice the distance when you have thousands of point sources. That fits very well with umol requirement for cannabis, which needs about 500 to grow well and limits at 1000-1500 umols depending on environmental conditions, like CO2 and VPD. 40% of 1200 is 500. So ...

If the top of the canopy is getting 1200 umols, then at twice the distance, there should be 500 umols.

If you have one foot of headroom between the source and the top of the canopy, then at twice that distance = 2 feet, you'll have 500 umols. You have a canopy that is one foot deep. If you have 2 feet of headroom, your canopy is 2 feet deep. Right?

So you want your source higher and farther away, and you want it more spread out.

Secondly, the foliage at the walls and corners is only exposed to 1/2 to 1/4 of the sky. The rest is blocked by the walls. Having light directly above helps a great deal. The only source otherwise, is either at an angle or shaded by other foliage. Every PAR grid I've seen is weak at the corners and walls. I agree entirely with your 4-8" gap at the wall, but I'd be closer to 2-4". Snugging right up to the wall accomplishes nothing and impedes air circulation.

At some point though, I'll add strips along the walls, pointed straight down, about 30 inches below the sky. They have these nifty commercial housings for strips, complete with clear waterproof covers, for jewelry case lighting. That should increase penetration into the lower canopy without adding more at the sky.

 

[PurpleGunRack]

@StoneyMemoirs

To answer your question: You want at least 32 of the 560mm single row F-series strips for 8x4.
Making 8 lamps each with 4 strips on a HLG 120h - C1400 would be a fine way to light that space evenly.



My advice is to not look at watt, but to get at least 800x LM561c diodes per m2 / 70 diodes per sq. ft. and run them at 75% of their rated max when flowering.

3000K all the way, never sell out on flower power to improve veg, a day or two longer veg will make up for any lower vegging qualites of the diodes, if there are any at all, 3000K vegs amazingly well
If you have a separate veg room, then get 4000K for that, 5000K has higher efficacy, but also has a lot less photons in the red wavelengths which are by far the most efficient photosynthesis wise.

 

[Graytail]

Great advice PGR, especially the diodes/sqft and the ma/diode guidelines.

I run mine very soft, but you know the hotter end of things better than I do. Do you know what kind of Lu/W you get in real life at 75%?


[Edit] I just did the diode/sqft calc for my room. 4256 diodes/16.5 sqft = 258.

 

[Derbybud]

75-80% max and they sing.

 

[LEDXpert]

Its more about junction die temperature of the led. I've spoken with Samsung engineers about 3 years ago and they have said they rate their modules so survive their rated temperature ratings and their rated currents with no heatsinks. They've since come out with higher wattage boards so I don't know but you should always heatsink. Hotter the LEDs get the lower the efficiency even at the same current and the longevity decreases which in turns makes them less bright faster over time.

The real way to test is on your modules there is a Tc point. Solder a temperature probe to it and then measure. Then put in your system to Samsung Module Calculator and input Tc. It will tell you good information.

As a rule with the LM561C I don't go above 150ma per string just based on the longevity testing Samsung has done (LM80). Calculator link there will show you how many series and parallel strings there are.

 

[HiddenGrow]

Bless you RIder for this thread. I've learned a lot and plan to get a LED strip setup.
Merry Christmas.
And thank you everyone that asked the hard questions for me

 

[TheMadDabber]

Wish rider would make an appearance again... merry Christmas wherever you're at buddy!

 

[Picklecloud]

Either/or. Depends on the driver you use. They run at 46V typ.

hey don’t mind the quote I just needed to reply to get your attention. first off solid thread, good work it’s awesome. I wanted to build using these strips on a 480h also. I was planning on doing 6 strips per 480 until I saw this thread. How are they holding up heat wise with the 4 Strips per 480?