Your Opinion on These LED Specs Please?

Lights with these specifications will be available in the near future. I know absolutely nothing about the mfg, only numbers.

Based on the data I was sent...

vendor said:

For our 660W Spectrum Tuning LED grow light -
(1) SamSung White LEDs 301B V2(Luminous Efficacy: 223 lm/W, 3500K and 6500K) + SamSung Red LEDs 3535(PPF: 2.32 μmol/s @ 350mA, Efficacy: 2.32 μmol/J)
(2) PPF1500umol/s, efficiency 2.3
(3) 650-660W, 90-277VAC input, 48VDC output, safety voltage, Meanwell power

Click to expand...

What questions would you ask the manufacturer about them? What details do you look for in a light for cannabis?

Douglas.Curtis said:

Lights with these specifications will be available in the near future. I know absolutely nothing about the mfg, only numbers.

Based on the data I was sent...


What questions would you ask the manufacturer about them? What details do you look for in a light for cannabis?

Click to expand...

lm301b its actually one of best smd on market

The specs are a bit off,
LM301B's produce 220lm/W only when driven very soft, at borderline usable intensity.
Realistically if I imagine these are basically QBs, they will have around 160-170lm/W or even a bit less.
Another thing, Samsung 3535 leds are white leds, the reds have another code, so can't say which leds the sale man is talking about.
It would be important to see(know) the layout of the leds, are they on a custom made board or driven separately, maybe they have a dimming option or anything of the sort to let you dial the spectrum, if wanted.
Price? Honestly for a well built led light with custom options, it can't go over 1200$ tops. I suspect they'll want more, but who knows.


Cheers

warpese said:

lm301b its actually one of best smd on market

Click to expand...

This is definitely good to know. Thank you.

Koondense said:

The specs are a bit off,
LM301B's produce 220lm/W only when driven very soft, at borderline usable intensity.
Realistically if I imagine these are basically QBs, they will have around 160-170lm/W or even a bit less.

Click to expand...

Interesting. This, I imagine, is strictly due to heat reducing efficiency at higher power levels, yes?

Koondense said:

Another thing, Samsung 3535 leds are white leds, the reds have another code, so can't say which leds the sale man is talking about.
It would be important to see(know) the layout of the leds, are they on a custom made board or driven separately, maybe they have a dimming option or anything of the sort to let you dial the spectrum, if wanted.

Click to expand...

Great question! I was asked today what spectrum I would want in a sample unit. So this is a really good question all of a sudden.

Thanks for your plant to test our lamp. We use full spectrum, mixing spectrum or your customized solution for your sample.

Click to expand...

They want cannabis specific.

I'm completely unprepared to answer this question. What do the top of the line units use? Spectrum? Configuration? Any recommendations on where I can look this up myself in the next few days? I'd rather get as close to ideal as possible.

Where's a good place to upload pdf's and link them? I have some spec docs they sent. The 800w (actual watt) specs say 10 bars, so 80w each bar? It looks like the fluence spyder and other DIY, strip led light arrays. I'm trying to find a photo or led distance in the specs.

Dimming goes from 800w down to 740. This does not seem like a lot to me, but then this is all new to me. Is this a little or a lot?

Koondense said:

Price? Honestly for a well built led light with custom options, it can't go over 1200$ tops. I suspect they'll want more, but who knows.


Cheers

Click to expand...

Absolutely zero clue. I'm in a few cannabis groups on facebook to help others with growing, and one of them is LED lighting based. The mfg contacted me and I responded. As far as I can tell, they want real world testing with their lights and cannabis. I'm game.

I honestly wanted to see what they had to offer, since the LED game is heating up. I'm interested in seeing where this goes.

Here's some more they sent, and after this I only have the pdfs.

Features
◆ Tuning Spectrum - Allows digital tuning of spectrum to match any plant type at any stage of growth, from veg to bloom, or to the Sun’s circle, with Spectrum Irradiance Colorimeter for precise control over PPFD, and offers flexibility to instantly accommodate a new crop with varying light requirements.
◆ Tuning and Dimming - Tunable spectra with one gateway by WIFI or triac. For convenience of cultivators' use, each light designed with digital WIFI(overall tuning) AND a triac(single tuning). For overall tuning, 65535PCS lights to be tunable, to the maximum.
◆ 48VDC output -Safety protection for water spraying and outdoor lighting
◆ WINGS series features a built-in power supply, or a detachable power supply that can be mounted remotely on the side of a rack or on an adjacent wall to remove heat from your grow space.
◆ Quick and Easy Installation - Each lighting bar is updated with a push-lock connector, snapped together with the brackets.
◆ Best Replacement of HPS and HID – Low working temperature, 55.8℃; Large Coverage, 4’x4’, 6’x 6’ area at only 6-12” above canopy; Illuminance uniformity, >0.836; Full spectrum or mixing spectrum (R, G, B, IR, UR) for varied plants

Click to expand...

So again, does anyone know anything close to ideal for flowering cannabis with these chips? Placement and spectrum? I'm open to all suggestions, especially when you include technical details. I'm not slow, I'm simply having difficulty absorbing the enormous amount of info on LEDs out there. lol

Any links to additional study material greatly appreciated. Thank you so much for the questions so far.

It all seems about right, but still a little hazy.

Is the 301b available in various white spectrum's? A lot of the brightest leds have a bad cri, using orange not red.

The reds are hopefully the newest ones, that I have been thinking about myself. Much the same project actually. Using good whites and extra red.

Dimming from 800 to 760? That should be clarified.

Plus what's already been said. Particularly about the number used, and drive current expected. If they actually have a product themselves. Not just the idea that you will design it.


Manufacturers should really move away from the cumbersome meanwell's towards smart chips. You can't control what an end user will do with your kit. Lots of led lights fail with burnt out leds. Having the lights dim to avoid cooking themselves is a huge advantage. The smart chips take zero space. They can be parallel connected(and usually are) but use them for groups of leds. Literally slap down a smart chip and circle it with ~10w of leds. It will look after them, and if an led goes pop in a couple of years, you only loose 10w.

What we have now is stuff made in sheds. Just off the shelf bits screwed together. The height of technology is etching a circuit board and sticking leds on it. The parts choice is great, with metal foundries knocking out some decent cooling. But at our end, it's pretty sucky.

I decided to go off and design it. I failed, so don't get excited lol

This is a typical chip https://www.bpsemi.com/uploads/file/20161215114728_476.pdf
It's a line voltage current regulator. You need a string of leds that equal mains voltage. Little groups won't work.

Lets say you had the 10w 12v 9led cobs. 10 of them gives you a 120v requirement. 120w. The chips will manage about 5w each, so you need about 20 in parallel. The thermal limitation is 130c. They will shut off as much as possible to not pass that temperature.

130c sounds hot. The chips are designed with 40mA in mind, but it's all about heatsinking. The boards we see run them harder. Keeping them hotter. There is a balance where you can keep the chips hotter than the leds, so while the chip is regulating 130c, the leds are cooler.

Having line voltage at the leds means having a cover over them. It's electrical safety. 10% of the light is lost. Unless using isolation transformers, but the bulk and expense is back.

I'm shocked how much the latest led's cost. Just to get 10% more than the last generation. You wouldn't pay it, then put a cover on. Though you could have a removable cover, that you know full well people will remove. It seems many lights do have a cover, and good leds, but It's not a route I'm following.


The 301b is classed as a 3v led so something like 40 of them for a 120v supply. A couple a these 40ma chip turned down to allow 65ma and your at about 200lm per watt. It's only 6w total though, per small group like I was talking in my last post. If 40 leds for just 6w is a small group. That's about $20 of leds. $200 for 60w, $2000 for 600w. Just leds. So I can see why they drive them 3 times harder and get ~170lm per watt instead. $650 of leds for 600w.


I'm glad I looked at this, even if it is shooting myself in the foot. It helps me see that 3 72v cobs such as the citizen 1212 90cri, will run off a 150w china floodlight board in 230v regions. Them chips like 50w each, but go over 100w each if asked. So some resistor swaps in the board, and you can get the chips hot enough to match nicely. Or just run it.. As I'm not geared up towards surface mount work.

Edit: I guess you must need a string of about 40 3v leds for a 120v line. But it can be 20w 170lm/watt. The chips are aimed at 220v 40ma use, but maybe at 120v you could get 100ma from one (a look at designs on ebay would tell you) So just 2 or 3 chips could be circled with the 40 leds as a 20w light. Then repeat 20 times for 400w light, that might cover a meter. It's not terrible..

Who is making this? HLG?

That's a lot of info to process, thank you f-e. lol

So I want to know:
The CRI of the chips being used
The type of red chip used
The layout pattern of the chips
The dimmable wattage range
The expected drive current (I think this is in the pdf spec sheets)

So what mix of white/red do I really want? Would two dimmers be useful for separate spectrum chips, groups of chips?

P-NUT said:

Who is making this? HLG?

Click to expand...

The spec sheets say OriGulm and it looks like they already make these units. They're asking me what custom spectrum I want for cannabis. I've never heard of them before reading the spec sheets. lol

they ask if you want them 3000k 3500k 4000k 6000k

Douglas.Curtis said:

That's a lot of info to process, thank you f-e. lol

So I want to know:
The CRI of the chips being used
The type of red chip used
The layout pattern of the chips
The dimmable wattage range
The expected drive current (I think this is in the pdf spec sheets)

So what mix of white/red do I really want? Would two dimmers be useful for separate spectrum chips, groups of chips?


The spec sheets say OriGulm and it look like they already make these units. They're asking me what custom spectrum I want for cannabis. I've never heard of them before reading the spec sheets. lol

Click to expand...

The cri would be nice, but it's really just an indicator of the red wavelength used. To really nail in down, you want to know the exact parts they're using. So we can look at the spectrum they give out. No guessing when it's this kind of outlay.

The layout, or rather the count, will allow us to see just how hard they are driven. Then we will know the efficiency at full power. Most manufacturers can say their light is dimmable down to whatever, then give a maximum efficiency value, not telling you it's measured at that dimmed down power level. Sneaky..


The brand just took me to a FB login page.

Specifying a spectrum is a bit challenging. All we really have is that par curve that governs all plants. Nothing specific to ours. Not that I know of anyway. Perhaps we should get some petri dishes and grow out some tissue samples, using a low level of full spectrum light, and a big spike of different colours over each dish. Weighing the resultant green matter grown at one month, then at 2 months, to see how it multiplies over time. Then we have our own curve. It does seem a lot of effort to buy the right led's though

I reckon we just want a shed load of red and blue, with some high cri white to plug any gaps. It's little more than guesswork though.

It's funny they ask you about the spectrum, if they were serious, they would tell you not ask. Seems like a zero research/development company.

The easiest and most secure way to go is to copy the spectrum from Fluence, add dimmers and timers for far red(if present).

But here we're doing R&D instead of them, so I'll just stop talking

Cheers

warpese said:

they ask if you want them 3000k 3500k 4000k 6000k

Click to expand...

Ok, that makes sense. My first thoughts are a mix of 4000k and 6000k chips, with some red in there. If you look up "6000k led spectrum chart" and switch out the kelvin number, I've found you can get pretty decent looking representations of spectrum output.

f-e said:

The cri would be nice, but it's really just an indicator of the red wavelength used. To really nail in down, you want to know the exact parts they're using. So we can look at the spectrum they give out. No guessing when it's this kind of outlay.

The layout, or rather the count, will allow us to see just how hard they are driven. Then we will know the efficiency at full power. Most manufacturers can say their light is dimmable down to whatever, then give a maximum efficiency value, not telling you it's measured at that dimmed down power level. Sneaky..

Click to expand...

Got it. Exact chips available and the layout pattern.

f-e said:

The brand just took me to a FB login page.

Click to expand...

This does not surprise me. heheh Never heard of them before either.

f-e said:

Specifying a spectrum is a bit challenging. All we really have is that par curve that governs all plants. Nothing specific to ours. Not that I know of anyway. Perhaps we should get some petri dishes and grow out some tissue samples, using a low level of full spectrum light, and a big spike of different colours over each dish. Weighing the resultant green matter grown at one month, then at 2 months, to see how it multiplies over time. Then we have our own curve. It does seem a lot of effort to buy the right led's though

Click to expand...

Interesting. Well, there's got to be this amount of effort going on somewhere. I know the top mfgs working with cannabis are probably doing it. Gotta wait for the trickle down, or figure something out in the meantime. I'm extremely interested in working with fully discrete led panels, with complete control over each wavelength by the nanometer. The down side is the experiments I would want to run will take soooo many lights, or an amazingly long time of repeat tests. lol

f-e said:

I reckon we just want a shed load of red and blue, with some high cri white to plug any gaps. It's little more than guesswork though.

Click to expand...

Argh!! lol

Perhaps a mix of blues, and a dimmable set/mix of reds? Which chips in the Samsung lineup would you want to use for the reds?

Very thought provoking information, thank you!

Koondense said:

It's funny they ask you about the spectrum, if they were serious, they would tell you not ask. Seems like a zero research/development company.

Click to expand...

My first guess is zero into cannabis for sure. You could very well be correct about the rest as well. lol

Koondense said:

The easiest and most secure way to go is to copy the spectrum from Fluence, add dimmers and timers for far red(if present).

But here we're doing R&D instead of them, so I'll just stop talking

Cheers

Click to expand...

I was just thinking about far-red a minute ago and wanting to limit it. Dimmers and timers are a good solution, thank you.

Copying the spectrum from one of the best lights was also one of my thoughts. lol I'm looking for the best combination for me, for cannabis. Them benefiting is ok with me. Heh I figure to share as much information as possible, because it's the best way to get what we want the quickest.

Thank you, good stuff!

Pic I found of the spectrum from the Fluence SpyderX
https://2xuwao2gok1v2wn2em9n5ys8-wp...oads/2018/11/SPYDR-2p-spec-sheet-181119-1.pdf

385nm is very expensive if done with LED. 385nm is also damaging to skin, eyes, equipment, and even plants at higher levels. Not going to go this route just yet.

Here's the link to the LM301B chip specifications.
As long as they're using the 90cri chips, my thoughts are to combine the 5700k and and 6000k chips 50/50 for the blue range, with the 2700K chips for the red.

Approximately a 2:1 blue:red ratio, with "dialing more red" increasing the power output of the red, while dimming the blue just a bit. Hoping their layout pattern doesn't mean they're already driving these chips at the max.

Thanks for all the responses, I've sent my request and questions.
We'll see what happens.

Here's the proposed layout at this point. I've been given advice to drop the 6500k chips and go with all 3500K for the white. I'm also looking at the 660nm wattage being too low. Any thoughts?

IP65
168pcs 3500K, 72pcs 6500K, proportionate layout about 2:1
7pcs 3W 660nm red
7pcs 3W 730nm red
the two red in interleaving layout

And a (mock up) graphic of the 3-route mixing spectra

Cree's red LED is more efficient that that one, was the Samsung significantly cheaper? Cree's photo red XP-E2 comes in 3.2 μmol/J and a 3.44 μmol/J bins vs the Samsung's 3.12 μmol/J (all three are specd at 350mA)

If this doesn't need to be in your hands urgently then it might also be worth waiting for Samsung's new LM301H and Cree's new JB5630HE pricing to come out. We have the specs for them already, but currently you can't work out the price/performance trade-offs.

Makes sense to drop the 6500K if you will mostly flower the plants.
But for a fully versatile cannabis fixture there should be 3500K leds, 450nm leds and 660nm leds in a ratio like 60% white, 15% blue and 25%red. Something like that. For flower only a mix of 3500K and 660nm(70%+30%) should do the trick well.

But then again, if I see they start selling a fixture like that, I'm gonna expect some %$$$!


Cheers

popta said:

Cree's red LED is more efficient that that one, was the Samsung significantly cheaper? Cree's photo red XP-E2 comes in 3.2 μmol/J and a 3.44 μmol/J bins vs the Samsung's 3.12 μmol/J (all three are specd at 350mA)

If this doesn't need to be in your hands urgently then it might also be worth waiting for Samsung's new LM301H and Cree's new JB5630HE pricing to come out. We have the specs for them already, but currently you can't work out the price/performance trade-offs.

Click to expand...

I can specify CREE or Samsung for the reds, I'll take a look at both. I don't have a timetable for when they'll build and send it, so the rest I don't know.

What CREE blues would you use, if the white were all 3500k?

Koondense said:

Makes sense to drop the 6500K if you will mostly flower the plants.
But for a fully versatile cannabis fixture there should be 3500K leds, 450nm leds and 660nm leds in a ratio like 60% white, 15% blue and 25%red. Something like that. For flower only a mix of 3500K and 660nm(70%+30%) should do the trick well.

Click to expand...

This is a very valid point. I have other LEDs for veg, the possibility of building this for "Only" flower hadn't crossed my mind. Hrmmm...3500k + Red and Far-Red? I did want blue for initial stretch, then red for bulking, then a bit more blue and less red for finishing. Are there companies making lights specifically for flower only?

I requested royal blue, but I don't see it in the chip list they sent back. heh Now I'm wondering what CREE chips in blue would be best, and if they'll include them.

Koondense said:

But then again, if I see they start selling a fixture like that, I'm gonna expect some %$$$!
Cheers

Click to expand...

Indeed. lol