Off the shelf retail store screw-in LED and CFL bulb comparisons

[Drop That Sound]

IMO: At the end of the day price should be the only thing that matters. If you look at the reviews for any SIL there are people complaining they burned out within 1 month to a year, all exept the ones that werent received free in return for an honest review anyway. If they do last longer as stated then why havent they just stuck with original designs from 5-6+ years ago?. The prices are dropping which leads me to beleive the quality is somewhat too.

Its kinda sketchy when your buying loads of bulbs, like 60+ especially. Kinda like gambling if you ask me, when I could buy years worth of HID bulb replacements for the price of 60 SILs (just paid around $140 for 60 15.5s which seems awesome if they last!). I refuse to pay much more than a dollar per 8-9 watts, knowing they could burn out at any time. And 9w are a buck at the dollar store.

I havent tested yet (maybe one of you guys can?) but im sure that 2 lower wattage bulbs right up against each other would put out almost the same umol reading as 1 higher rated bulb. Leading me to beleive an 800+ watt fixture crammed with 8-16w cooled off bulbs is just as good as a COB fixture that cost way way more. Or any other LED light for that matter.

Heat is the real enemy and you can only pack so many watts into a standard bulb size, and expect it to last as long as they claim too. 14+ watt bulbs get pretty hot to the touch!

Bases cost almost as much if not twice as the bulbs too unless you buy chincy ones from hong kong that will probably melt with anything more than a 14-16 watter., and possibly burn your house down. They use aluminum connections coated to look like brass, I would steer clear unless I was using 8-9 watt sils.

Thats why I planned to experiment and come up with a cheap easy way to cool the bulbs, while also doing away with the added cost of the base sockets too. Even for smaller fixtures, with say 20 bulbs or less. If those 20 bulbs burn out in a year then SILs aren't an effective solution at all the way I see it. Cooling is what will get your moneys worth, and make them last, i feel.

I thought about drilling many holes in the side of bulb housing, to allow forced air to flow through, but pulling the screw end terminal off will open up a bigger hole and act like a chimney. I think the extra work will pay off and make the bulbs last for years longer than not doing it.

I'm basically talking about building a 4+ inch high X whatever dimension box, with all the SiLs contained inside and air flowing through. The main panel underneath with all the bulbs mounted on could be removable or hinged to allow access and replace bulbs, just like any glass hood opens. I even though about re using old hoods crammed with sils, and replacing the glass with the new panel full of holes!

If anyone is interested, i'm gonna start a new thread, for high wattage SIL grows and designs. Ill let ya all know when I do

EDIT: Here's the link https://www.icmag.com/ic/showthread.php?t=358871

Like to see more grower's and DIY setups, with excess of 300-1000+ watt SIL gardens, compare results, and toss around idea's.

 

[noknees]

i've now grown the same exact plant in the exact same soil (batch) three times, under three different lights as part of my experiments.

for the most part, predictable results.

i wouldn't personally stray from 2700/warm in bloom unless i really really really needed stretch control.

YMMV

 

[Terpene]

IMO: At the end of the day price should be the only thing that matters.

I havent tested yet (maybe one of you guys can?) but im sure that 2 lower wattage bulbs right up against each other would put out almost the same umol reading as 1 higher rated bulb. Leading me to beleive an 800+ watt fixture crammed with 8-16w cooled off bulbs is just as good as a COB fixture that cost way way more. Or any other LED light for that matter.

Heat is the real enemy and you can only pack so many watts into a standard bulb size, and expect it to last as long as they claim too. 14+ watt bulbs get pretty hot to the touch

Cost, color temperature and wattage should all play a factor in your build.

Wattage:
Higher wattage units offer better depth of light penetration than smaller units, even when placed more closely together. However, the extra intensity means you need more space off the bulbs so you don’t cook the plants. So for taller cabinets, higher wattage works, for limited vertical space common in micro setups, lower wattage bulbs are better.

Color:
The yellower/redder the bulb (2700k for instance) the more stretch / larger the leaves. The bluer the bulb (5000k) the smaller the stretch and leaves. Bud size is larger with yellow bulbs, but the blue bulbs produce more frost. In tall setups, run all 2700-3000k bulbs. In small setups not using LST / scrog, 4000-6500k bulbs are better. Or run mixed if you’re worried you’ll miss out, not like it’s hard to change them to suit your mood.

Heat:
As someone who’s played with high wattage lights, CFL arrays and now LEDs, I can say complaining about LED heat output per light output is comical. I can cook dinner on HPS, and CFLs produce 2x as much heat per watt in every array I’ve built. I’m cooling 180w (25w per sqft) to a degree over ambient with 2 PC fans. Try that with a 180w of HPS.
On the argument of little bulbs = less heat - not really. The heat is more or less the same as the wattage. 9w units produce less heat than 14w bulbs but 100w of either will produce near as makes no difference to the same heat overall. I’d wager that due to the additional surface area of the 9w bulbs requiring more bulbs they’d impart more heat into the grow but that’s a negligible diffference.

Cost:
Are we really complaining about 10-15k hour life bulbs costing over 1$ a watt? Has anyone run an HPS for 15,000 hours?
At this point I’m waiting for everyone to upgrade to CREE SILs to see if there’s any improvement in flower quality. They aren’t cheap, but what’s another 30$?

 

[Drop That Sound]

For sure Terp, don't forget all you guys and the results play a factor too

Cost: No complaints here, I just picked up 930 actual watts of basically complete (minus the Wiring, and the base sockets that i wont be needing) LEDs for $140. For not much more and some fun time in the shop, with materials already mostly laying around, the 930w of bulbs will be lighting a smaller 18sq ft area with a 6ft ceiling. Another 930 will go next to it in a matching space after testing the first prototype, if it works well. So lets say just over 1800 watts for roughly $400 in supplies and some work. Compare that to 2500-4000 that I could spend on 2 pre made 900+ actual watt fixtures!

Color: planning 50/50 soft white to daylight, wired up with multiple switches to select different ratios ,patterns, and total amount of bulbs lit. i wont be able to unscrew bulbs with my newest design. The bulbs will be glued on, but easily popped off for replacement if/when needed. Thats why im going extra, because SiLs are so cheap now as is, and i wont have to screw just switch.

More cost: If they do burn out in a years time i'm not gonna cry. But if i bought a 900-1000w actual fixture (any current type of LED fixture) which are in the 1000's of dollars range and it burned out before and especially just after the warranty expired i'd be pretty bummed! Id rather make the guys that did spend that cry when i pull more gpw's than then with a homemade lightbulb rig in the same sq ft, and spend that extra money on more of my crazy ideas.

I wouldnt even purchase COBs unless i went with water cooling, and when spending 100's on that, whats 100s more? Cooling those chips is what makes them possible to even use. Very costly IMO when I still think SILs will better accomodate the way I grow, and the space I have. Ive looked into buying just the round/strip Led pcbs and powering with a few drivers, but still cost way more. Im all about DIY, and spending time over money.

Heat: not worried about heat in terms of heat in the grow, but heat affecting the life expectancy of the SILs themselves. Sure a few of them is no big deal, but add 120x15.5w of them at full blast in a small room and im pushin almost the same btu as my 2 1000w hps bulbs, and cooking SILs for dinner, let alone the poor plants lol..but.. If the components are opened up and cooled 200% or better(they really arent at all as is) the stated 15-20,000 life hrs, which I doubt they'll actually last, could possibly go up to 50,000 or more. That sounds nice if you are investing in 100's of SILs.

Air cooling the internal SILs is the missing link the way I see things. Hasnt exactly been done yet on a bigger scale. But will be done easily/practically as I hope to do.

The most powerful diy led fixture, the lowest possible cost, and longest life expectancy that I could possibly build is the goal. With SILs that come off the shelf no less.

 

[Douglas.Curtis]

Cost: No complaints here, I just picked up 930 actual watts of basically complete (minus the Wiring, and the base sockets that i wont be needing) LEDs for $140. For not much more and some fun time in the shop, with materials already mostly laying around

That's the beauty of the SILs, a cheapo base and some wiring with a plug and you're done. Amazing how well they grow plants.

 

[Dr.Young]

If I can just figure out the sockets and wiring on the cheap.... I could rig up some cheap $1 8.5w.....
I'd need to wire up about 30-60 sockets depending on if I used $1 8.5w bulbs or $3 14w bulbs.

 

[noknees]

If I can just figure out the sockets and wiring on the cheap.... I could rig up some cheap $1 8.5w.....
I'd need to wire up about 30-60 sockets depending on if I used $1 8.5w bulbs or $3 14w bulbs.

https://imgur.com/a/KklNG

 

[Drop That Sound]

Here is my current fixture design rough draft, which I should be able to build for much less than the base sockets would cost, while increasing the life of the bulbs. Also keeping most of the heat outside the room all at the same time.

Materials for the panel and box I haven't figured yet, may end up using metal, plexi glass for a see through effect, or just some painted lightweight 1/4 plywood. Or a combo of stuff, weight being the main concern.

Cheap wire nuts, good wire, and standard light switches mounted in the same box to toggle rows\colors of bulbs on and off.

With integrated timers built in I could emulate the suns intensity throughout the day, another cool feature that could be added!

 

[Terpene]

Everyone likes overkill but why are you planning to run 60 bulbs in the same space I run 12-15? Even with 9w bulbs that’s 75w per sqft and we’ve already seen that you hit diminishing returns past 40w per sqft.

That said, cool fixture design. How much ventilation are you planning to run? Multiple 90 degree turns might slow the air down a bit and work your fans harder - have you thought about just pulling the air left to right with two openings?

 

[f-e]

40w X 10.75 = 430w per meter.
430w x 108lpw = 46440 lux
A 600 sodium is double that. Like equatorial sun.

I'm vegging with about 33w per foot, using 150lpw leds, and I could use more.

 

[Terpene]

430w x 108lpw = 46440 lux
A 600 sodium is double that. Like equatorial sun.

I'm vegging with about 33w per foot, using 150lpw leds, and I could use more.

At 46k lux, (43k foot candles) you're as bright as a 600w at optimal distance (10") or 1000w at optimal distance (12.25") from the canopy:

I know of very few people running 600w or 1000w units a foot or less off the canopy. Usually they're spreading the light out a bit, 14-18 inches off the plants. (20k or so)

 

[Drop That Sound]

Everyone likes overkill but why are you planning to run 60 bulbs in the same space I run 12-15? Even with 9w bulbs that’s 75w per sqft and we’ve already seen that you hit diminishing returns past 40w per sqft.

That said, cool fixture design. How much ventilation are you planning to run? Multiple 90 degree turns might slow the air down a bit and work your fans harder - have you thought about just pulling the air left to right with two openings?

930/18 = 51.66667w per sq ft.

Your right its still a little overkill from 40, but then again is my whole flower room. Ill be running sealed with c02 and RDWC hydro so the diminishing return might go up, we'll see. I may very well be underestimating the SILs, and have to raise the light considerably, but thats cool. I already do good with a single 1000hps over the 18sq ft, and want to try and out perform it, or at least get my gpw still.

Ill probably only use 75% of the bulbs max but want the extra ones to have more circuits to play with color combos. Maybe go full blast for an hour a day etc, and start out and end the cycle at 1/3rd power. The wiring to do that will be the hardest part to figure out. Cheap onboard digital timers could make that happen, and something ive always wanted to try out. For example, start the day with daylight, and end with soft white, and mostly soft whites towards the middle of flower, etc.

My 6" inline fan should pull air through 2 of my fixture designs, i may just go from one side to the other. Not dead set yet but wanted to get my idea out there.

I think it could work well with any amount of bulbs. Just the flat panel and holes itself, without cooling could be a great way to make a quick fixture, and save money on sockets.

The great value walmart bulbs are flat on the housings rim, after you remove the globe, so glue ups will be a cinch! The PCB is about 1-5/8" so ill drill 1-3/4" holes and call it good, a nice fit.

 

[noknees]

...

 

[f-e]

At 46k lux, (43k foot candles) you're as bright as a 600w at optimal distance (10") or 1000w at optimal distance (12.25") from the canopy:
View Image

I know of very few people running 600w or 1000w units a foot or less off the canopy. Usually they're spreading the light out a bit, 14-18 inches off the plants. (20k or so)

Ok, I will try that another way, as it's an unfair comparison to look how much one has at source, and the other on a surface. The chart used to do so is possibly computed, based on source and spread over 360, not with a shade. Both variables that effect the result by many fold. Plus it's in foot candles.

Without using any math, a 600 puts 92,000 lumens over a meter. Or it's 92,000 lux. However you want to say it. That is what the rating means. If you take all its light and spread it over a meter, it's 92k

40w of these 14w lamps per foot, when times by the 10.75 foot you get in a meter, sum up to half that. It's 430w of something about 2/3rds as bright.

92,000 divided by 10.75 gives 8500 lumens per foot
I have 33w of 150lpw citizens. 5000 lumens per foot.
40w of 108lpw is reasonably similar at 4350 lumens per foot.

I'm not just talking theory though. I like messing with lighting. I'm on my second air cooled LED install, thinking about water cooled. I'm doing green light experiments. Using Led uplighters. Pushing the frontier like anybody here. But I won't bullshit. I have not flowered with LED alone.

One real edge leds have, is not needing shades. A 600 might paint a meter with 92k, but just try it. It's a certainly that I can't paint a meter with 92k. If I only lost 20% that would be a good result. If the led hid comparison involves a single hid at a fixed point, again the gap closes.

I hope nobody thinks I'm hating in any way. I like to think I'm seeing all the variables, and that they align with the results I see.


Hastings is experimenting with a low maintenance approach. His grow looks the most like one of mine. It actually has a healthy appearance beyond what I would be happy with. I'm impressed. But it's growing a lot of big leaves. I don't know the strain, but I have to presume it wants more light. I think if he stripped out a lot of that upper foliage and stuck in more light, he could fill that 1.5? tent with bud. (PM: Hastings, I 'know' you could. You are one to watch)


In closing (about time hay) A 600s 92K divided down to a foot, then with 20% more lost, is still 50% brighter than 40w of 14w led lamps. It's a gap that I find shows.

 

[Terpene]

I have not flowered with LED alone.

I hope nobody thinks I'm hating in any way.

In closing (about time hay) A 600s 92K divided down to a foot, then with 20% more lost, is still 50% brighter than 40w of 14w led lamps. It's a gap that I find shows.

No one thinks you're hating, I said the same thing before I tested all this ish myself.

I've had the opportunity to visit lots of other very productive gardens and bring my ghetto light meter with me to check on canopy brightness to see what works and what doesn't. 600w gardens are typically between 950-1200umol and 1000w gardens sit between 1100-1400umol on my little green meter.

So I built all of my SIL gardens to match or beat what I saw with the big lights:

Over a broad area, you can get the same yield (typically with about 15-20% improvements in GPW) as the big lights. This plant (alien orangeADE) produced about a half pound from 240w:

Or if you prefer, heres me doing a vertical effectiveness test using landrace lines (mauritius x ethiopian from Ace Seeds) about 7 feet tall. The SILs fall off about 24" into the canopy, but produce very similar buds to anything I would expect the big lights create:

 

[delta9nxs]

Ok, I will try that another way, as it's an unfair comparison to look how much one has at source, and the other on a surface. The chart used to do so is possibly computed, based on source and spread over 360, not with a shade. Both variables that effect the result by many fold. Plus it's in foot candles.

Without using any math, a 600 puts 92,000 lumens over a meter. Or it's 92,000 lux. However you want to say it. That is what the rating means. If you take all its light and spread it over a meter, it's 92k

40w of these 14w lamps per foot, when times by the 10.75 foot you get in a meter, sum up to half that. It's 430w of something about 2/3rds as bright.

92,000 divided by 10.75 gives 8500 lumens per foot
I have 33w of 150lpw citizens. 5000 lumens per foot.
40w of 108lpw is reasonably similar at 4350 lumens per foot.

I'm not just talking theory though. I like messing with lighting. I'm on my second air cooled LED install, thinking about water cooled. I'm doing green light experiments. Using Led uplighters. Pushing the frontier like anybody here. But I won't bullshit. I have not flowered with LED alone.

One real edge leds have, is not needing shades. A 600 might paint a meter with 92k, but just try it. It's a certainly that I can't paint a meter with 92k. If I only lost 20% that would be a good result. If the led hid comparison involves a single hid at a fixed point, again the gap closes.

I hope nobody thinks I'm hating in any way. I like to think I'm seeing all the variables, and that they align with the results I see.


Hastings is experimenting with a low maintenance approach. His grow looks the most like one of mine. It actually has a healthy appearance beyond what I would be happy with. I'm impressed. But it's growing a lot of big leaves. I don't know the strain, but I have to presume it wants more light. I think if he stripped out a lot of that upper foliage and stuck in more light, he could fill that 1.5? tent with bud. (PM: Hastings, I 'know' you could. You are one to watch)


In closing (about time hay) A 600s 92K divided down to a foot, then with 20% more lost, is still 50% brighter than 40w of 14w led lamps. It's a gap that I find shows.

if i may, lumens is a measurement of light at the source only. it has no relation to a measurement of light at the target.

lux, foot candles, and umols are units of measurement at the target.

lux and foot candles are measurements of what humans see and umols is a measurement of plant available radiation.

from my post above you can see i got different readings based on distance.

with a 1000 watt de bulb i got 1200 umols at 30".

but i can get 1200 umols at a little under 4" with these 14 watt sils.

so if you have enough of these little bulbs close enough together at 4" distance from the plant you can equal the output of a 1000 watt single bulb.

if you get uniform readings on every square foot you have enough bulbs.

you can achieve the same illumination with many more bulbs but they will have to be further away.

you do not want to exceed 1500 umols as there is no point in it because 1500 umols will drive the photosynthetic process at the highest metabolic rate possible with cannabis.

the metabolic rate plateaus off between 1500 and 2000 umols and then above that the plant goes into photo-inhibition as it cannot process synthate faster than that.

1200 umols is a good delivery rate for a 12 hour period because driving the plant at 1500 umols you will usually hit a "wall" around 8 hours into the period as the plant has built up synthate faster than it can process it whereas about 1200 umols will allow the plant to continuously process.

i would say that with these 14-15 watt sils about 6" apart on centers 4" above the plants you will produce the most dry flower weight so there is little point using more bulbs than that.

i hope this helps folks make decisions about how much lighting they need.

 

[psyphish]

Or if you prefer, heres me doing a vertical effectiveness test using landrace lines (mauritius x ethiopian from Ace Seeds) about 7 feet tall. The SILs fall off about 24" into the canopy, but produce very similar buds to anything I would expect the big lights create:

View Image

Some genetics just make long colas, wether they're thick and heavy is a different story.

Here's a grow from years ago. The plants look really bad because I had no idea what I was doing, but the colas are long.

 

[Hastings]

Hastings is experimenting with a low maintenance approach. His grow looks the most like one of mine. It actually has a healthy appearance beyond what I would be happy with. I'm impressed. But it's growing a lot of big leaves. I don't know the strain, but I have to presume it wants more light. I think if he stripped out a lot of that upper foliage and stuck in more light, he could fill that 1.5? tent with bud. (PM: Hastings, I 'know' you could. You are one to watch)

Did someone mention me?

The photo kinda lies a bit. I took literally pounds and pounds of fan leaves off, leaving only the upper 2 nodes of real fan leaves per cola. The strain is custom breeder strain "White Widow X Jacky White", and the tent is 5X5X7 footer. 33.6 watts per square foot seems optimal light to me. Nodes were too tight, but that's also strain related.

There is a second tier of buds in there that do get some light, but I plan a staggered harvest. After harvesting all the top main colas, I'll be planting a cover crop on the same day after top dressing with fresh worm poop, then harvesting the rest of the bud over two more weeks as the cover crop grows and gets a nice head start. Part of this is more laziness on my part. I hate having to trim a kilogram (2.2 pounds) worth of dry bud all at once. Plus I have a day job, so that's a weekend thing. And that's what I hope to harvest, about 1K which certainly isn't amazing for 25 square feet.

The buds today in the start of week 7 are very dense, better than my last grow but that could also be strain related, or maybe this soil is getting better and better after each reuse. I'm using roughly the same watts per square foot as my last slightly smaller grow that I pulled 1.2 grams per watt from.

It won't bother me too much if I only get 900 or less grams from this grow because it's all for personal use. But now it's become a challenge and a hobby - just to see how little maintenance and cost can be put into a grow this large for a Canadian staying within his 4 plant limit and still harvest a respectable amount of top shelf bud.

 

[Hastings]

If anyone is interested, i'm gonna start a new thread, for high wattage SIL grows and designs. Ill let ya all know when I do

EDIT: Here's the link https://www.icmag.com/ic/showthread.php?t=358871

Like to see more grower's and DIY setups, with excess of 300-1000+ watt SIL gardens, compare results, and toss around idea's.

This is a good idea. Although this thread taught me everything I know so far, it is listed under "micro grows" and people who don't grow in small cabinets likely don't even know this exists. More visibility and wider input among growers could prove interesting.
I'm not sure why this thread hasn't been made a sticky in all these years, but it remains active because we have no other place to talk about these "secrets".

It's really hilarious on Reddit. I keep posting cola photos and saying this was grown under regular everyday household LED light bulbs. And 100% of the time, the response is "BS! No way! Are you stupid?". I get down voted like crazy. Then I post a pic showing the lights in the photo with the plants, and it's like "OMG WTF!?! I couldn't believe it, but you're right! I'm really sorry for not believing you." Then all the people who down voted me, up-vote. Hahaha!

 

[noknees]

Did someone mention me?

The photo kinda lies a bit. I took literally pounds and pounds of fan leaves off, leaving only the upper 2 nodes of real fan leaves per cola. The strain is custom breeder strain "White Widow X Jacky White", and the tent is 5X5X7 footer. 33.6 watts per square foot seems optimal light to me. Nodes were too tight, but that's also strain related.

There is a second tier of buds in there that do get some light, but I plan a staggered harvest. After harvesting all the top main colas, I'll be planting a cover crop on the same day after top dressing with fresh worm poop, then harvesting the rest of the bud over two more weeks as the cover crop grows and gets a nice head start. Part of this is more laziness on my part. I hate having to trim a kilogram (2.2 pounds) worth of dry bud all at once. Plus I have a day job, so that's a weekend thing. And that's what I hope to harvest, about 1K which certainly isn't amazing for 25 square feet.

The buds today in the start of week 7 are very dense, better than my last grow but that could also be strain related, or maybe this soil is getting better and better after each reuse. I'm using roughly the same watts per square foot as my last slightly smaller grow that I pulled 1.2 grams per watt from.

It won't bother me too much if I only get 900 or less grams from this grow because it's all for personal use. But now it's become a challenge and a hobby - just to see how little maintenance and cost can be put into a grow this large for a Canadian staying within his 4 plant limit and still harvest a respectable amount of top shelf bud.

stay warm dude. although with that much flower, i doubt you're lonely long

stick around and continue to slay