Is it a good idea to DIY my own HLG 600 or HLG 550?
- Thread starter kdt15
- Start date
yankeeny20
Active Member
Would one of those 260w be good for a 2'x4'? I had a viparspectra but switched to a 600w hps. It works for now bc it's winter but once spring/summer come around I'm going to have heat issues. I was originally checking out the Vero29 DIY kit from Rapid LED but saw the ones you were talking about and they seem better for the price.
coreywebster
Well-Known Member
Yeah they cover a 2x4 perfectly. I'm sure the vero kit is good, but the HLG 260w kit takes 5 min to assemble. Its almost too easy.
WeedSexWeightsShakes
Well-Known Member
I have a 320w kit in my 4x2.5
I’m sure a 260watt would be enough tho
I’m sure a 260watt would be enough tho
coreywebster
Well-Known Member
Yeah its enough, I have two covering 4x4 and pull solid buds and good weight. Not saying more wouldn't do better but the 260watters defo don't fall short by any means.
yankeeny20
Active Member
Any Uncle Sam tax sales for the smaller 260w/320w?
yankeeny20
Active Member
Would the 260w XL kit be better for a 2x4? It looks like it has room to add an extra panel later on if needed.
https://growerslights.com/collections/horticulture-lighting-group-quantum-board-led-kits/products/horticulture-lighting-group-260-watt-xl-quantum-board-led-kit
https://growerslights.com/collections/horticulture-lighting-group-quantum-board-led-kits/products/horticulture-lighting-group-260-watt-xl-quantum-board-led-kit
GrowLightResearch
Well-Known Member
Absolutely. I found the Bridgelux Gen 2 EB strips run cool. No heatsink required.
Just screw down the ends of the strips to an aluminum angle bracket frame and add something to support the center so they don't bow. I've the 22" come out of the box already bowed.
The HLG-550 has 1152 LEDs. The S6 bin are 37 lumens typical.
The Bridgelux Gen 2 EB 175 lm/W (3000K-3500K) 180 lm/W (4000K-5700K), up 4780 - 4930 lm per strip at test. 19.5v, 700 ma Test, 1200 mA max, and 224 LEDs per 44" strip. The Samsung have 8670 lm to 9300 lm @ 1120mA
At 1120mA Bridegelux would be 1.6x the lumens at 700mA or 7648 lm to 7888 lm.
So 1400 Bridegelux LEDs would do better than the 1152 on the HLG-550.
@ 224 per 44" strip 7 strips would do the trick. And I have simulated 7 of those strips and the uniformity was excellent.
@14.00 each x 7 = $98. They can be run at 1400mA. Sounds too good. May need to check my math.
1152 x 37 lm = 42,624 lm
7 strips @1.1A = 7500+ lm = 53,536 (plenty of margin)
Even at 700mA total lm = 4,800 x 7 = 33, 600
Math is close enough.
I have run one of these 22" Gen 2 EB strips at 1500mA (100mA over max) for a couple of days. Did not go much over 50° C.
1.4A x 7 ≈ 10A
Max Vf = 41V.
HLG-480H-42 $140-$150, can push 11.4 A, ≈2 Amps to spare if you wanted to wire in parallel. Not recommended by me.
Best bet electrically is HLG-480H-C1400 but has lethal voltage. $140-$150.
7 x 41v = 287v That can kill in many circumstances.
No heatsinks required. You do not have to put an aluminum bar under each strip.
On mine I used 3/4" PVC water pipe to strengthen the rigidity and prop up the strips.
For the frame two 96" x 1" aluminum angle brackets.
I used the thinner 1/16in thick at $10 per 96" or $20 for the 1/8" thick.
Get a box each of 8-32 nuts and 1" machine screws ($5-$10)
18 gauge wire (16 Amp) would be fine.
16-14 would be better.
Strips $100
Driver $150
Frame $50
You can beat the hell out of of the HLG-550 PPFD wise for ≈ $300.
For another $60 you could mount the strips on a 1" x 44" heatsink. That would boost your output a few percent.
Don't waste money on an a flat aluminum bar to mount the strips to. A flat bar will not act as any type of heatsink. You would only end up with an aluminum bar about the same temperature of the strip with no heatsink.
You can go another $250 for 7 more strips and driver.
At $550 you kick ass and are still way ahead of the game.
I did that quickly and it's possible I made an error. But looks okay.
If you do the parallel wiring, add a 0.01Ω 1% (less than $1 ea.) resistor to each strip so you can measure the current in each strip with a volt meter. Highly recommended.
nfhiggs
Well-Known Member
Good analysis but I think you overlooked a couple of things.
The HLG-550 runs 1152 LM561C diodes at 510W, at that wattage they are running ~150 mA per diode - 20% above the Samsung F series test current (125 mA) and 2.3 times the 65 mA test current that the LM 561C S6 37 lm per diode figure is based on. So the calculated total lumen would be 98k lumens (37 x 2.3 x 1152) rather than 42.6K lm (incidentally, the value they give on the HLG site for the fixture is 78k lumens - typo? I dunno). Anyway, 7 of the Bridgelux strips gets you a little over half way there. Even at 14 strips though, you're still way ahead of the game with only $400 spent, compared to the $1000 pricetag of the HLG-550. You might even be able to do it with an even dozen strips if you run them a bit hotter.
Using the Samsung F series strips - you would have the same diode count as the HLG with 8 of the 44" single row strips - same diode, same quantity. Cost is $212 for 8 of them, compared to $196 for 14 of the Bridgelux strips. One HLG-480 series driver would be enough to power either setup.
One thing that irritates me about the Bridgelux pricing is that they don't give squat for volume pricing - a measly 50 cents (per strip) if you buy 50 or more - barely 3.5% discount. With the Samsung strips you get $3 per strip for 10 or more - buy 100 or more and its 1/3 off, close to $9 per strip. If anyone wants to organize a 100 strip group buy, I'd be on that shit...
GrowLightResearch
Well-Known Member
I looked at this as BX doesn't mark up the single quantity, giving a better deal on single units than Samsung.
Has anyone said you think too much?
That was a thorough analysis. Too much like work.I was being conservative and trying to keep it simple. I'm not sure if it's fair to use typical for one and not the other.
Typical is the easy way which leaves out additional variables that can skew results.
Speaking of curretn adjustment, this is something I have been wondering. Like here the current is 2.3x over typical. Does the buyer just leave it there?
I noticed in the CoB faction, they often leave the current up to the forward voltage of the CoB and the max voltage or the CC range of the driver. I still cannot wrap my head around that approach.
Do most growers adjust the current to an optimal level? Or do they just max it out even if the fixture has a current adjust?
Does most know how to adjust current to an optimal level?
Does most know how to adjust height to an optimal level?
My thinking is you first find the optimal height, then adjust the current.
It seems the biggest impediment to do this correctly is not having a PAR meter. Understandably so.
One reason I ask is I have found a very inexpensive 6 channel sensor ($5) that only needs a serial port to read the data.
It uses modem AT commands to read the irradiance values (µwatt/cm²) for deep blue 450nm, blue 500nm, green 550nm, yellow 570nm, orange 600nm, and red 650nm. You can set gain, integration time, and calibration for each color.
I think this would be the perfect tool to adjust the fixture to the optimal height.
SparksFun sells a Qwiic / Arduino board for $25. https://www.sparkfun.com/products/14351
I am writing an app for Windows to use it for estimating the irradiance at the canopy.
Example of data for the six wavelengths for a 4000K F series strip: 22151, 10459, 13380, 22725, 33261
I normalize the data to the datasheet's SPD. They do not match because the SPD is look at 400-800nm and the sensor is looking at a 40nm bandwidth. I need to convert the radiometric µWatt/cm² to quantum µmols/m²/s.
What do you think?
22151 67 deep blue
10459 31 blue
13380 40 green
22725 68 yellow
33261 100 orange
20851 63 red
The size of the PCB is about 0.5" x 0.75"
.
Attachments
nfhiggs
Well-Known Member
"Typical" or test reference current is not always the same as real world use. I try to use end user values.
Well, this varies by user - if its going to be used full cycle they may dim it during seedling and veg and bring it up to full power for flower. If its going to be dedicated to flower its more of a "set it and forget it" thing for the most part.
Most CoB users try to set an efficiency target based on the chosen CoB's spec sheet, and run them at that current. This is usually 30-50% of the CoB's max current. Running them at full capacity generally results in poor efficiency, so that is avoided.
Can't really speak to what most growers do, TBH. I can only speak for myself - I'm not necessarily seeking optimum efficiency I'm using DIY LEDs mainly because I just like to tinker.
Yeah, that is what's generally recommended, but lots of peeps seem to do it backwards. They select a current level to run them, then adjust height to bring them as close as possible without burning them.
I tend to disagree here - PAR meter is not really needed when you can get to within 10% of a target PAR level using LUX values if you have white light and a known spectrum.
verticalgrow
Well-Known Member
Were do u get strips for $9 or are u talking about 2 foot strips
Digi 4ft price for 100 strips = $17.88/strip
LINK = https://www.digikey.com/product-detail/en/samsung-semiconductor-inc/SI-B8V521B20WW/1510-2225-ND/6624009
Zerodayz420
Active Member
I bought these lights too and they are definitely lower quality than my QB 260 kit, but much cheaper in price. I can't put the driver on top of my fake kit because it gets scolding hot. I don't know if ill ever buy the fake kit again because the quality scares me. I didn't sleep the first night I hooked them up lol. I used the light meter on my phone and at the same height I am getting the same measurements from each board.
The hanging kit that comes with the light is made for a 50 foot high tent and i had to take the boards off and clean the heatsink also. The boards are wired in parallel and i had to buy a potentiometer because its not included in the kit.
Attachments
coreywebster
Well-Known Member
Looks like a lower profile heat sink. I take it you have real QBs too? If so how do the heat sinks compare in size and profile.
I can look at my QBs but easier to ask someone who has both.
Zerodayz420
Active Member
Its about half the height of the slate 2. I will try to get some comparison shots tonight.
Zerodayz420
Active Member
Attachments
nfhiggs
Well-Known Member
Yes, that was for the 560 mm single row strips.Were do u get strips for $9 or are u talking about 2 foot strips
Digi 4ft price for 100 strips = $17.88/strip
LINK = https://www.digikey.com/product-detail/en/samsung-semiconductor-inc/SI-B8V521B20WW/1510-2225-ND/6624009
robincnn
Well-Known Member
The driver will also need to change if you want to add extra board to 260XL Kit .
The empty space in center helps with an even light spread and no center hotspot.
robincnn
Well-Known Member
Sphere said Total Lumens: 77963.72
Ya i think 550 is just under 0.145 A. max is .200A.
Rated 65mA current of 561C matches with the LED Strips that use minimal heatsink . higher is fine with good heatsink
There are mid power 3030 can handle 1.3 watts. 6.6V *.2A
yankeeny20
Active Member
Didn't think about it. I ordered the kit Sunday. Can't wait until it gets here.