Build My Own Quantum Light Cheap

Edzzed

Well-Known Member
As the title say, gonna build my own quantum light. I think I can get it done for under 400.00 and it will match or exceed HLG or Atreum's offerings. I have learned they source their lights from Samsung using the LM301B diodes. Shenzen Meijiu lighting company in China gets the diodes and builds boards with them. We know they have various amounts of diodes per board style. The best by HLG or Atreum use 4 boards with 288 diodes each for a grand total of 1152 diodes. So I found a source who has the same diodes in a strip style layout. It's 15 inches or so x 1 1/2 inches and each strip contains 96 LM301B diodes. Math says 12 strips will give me the amount of 1152 diodes. Same output as HLG or Atreum.
Only I plan to make the strips spread out more in my 4x4 tent which in theory should evenly light the whole space. I found out what HLG and Atreum use for a driver and that's

1-board configuration: Meanwell HLG-120H-54A
2-board configuration: Meanwell HLG-240H-54A
3-board configuration: Meanwell HLG-320H-54A
4-board configuration: Meanwell HLG-480H-54A

You can do the math with each board having 288 diodes. looking at 130.00 for the driver. Each strip will be 22.00 x 12 strips 264.00 Total will be 394.00 but then I will need some 2 inch aluminum angle and heat transfer tape. I figure aluminum will be 50.00 and heat transfer tape 18.00 add in 10.00 for solid strand copper wire and I will be at 472.00 My only question will be what color temp should I get.
The seller has 3500K and 2700K. I am thinking of using my Viparspectra par 600 and I have two of those and I know they aren't 600 watts each its more like 285 each. I have never used them but think would they be best for vegging and then use my home made contraption for flowering. What spectrum would you get and why. I was leaning towards 8 of the 2700K and four 3500K. Just need to figure that out.
Or do you think I could sell the Vipars or is there even a market for them. Whoever wants them will assume them to be used but they are new and only turned on to verify they all work. so what color configuration to go with ?????
 
And I got an email from a supplier in China and they tell me the proper driver to get would be HLG-480H-48B Not being an electrical guru I know enough to know there is a difference between HLG-480H-48B and the HLG-480H-54A I don't want to do this wrong so if anyone has any input. Thanks in advance............. In the meantime, I will email the seller of the strips and ask for their advice.
 
Don't know if I'm too late, but I missed the profile post :volcano-smiley:


Have you checked out the original Samsung strips:

I have plenty recipes for these on hand :)




If you plan to flower with them go with 2700K over the 3500K.

Vipars are quite blue for blurples, so they would be great for veg.



Go with U-channel aluminium, flat and L-channel is gonna be a bitch to build with and transfer less heat.



To find the right driver(s) we have to know the voltage and max current of the PCB/strip/board.
If you give me the numbers I'll calculate it for you.
 
difference between HLG-480H-48B and the HLG-480H-54A

One uses a built-in potentiometer for dimming and the other uses an external one. I believe the A versions are the ones that have it onboard, but might be mistaken. <<==== Probably wrong, lol; watch the entire set of videos mentioned below.

There is a set of videos on YouTube that you should be able to find by searching for "The LEDG Guide to Mean Well HLG Drivers." I just discovered it and haven't had time to view the videos as of yet, but my guess is that the set would be useful to those interested in building their own lighting setups (or merely interested in understanding them).
 
I'll do my best to contribute to this topic. I'm no expert so practice due diligence with what I'm suggesting.

Driving an array of LEDs is about matching a power source to the way the emitters behave as a group. Wiring the emitters differently will split power in different ways. There's a dance that happens when you build an array because you need to match both voltage and current consistently across similar emitters. A single power source driving a string (oversimplified) of chained emitters needs to maintain a constant current acceptable to all emitters while also supplying adequate voltage to be divided up between them. Here's my best guess as to why they're suggesting 48V over 54V:

17 parallel rows of 17 diodes in series @ 2.75Vf, 289 diodes per panel (overdriven @ 46.75V on a 48V supply)
16 parallel rows of 18 diodes in series @ 2.75Vf, 288 diodes per panel (optimal @ 48Vf on a 48V supply)
15 parallel rows of 19 diodes in series @ 2.75Vf, 285 diodes per panel (overdiven @ 52.25Vf on a 54V supply)
14 parallel rows of 20 diodes in series @ 2.75Vf, 266 diodes per panel (underdriven @ 55V on a 54V supply)

I'm guessing they suggested that driver because 1) it matches your emitter design best while 2) minimizing your chances of over-driving the emitters causing premature failure (~156ma forward current across 64 rows). This all relates to circuit design and without knowing how your MPCBs arrange your emitters it's hard to say.

In reality either of the drivers could work because they're both adjustable (A versions). The 480H-48 has adjustments for both voltage (40.8 ~ 50.4V) and current (5 ~ 10A). The 480H-54 likewise is (45.9 ~ 56.7V) and (4.4 ~ 8.9A). Notice that you get more power with the 48V driver compared to the 54V driver. If you were flexible with the exact number of diodes you could find an optimal design that would make the most from the driver. I think this is what @PurpleGunRack was offering :) .

The A/B designation is about dimming as someone else mentioned. I personally would go with anything including A because you can fine tune voltage and current output.
  • A: current and voltage adjustable through built-in potentiometer
  • B: 3 in 1 dimming via PWM or resistance
  • AB: both above features

A few other things worth mentioning:
  • Confirm with your supplier what the bin range is for you emitters. Check out page 6 of the LM301B datasheet. Notice that some emitters will be 14% more efficient (4lm/W) then others but they will all be sold at the same price (unless you ask ;) ).

  • Your diode choice has a viewing angle of 120%. Depending on how far away your lights will be from the canopy this could be good or bad. Some diodes have a more narrow angle making photon delivery more direct (less bouncing off of reflectors and walls and such). With larger power emitters and COBs this is sometimes something you can order to spec.
I hope I'm not way off in my understanding of things. If someone sees something out of place let me know!
 
I thank you for the replies. Yes, I did go to the digi-key web site and saw. Of course that was after I had bought the Ebay ones. I wired up one set of 6 strips and OMG are they bright. I was seeing spots. I had ordered two drivers HLG240H-24A from two different sellers one being Amazon and the other a 3rd party seller on Amazon. 2 different prices as one was a warehouse return. That one works fine. The other one for whatever reason was sent to Hollis NH and I get parcels sent to Blaine Wa. Go figure, so Amazon refunded me on that. I then found one on Ebay and it is used but not a lot according to the seller. They wanted 36.00 with a best offer button and I offered 20.00 and it is on the way. I am using 2 aluminum cookie sheets and have 6 alternating color strips on each one to balance it out being 2700K and 3500K should be 3100K when done. Got double sided heat transfer tape. It will probably be about a week or more before I can get the 2nd driver wired up. Too much on my plate right now. Anyhow it is a total relief that one set is functioning. Was worried I might have wasted money by not getting it right.
20190118_140427[1].jpg
 
pretty sure it's 10% to full power,, mine is
 
Glad to hear things are working! If I understand correctly, I think you're using the voltage/current adjustment pots to dim the LEDs? If so be careful. Understand that if your boards lack protection circuitry the emitters can easily be shorted (burned out) with too much voltage. As I understand, dimming LEDs is best done with pulse width modulation which is why one of the driver versions comes with this feature. In any case congrats!
 
Glad to hear things are working! If I understand correctly, I think you're using the voltage/current adjustment pots to dim the LEDs? If so be careful. Understand that if your boards lack protection circuitry the emitters can easily be shorted (burned out) with too much voltage. As I understand, dimming LEDs is best done with pulse width modulation which is why one of the driver versions comes with this feature. In any case congrats!
Thanks, I am pretty sure I got a driver which won't over power the strips but will push them to near their limits. At least I hope that's the case. Each strip is supposed to be 85 watts. .
 
Good job on the build :)


Thanks, I am pretty sure I got a driver which won't over power the strips but will push them to near their limits. At least I hope that's the case. Each strip is supposed to be 85 watts. .

Depends on the max current (Ampere) the strips can take, I'm pretty sure they're +2A so the driver won't overpower them.

Watt = Volt * Ampere, and the voltage of a unit depends on the current -> higher current = higher voltage.
Ie. at minimum current your strips might pull 22V and on max they might pull 28V


BTW that driver will pull around 280w with both potentiometers maxed out.
 
BTW that driver will pull around 280w with both potentiometers maxed out.
I will say I am a TOTAL ROOKIE when it comes to building my own light. I had not even heard of LED drivers until last month. When I took the pic I had them wired in parallel and wonder what if I had them in series. Would that work or what wiring layout would be best. Never even grown before but since it is now legal here. Giving it a go. Guess with Purplegunracks comment I am looking at 560 watts from the wall for both drivers.
 
Serial wouldn't work with that driver

Judging by @Edzzed success I'd say your correct. My best guess is that each panel contains 12 parallel chains of 8 emitters in series? If the panels were 24 parallel chains of 4 emitters in series then you could drive two panels in series with the HLG240H-24A correct? The existing panels in series would simply not light up as there wouldn't be enough forward voltage across all emitters (1.5v). All I'm saying is the driver should match both the panel design and the configuration of multiple panels wired as a group.
 
Judging by @Edzzed success I'd say your correct. My best guess is that each panel contains 12 parallel chains of 8 emitters in series? If the panels were 24 parallel chains of 4 emitters in series then you could drive two panels in series with the HLG240H-24A correct? The existing panels in series would simply not light up as there wouldn't be enough forward voltage across all emitters (1.5v). All I'm saying is the driver should match both the panel design and the configuration of multiple panels wired as a group.


It's not that complicated, you don't need to worry about the connections inside the PCB, just the forward voltage and max current of the PCB, that's what the internal layout adds up to ;)

The driver is CV (Constant Voltage) and puts out 25,6V max so it won't even turn on 2 of these strips in series, if you want to do a serial connection you need a CC (Constant Current) driver that puts out a voltage range that the total combined forward voltage of the strips is within.
Ie. the MW HLG-240h-C1400a that puts out a constant current of 1,4A max and a voltage range of 89V - 179V.

You can also do a parallel/serial combination, or split the voltage on CV drivers etc. Check out LEDGardeners videos on the MW drivers if you'd like to learn more about them :)
 
@Edzzed Nice work! I wish I had a Spectroradiometer so I could duplicate some of these setups and post performance stats. Some day ;) .
it should be equal to the HLG 550 v2 as I used the same chips and driver power that they used, just went about it in a different way. Took a page from the Chinese how they reverse engineer things. lol
 
Plants are now under the DIY quantum board. I had them under a viparspectra for couple weeks and had to whack a few inches off the table legs to accommodate future growth. Figured it was a good time to swap lights out. The plants seem to like the new lights. Been just over 24 hours now since the change. If you look at the pics. The bottom pic was taken yesterday and middle pic was taken just minutes ago. It will be a 24 hour time difference. One PK is doing a little slower than the others and always was.
1732086
1732087
1732088
 
Back
Top Bottom