Attempting a DIY LED fixture - Part 2. New pics 9/27 (56k)

[malaybiswas]

My first step on making an alumimium enclosure. I got a real simple idea, which needs no complex tools (just tin snip, a block of rectangular wood, screwdriver, 22AWG aluminium sheets, plexiglass and some sheet metal screws)

I will post the design and progress of the model as I proceed with making it.

[evilc66]

Calculating the actual PAR of your setup will be very difficult unless you can find the weighted spectral output of the LEDs you want at the drive currents you want to run at. The datasheets don't do into that much detail and typically on list the relative spectral output for color comparison. I'm not sure where you will be able to get that information.

That spreadsheet is the closest thing to being able to calculate PAR, but without the SPD numbers it's no good.

[malaybiswas]

So I referenced a whole lot of information from different sites and got a whole bunch of mind boggling calculations.

Before I get into that, I decided on using white and warm white instead of white and blue as evilc66 suggested before. After reading many articles I agree that this combination would be better in terms of photometric and radiometric (is that a correct term??) light (or energy) output.

So some facts
1. Will use 7 white and 8 warm whites
2. Tank dimensions
20 gallon (22 gallon minus appro 2 gallon worth of substrate)
Substrate surface area approx 24" * 15" or 0.23 sq meter
Distance of LED from substrate surface approx 22" or 0.53 meter

Calculations
1. PAR-watt/sq meter.
Benchmark - 22 = low, 45 = medium, 75 = high and 135 = very high (found it one one site, cannot say if it's correct or not). Following the formula, I tried to calculate my setup of 15 LEDs. First thing required is PAR rating of bulbs. Did not find any authentic source, so I referred to several LED bulb products using Luxeon LEDs which had PAR ratings for the bulb. I divided the PAR rating by the number of LEDs used to approximate the PAR rating of each bulb.

Very approximate, but logically should be correct. PAR is amount of photons emitted by the bulb per unit time. If x number of bulbs emit y photons at at certain power (w), then the average number of photons emitted per LED is y/x (at the same wattage). Since these are Luxeon bulbs, the wattage is same as mine (1 watt or actually 1.2 watt max as per my calculations).

So what do I get. PAR ratings vary between 2-4 PAR-watt. I am assuming 2.67 for the LEDs I will use based on the proximity to the PAR rating calculations for bulbs manufactured with white LEDs

All calculations are approximated
a. Total PAR-watt for 15 LEDs = 15 * 2.67 = 40
b. Assuming 60% of the light reaches the substrates surface, PAR-wattage at substrate surface = 24 (why 60%, I refered to some documentation from a plant growth light manufacturer who suggest assuming 75% of the light reaching the target at similar distance. Since this is under water, I assumed 60% instead).
c. PAR-watt/sq meter = 24 / 0.23 = 104 (High light)

Note that, this assumes that there is no overlap which is not true. However with lenses, and the layout I planned, If I assume that a single LED lense directs 1 unit of light vertically down the axis and the intensity reduces radially on the bottom (e.g 0.9, 0.8 etc.), with the overlap, the cumulative effect of 2 or 3 lamps should always provide 1 unit of light at any point (or more if same point receives light from 3 lenses). E.g. one point on the substrate receives 1 unit from bulb 1 and 0 units from bulb 2 (this point is vertically below bulb 1). Move out a little (radially) and the next point should receive 0.9 units from bulb 1 and 0.1 units from bulb 2 (further from bulb2 and closer to bulb 1). Makes any sense???

2. Lumens/gallon
a. total lumen of 15 lights = 7 (white) * 45 + 8 (warm white) * 20 = 475 lumen
b. Lumen/gallon = 475/20 = 23.75 (High??)
c. If 30% is lost due to refraction etc that would be 16.625 lumens/gallon


3. Watts/gallon
a. Total wattage of 15 LEDs = 1.2 * 15 = 18
b. WPG = 18/22 = 0.82 WPG (high efficiency of LED = low WPG but high PAR??)

4. PGL (Plant Growth Lumen)-hr/gallon
Benchmark (again referenced but unsure about accuracy)
100 = low, 200 = med, 300 = high

a. Total lumen output of 15 LED = 475
Read some where that white light is approx 11% blue, 59% green and 30% red.

Approximated
11% blue light @ 460nm = 0.11 * 475 = 52.25 lumen
59% green light @540nm = 280.25 lumen
30% red light@650nm = 142.5 lumen

b. Spectral output for plant (P) = 52.25*0.198 + 280.25*0.263 + 142.5*0.509 = 166.94

c. Spectral output for human (L) = 52.25*0.06 + 280.25*0.951 + 142.5*0.108 = 285.05

d. P/L = 166.94/285.05=0.59

e. PGL = 475 * 0.59 = 278

f. Running lights for 10 hrs a day for 20 gallons = (275 *10)/20 = 139 PGL-hr/gallon (medium low??)

If you are wondering where I got all these weird factors and formulae, refer here.

So as you can see I have beaten it to death and have not been able to reach any justifiable clue.

However anything more than 15 LEDs do seem to point towards wastage of the light energy (or may be not = algae).

So for now I will give myself a break from this and finalize with 15 LEDs as I planned. We'll see how it goes.

evilc66, I checked my layout plan with 45 degree lenses. Pratically I need only 2-3 LEDs to cover the ground surface and a lot of light wasted outside tank (or too many LEDs overlapped). So I will stick with the 25 degree lense for now.

[malaybiswas]

Ordered the luxeon LED and accessories today. projected to come in by March 12. So I have sufficient time to check on the dimmer circuit by then.

Meanwhile I finished the first of 2 fixtures. I will post the DIY process tomorrow.

[Shoey]

EBay has several at $10 ~ 20.

I am going to use 3 ~ 12" over my 90 gal

[malaybiswas]

Originally Posted by Shoey

EBay has several at $10 ~ 20.

I am going to use 3 ~ 12" over my 90 gal

How many LEDs do you have?

[Shoey]

Plan on using 3 banks of 10 each 3w CREEs, but.............

I will be using optics which are specific to my scape. For example in the foreground, more spot type (~45 deg) optics will be used to penetrate to the substrate, the background optics will be more flood type (~90 deg). I know this can create hot spots but I hope to balance flood and spot to eliminate this.

Worst case I will use 90 deg on all of them. The optice are cheap on EBAY


Al

[malaybiswas]

Originally Posted by Shoey

Plan on using 3 banks of 10 each 3w CREEs, but.............

I will be using optics which are specific to my scape. For example in the foreground, more spot type (~45 deg) optics will be used to penetrate to the substrate, the background optics will be more flood type (~90 deg). I know this can create hot spots but I hope to balance flood and spot to eliminate this.

Worst case I will use 90 deg on all of them. The optice are cheap on EBAY


Al

Cool. I am not planning to use any heatsinks separately (unless they really get hot). My fixture is all aluminium and the LEDs will be setup on a aluminium sheet. Based on my calculations that should be enough and my enclosure is all perforated aluminium sheet so the heat should escape easily. So I will get heatsinks only if necessary.

I am getting 25 degree optics as well. I calculated the spacing of the lights with the optics to ensure that I am getting an even distribution of light on the substrate.

[fishman9809]

i thought u were gonna post pics? D:

[malaybiswas]

Originally Posted by fishman9809

i thought u were gonna post pics? D:

I am. Just had to finish the first part of the project. But promise...within the next couple of hours

[Shoey]

Watch your heat, it is a

You may want to start at a lower current check your temps and work your way up from there. If you see the the LED discolor (yellow) when turned off you most likly got too hot. Been there done that, but that is another story, bicycle light.

Al

[malaybiswas]

I realized that I have got too many pictures for this, so I need to to breakup the posts instead of one gigantic one.

First, the raw material

1. Perforated aluminium sheet. Many reasons why I chose this....light weight, easy to cut and bend, almost no drilling required, will allow heat to dissipate out of the enclosure from all sides and will look cool (probably)

The only downside is that when you bend it along perforations, they tend to break off, but there is a fix for that too.


A closer view


And the exact name under which it is sold in Lowes


2. Sheet aluminium (1/2" wide)...this will be used as reinforcements


3. Angle brackets. Also for reinforcement, hold on the bent sheets and maintain the 90 degree angles on the edges


4. nuts and bolts. 6-32 (1/2") works best. Also had a few 4-40 in my garage but probably they won't be required.


5. Aluminium angle bracket (48" long)


6. Plexiglass sheet (1/4" thick) cut into 24" X 15" piece


7. Flat aluminium sheet. 2 pieces (18" X 12") to be cut to proportions. They will be the base for the LEDs and the heatsink.

[malaybiswas]

1. Measuring ruler


2. my beloved portable work bench


3. tin snip


4. wood blocks, for bending aluminium sheet


5. hammer (also for bending sheet) and screw driver


6. clamps


7. marker


8. drill. I'll still need it to drill holes in the other aluminium pieces without holes

[malaybiswas]

The plan is simple. The perforated sheet will be cut as shown below. The overall dimension of the enclosure is 24" X 15" X 3" (1" depth reserved from the bottom for the optics. 1" from the bottom will be the aluminium plate with the LEDs. The 2" about it will house the adaptor and other circuits (if I make any)

.I kept the side wings to 2" depth so that the LED base plate can be slipped in or out of the enclosure from the sides without any disturbance. That's why I chose 2 - 12"X18" base plate sheets instead of one so that they can be slipped in from either side (the plates will be cut into 12" X 15" plates so that the combination provides a 24" X 15 " surface area).

[malaybiswas]

1. Cut the enclosure aluminium sheet into 26 " (24 + 2 + 2) X 21" (15 + 3 + 3) proportion



Pretty easy to cut with the tin sip


2. Bent the edges using the wood blocks and hammer. Actually the plate is thin enough to be bent with hand. The hammer can be used at the end to sharpen the corners.


And this is how it looks.


3. Next I fixed all the edges with the angle brackets






4. Cut the 1/2" wide aluminium sheet into 15" pieces (4 of them in total)


5. Two of the 4 pieces were fixed on the sides to provide rigidity to the wings


6. I cut the 48" L bracket into 2 24" strips...


...and fixed them at the 2 long bottom edges.




These will form the rails on which the plexiglass cover will be slipped in...


...like this


7. I fixed a series of L brackets at 1 " depth from the bottom along the 2 24" long wings. These will hold the base plate with the LEDs as shown below (actually the plate will be on the other side of the brackets so that they don't crash into the plexiglass sheet when the enclosure is hanged upside down



And this is approximately how it will look from the bottom


8. Finally the other 2 pieces for reinforcing the rear cover.



I did a minor weight check to make sure it will be able to hold on to the total weight of the stuff inside. Did pretty well given the thickness of the perforated sheet.

This is as much I can finish up for now. The rest of the construction will commence after I get the LEDs in hand. Finally I have to device the hanging mechanism as well.