DIY LEDs - The write-up

[The Grim Reefer]

Submitted for your approval

Submitted for your approval

Or maybe not :crazy1:

Presenting the Ghettomaster XB1138 Photon Regurgitator.

I set out to build something that could be done with simple hand tools. The only thing I had special was a Dremel with some cut off wheels. Other than the drivers, LED's and Aluminum channel everything was accessed locally, mostly at Lowes.

Heatsinks are just 2x1 8th inch aluminum channel with the LED's mounted inside. I had a bunch of little heatsinks with adhesive on them so just for giggles I mounted them on the back of the channels behind the LED's. I only have 3 of the 6 Plexi shields installed. Amazingly enough I was able to mount the shields with just zip ties. I thought I was going to have to make a bunch of clips, THANK YOU GOD. The plexi keeps the spray off the LED's and wiring and also should the thermal tape not hold the LED's as advertised keep the LED's from taking a bath

Had a Reefkeeper laying around so I figured what the heck. Self contained 7 channel programmable lighting. Used 12V Automotive connectors to make up the connections between the Drivers and strings.

I attempted to stagger the LED's and did a pretty crappy job of it but then again doesn't really matter. The spacing is 3" fore and aft except for the first 2 rows. Forgot to account for the overflow when I spaced things out so rather than rebuild the whole thing I just moved the front channel back. I figure that will bump the intensity a little more out on the sandbed where it's needed more anyway. That is part of the reason I built it like an erector set. I wasn't sure if I would need to add another row of LED's or change the spacing. LED's don't have any optics so the light blends really well and sitting on the top of the tank it's putting 170 to the bottom.

52 XRE Royal Blues
6 XRE Q5's (Front row)
8 XRE Warm Whites
18 XPG cool whites.

There are a lot of things I'd do different/better knowing what I know now but overall I am pretty happy. Now I just have to set the dang tank up :hmm4:

 

[laverda]

Laverda- You do have a good point. How much do the HLG-120-54B Meanwells cost anyway? I guess this tangent is just for my interest, being that no time soon will I be running anywhere near enough LEDs for the parallel strings to be worthwhile. I just run the little tanks, and don't see myself ever going over 24 LEDs, where I would need at least two separate drivers anyway to be able to adjust the blue balance.

Furthermore, my original LED build, that I don't see myself replacing anytime soon, was done with a 24v supply plus three buck pucks, and as I add a couple new little fixtures to light a rimless fuge and then some, I'm just going to be using a couple buck pucks and tethering them to the same power supply.

Interesting to know the various options nonetheless.

I have just made two requests for price quotes. So I will let you know.

 

[kcress]

Presenting the Ghettomaster XB1138 Photon Regurgitator.

:eek1::lol::lol::lol:


Rad! All Road Warrior, Back O' the bus like.

 

[sammy113]

Looks great Grim Reefer! it will be interesting to see how all these LEDs combine the colors over a tank. Gotta admit I love the name :lol:

 

[tahiriqbal]

lovely stuff.... the only way to learn and enhance your skills and knowledge to try, try, try and try... Would love to see warm white kicking inn. Please keep us updated.. Just keep an eye on your zip ties, with heat they may become a bit of an issue.

I got this aluminium architectural profile from local supplier, £13.80 per 4 meter strip. Ideal for sliding 3mm plexi sheets..

 

[The Grim Reefer]

Looks great Grim Reefer! it will be interesting to see how all these LEDs combine the colors over a tank. Gotta admit I love the name :lol:

Yeah, The name might be the best part of the whole project :lmao:

This is what it looks like over an empty tank. Notice the junk in the bottom

 

[Altpers0na]

want!!

 

[Obi-dad]

I have not heard anything bad, but someone (maybe widmer) found where Cree said to use screws.

I have done some experiments with different mounting methods (thermal adhesive vs screws/thermal compound) on different substrates (aluminum bar vs individual heatsink), and measuring not just the substrate temp, but the temp of the dome on the LED - this is the point that gets hot, and is closest to junction temperature. The thermal adhesive was within a few degrees F of the screws/thermal compound. Not significant IMO.

These temp readings were made using a very small thermocouple, not the typical hand held IR guns which have a much larger diameter of measurement than people realize (it is not the little red circle of the red guide laser that you can see).

And in spite of what Arctic Silver claims, the adhesive is not permanent. You can use a chisel between the star and the heatshink, and it pops off clean - no residue on the starboard, and very little on the heatsink (provided you use the adhesive the correct way, which is to use a small amount, then swirl the star while pressing down hard on it so that it achieves a very thin layer for max heat transfer).

 

[TheFishMan65]

Obi-dad, since you appear to have more temperature experience than most do you think Grim is going to be in trouble? I know they say temperature will shorten LED life. And you are saying the lens gets the hottest. That means Grim has trapped all the heat in the channel with the shield. I did not notice fans blowing through the channel so let's assume they are not there.

Grim. can you measure the inside temperature of those channels?

Thanks all

 

[bstohrer]

Speaking of U-channel... I'm planning to make some undercabinet lights for the kitchen, and tentatively am planning on doing one XP-E @ 700 mA per 6" of the 1" wide aluminum U-channel. Think that's enough space between them for passive cooling to keep the temp low enough if the lights were left on for any amount of time?

I"m not sure. If your U channel was in the open, I'm pretty sure you would be fine. Not sure about up under a cabinet where you would get little convection. You may just have to build a prototype and give it a try.

For reference, I used 1" X 1" X 0.125 channels with the XR-Es spaced at 3" running at 1000 mA. With no air flow, in the open, the temperature at the back of the star (probe in tiny hole drilled in channel) leveled off at about 140 F. With very gentle airflow, temperature leveled off at about 100 F.

Bob

 

[The Grim Reefer]

Obi-dad, since you appear to have more temperature experience than most do you think Grim is going to be in trouble? I know they say temperature will shorten LED life. And you are saying the lens gets the hottest. That means Grim has trapped all the heat in the channel with the shield. I did not notice fans blowing through the channel so let's assume they are not there.

Grim. can you measure the inside temperature of those channels?

Thanks all

Haven't yet but I can. The original design was going to have air blowing down the channels but after thinking about it I decided that would be a bad idea, cool outside air being drawn into hot channel could lead to condensation build up unless I was pushing a whole lot of air. There is a lot of room inside those channels and the ends are open so I don't foresee any real heat issues.

 

[jtma508]

Nice Grim. I'm in the early stages of designing a system myself and am gravitating toward using channel as well. In my case, I was planning on using the channel as a heatsink and mounting the stars on the flat profile (opposite of the way you used it) thereby making the vertical profiles act as fins. Like you I also plan on using a mix of warm and cool whites to get that 'bump' in the 500nm + range. Well done.

 

[TheFishMan65]

Grim, I would be interested in knowing what the center is running if you can do it. Thanks

 

[GEORGEDOPE]

I have just made two requests for price quotes. So I will let you know.

I've also made a request for a price on the HLG-100-48B.
One question. The adjustable current range on the HLG's does not go below 1A.
Is that a problem for running the LED's at 750mA? or don't I understand yet?

 

[TheFishMan65]

It is if you only have one string. Or if the string are very unbalanced.

 

[GEORGEDOPE]

It is if you only have one string. Or if the string are very unbalanced.[/QUOT

Can you explain more please, for I am very dense when it comes to W,A & V.

 

[GEORGEDOPE]

Maybe I got it. If the lowest current setting on the driver is 1.9A and you have 3 parallel strings, then each string only gets .063mA. Right?

 

[TheFishMan65]

Lowest current is 1.9 (if you say so). Now if you have one string then you just lost all your LEDs since they got 1.9 amps. If you have to string and they are balanced exactly they get .95 (2 * 0.95 = 1.9) amps. But if they are unbalanced or change with temperature (or something else) you might get 1.1 and 0.8; my guess it not am immediate loss but I am not going to find out. Three strings .6333 (decimal place wrong above), and you are probably OK with a slight imbalance. However there is a risk that you loose an LED pop a wire loose, etc and you are back to the 2 string model. Loose one of the remaining 2 and fry the last string. You need fuses or balancing resistors.

NOTE: I don't like balancing resistors, because they are just a waste of power (read money). And as someone pointed out (kkcres?) fuse holder are expensive. So small fixture are probably still better of with multiple drivers IMHO.

 

[The Grim Reefer]

I did a temp and PAR check last night.

Start up temp was 77 degrees measuring against the inside slot of an XPG star. PAR was 113.

After 10 minutes temp was 84 degrees, no PAR change

After 15 minutes temp was 88 degrees PAR was 112

After 30 minutes temp was 94 degrees, no PAR change

After 1 hour temp was 97 Degrees and PAR seemed to stay at 112.

The room temp was in the low 70's with a ceiling fan running above the fixture.

Grim, I would be interested in knowing what the center is running if you can do it. Thanks

Those readings above where taken at the center. While the acrylic wasn't installed the fixture was sat on a table which I am pretty sure was holding in a lot more heat than a piece of acrylic would.

 

[kcress]

cool outside air being drawn into hot channel could lead to condensation build up

Grim; You have that exactly backwards. :hmm5: There can be NO condensation in that case. It would have to be warm moist air drawn into a cool place...