DIY LEDs - The write-up

Status
Not open for further replies.
Looking at the XP-G data shoot I think that if you're trying to stay under 48v at 1.3A than 14 is going to be pushing it, I think the vf at 1.3A (as depicted in the data sheet) is going to be closer to 3.47 V. However the ELN-60-48 lists that the voltage range can be adjusted up to 52.8V. I'm not sure if there are any other consequences of doing something like that but it might be worth looking into.

But running 14 at 1000 mA should be no issue. I am most likely going to run 13 per driver at 1000 mA and leave room to power higher. My likely build is 13 CW, 13 NW, 35 RB (@700) mixed with 4 blues, for a total of 65 led's. And I will likely need to dim those whites some. Starting to think the XM-L's for what I need to do are not worth the added cost. The output of the rb appears to be around 175 lm @700mA, am I reading that right? Whatever, that will put out an amazing amount of light regardless.

I would imagine pushing the voltage on the driver would shorten it lifespan and heat could become an issue.
 
I have never seen royal blues listed for lumen output. So what are you reading and I will tell you (if I can) if it is correct.

Read the summary one heat kcress rethought it and did some research.
 
I have never seen royal blues listed for lumen output. So what are you reading and I will tell you (if I can) if it is correct.

Read the summary one heat kcress rethought it and did some research.

The blues are listed at like 31 at 350, the rb for some resone are listed in flux, and I am not sure how that traslates.

What did kcress rethink? The heat on the drivers or the heat for the linear setup? My setup will be 4 or 5, 40 inch u-channels in the mh fixture, and I plan on running 2 or 3 120 mm fans. I am not sure if I will put acrylic covers (the fixture has slits to accept the covers), or I may get glass covers for the tank itself.
 
RedM3; I don't believe the OP even has a tank anymore so he's not going to have a problem with that in any case.

Further, you're correct in theory. But! Since these can't have water sitting on them there won't be any corrosion due to that leaving only condensation. I wouldn't expect much condensation on something many degrees hotter than the surrounding atmosphere. Ever seen a hot cup of coffee have condensation running down the sides?:)

The little amount that might somehow condense on the surfaces at 5am probably wouldn't be enough to capillary into the threads.

With a LED fixture I wouldn't worry about that. That said, I'd probably just use zinc plated hardware to avoid the issue completely since the same logic sez they won't rust. :beer:
 
The blues are listed at like 31 at 350, the rb for some resone are listed in flux, and I am not sure how that traslates.

Luminous flux is perceived light...how the eye sees it. Since RB's are approaching ultra violet the 'perception' becomes a poor rating. Radiant Flux in mW is a meaure of total power output in this case. So the XP-E RB has a radiant flux of 500mW @1A...which translates into 1/2 the power being put in is coming out.
 
Luminous flux is perceived light...how the eye sees it. Since RB's are approaching ultra violet the 'perception' becomes a poor rating. Radiant Flux in mW is a meaure of total power output in this case. So the XP-E RB has a radiant flux of 500mW @1A...which translates into 1/2 the power being put in is coming out.

1A current, with a forward voltage of about 3 volts is 3 watts. So isn't 500mW about 1/6th of the power?
 
Ok, officially ordered. 35 RB, 4 blues, 13 CW, and 13 NW. Will be run at 13 per string on 5 x eln-60-48d drivers. This will be over a 120g with a footprint of 48 inches by 24 inches. I plan on running the blues at 700 mA and the whites at 1000 mA. I went with a little fewer after reading SOOOO many threads about people setting these up and then needing to dial them back to 75% or so. They will be about 8 inches over the water in a retrofit 48 inch MH fixture. Wish me luck.

I have not decided how I will dim them yet, I may just go with pots, but getting a Neptune and running these on the 1-10 v dimmer could be cool.
 
Leds

Leds

True, and I guess this further goes to show how you really can't name an all around "best" LED. There are simply to many factor that go into that are specific to every project (and where you are purchasing from).

Has anyone seen any practical (for our applications) optics for the XM-L's yet?

Since XM-Ls have such a high lumen output and a 160 degree viewing angle, you may not want ANY optics on them. Try THAT first. Then if you want, apply any lens that will come out for them, later on.
 
I thought I would post this link. After looking at this thread and the LED aesthetics thread I tried building a 12 LED array to see if I got the color balance I wanted. What surprised me was the low PAR numbers I got versus those shown on the forum for full panel buildups. I thought localized results would be similar to the full panel. They were not. If you are curious the discussion can be found here.

http://reefcentral.com/forums/showthread.php?p=18554889&posted=1#post18554889

Regards,
Mark
 
I thought I would post this link. After looking at this thread and the LED aesthetics thread I tried building a 12 LED array to see if I got the color balance I wanted. What surprised me was the low PAR numbers I got versus those shown on the forum for full panel buildups. I thought localized results would be similar to the full panel. They were not. If you are curious the discussion can be found here.

http://reefcentral.com/forums/showthread.php?p=18554889&posted=1#post18554889

Regards,
Mark

FWIW....I built a 12 LED test setup and put it over a 14g biocube 11 months ago. 6 XPG R5 cool whites, 6 XRE RB's. No Optics, 2 strips of 6 alternating CW/RB, each LED 2.5" apart, strips 3" apart. Running a Meanwell 60-48D at 1.3A (yes, a torture test for the XREs and they're still going strong). Growth has been tremendous but color temp not quite as cool as I would like it but still very good. nonetheless, its been a great setup. 2 months ago a couple fellow reefers decided they want to come check it out as we were discussing moving the idea to larger 120/150g's. So we took some PAR readings in the little cube with an Apogee.

At the surface we got 1100 which was about 4 inches below the LEDs. At the far front corners we got 200 which is about 12". I am growing several monticaps, couple chalices, bunch of zoas, and after that test one of the guys gave me a tri-color acro for a test and it intially lost some of its purple tips but now has come back very strong, bright and the branches have really thickened up and the polyps are flying fully extended. Last point is I have been running it with the dimmer turned back to about 3/4 (by eye not measured current) because it was just too bright. Last weekend I took out 2 whites and added 1 RB...no new measurements but the color is better and the everything looks a tad happier especially the acro.

Next we went to my 29g cube with a 150W 20,000k MH. At the surface we got 700 which was about 6 inches. At the furthest point we could reach we got 100 which was about 15". In here I have an open brain, favia, monti caps, hydnaphora, zoa's, some softies, and a purple clam....so a large mix. Very established tank 2+ years.

Last thing we did was hold the little 14g hood over the 29g cube and took 2 quick measurements. Center bottom of tank we got 500, far front of tank we got 150 (MH was 100). By looks alone all 3 fellow reefer thought the little 12 LED setup would be fine....I disagreed because I like my tank brighter.

I really suspect your NWs are the cause of your low readings. It probably helped mine because I'm over driving the RB's. Finally I agree with your theory to some extent on larger setups and over lapping cones. One thing that would be interesting would be a no optics test at the approximate same distance as mine for a more definitive comparison.

Regardless though...in my little setup its clear that PARs of 100-200 are sufficient to grow some nice stuff. I don;t know the model of the Apogee we used but can find out.

ps...crank up the whites to 1300ma and while you might not like the mix it'll ump the output 30% I suspect. Right now I am convinced 60% RB and 40% CWs and all pushed to high end is way to go. (1300ma for XPGs and 1000ma for XPE's)
 
Last edited:
Thanks for the feedback brassmonkyballs. I will try a reading at 4 inches for comparison. And then factor in 30 % for the difference in our drive current levels. and see where things stand. I am ordering a full set of LED's for a 49 LED bank and then will be able to see the effect of a full bank and the mutual overlap.

I am thinking of going with Thomas Research supplies so that I can run straight series strings of approx 24 leds per string. Only downside is you have to pick a fixed current up front. 700 ma, 1050 ma or 1400 ma. I have been thinking of the 700 for the RB's and the 1050 for the NW's.

Regards,
Mark
 
I like the sounds for price and all, but kcress posted (it maybe in the summary) a really good reason not to go above 50 volts for your string. Lets me see if I can track it down.

[EDIT]
Nope not in the summary and I am not finding it. If someone remembers where it is please PM me. Thanks
 
Last edited:
It is always the last place you look :)
There is some confusion running rampant here...

Current can kill you, yes. But without voltage there is NO current.

The only current that kills you is the current that runs through your body and specifically across your heart.

You need about 5mA running across your heart to cause fibrillation which is a mis-beating heart that pumps no blood. This needs to be avoided... For obvious reasons. :D

How much current a source can provide is completely immaterial with respect to how dangerous that source is. That is - if the source can supply more than 5mA. It doesn't matter if it can supply a million amps it is not even vaguely hazardous if it can't force 5mA of those million amps through your chest cavity.

You are no more than wet resistor to a voltage source. The resistance of your body is squarely in the equation of electrocution.

What is your body's resistance situation? Well, your innards are a wet goop of salt water. A great conductor. Through your chest cavity probably a measly 1 to 10 ohms. Luckily our skin is a very different story. Skin is about 25,000 ohms. So what do we have? 25k + 25k +10 = 50,010k ohms. Lets just stick with 50k.

You can walk up to your car battery and grab both terminals and what do we get? Let's keep in mind that a typical car battery can dish out 800A, but it's at only 12V. Using Ohms law which COMPLETELY describes what will happen we get:
12V / 50,000 ohms = 0.24mA. Chances are you will not feel it at all. I will point out that you actually can feel it but only through your forearms and just at the skin surface.

Note that even though 800A are available nobody is dying. Current availability is un-important with regard to electrocution danger level. Why? Again. Because without the voltage there is no way to drive any current anywhere.

Let's move on to our 48V current drivers. Manage to hook yourself up to one in the worst possible way and Ohms law tells us that we could get (at most): 48V / 50,000 ohms = 1mA running through our meat-sack. You will feel that... It's like someone overhands you with a tennis racket in the chest. But statistics and a 100 years of electricity, tells us that won't kill anyone. (your mileage may vary)

Moving on to these 230V drivers.. The story changes here.
230V / 50,000 ohms gets you the magic number.. 5mA! Since the drive is up for supplying 700mA it certainly has no problem loaning someone's chest 5mA.

It should be noted that the 50k People Resistance is heavily dependent on your skin at the time. If your hands are dry and recently washed in something like dish soap your skin resistance will be much higher. This reduces the amount of current that can flow through you. If your hands are a little greasy and you've been handling a bunch of metal and touching this and that during a build your skin resistance will be a little lower and shock severity will increase.

Unfortunately in our hobby we go a whole lotta steps further and end up with soggy, wrinkled hands, dripping with salt water. This severely dumps our skin resistance in the toilet. Perhaps down around 10k, greatly, increasing shock hazards. Moving many cases that would normally be a shock to a very painful shock, and something that was a painful shock into a painful episode for your surviving relatives.

It's the voltage folks. The voltage when hitched to any current availability that's greater and 5mA.

This is why I suggest that unless you're a consummate electrical assembler you don't dabble in these high voltage LED drivers. The result of an accident is just not worth the piddly savings of a few resistors and terminal blocks needed to have a relatively hazard free build.

Heli has done a really clean job here building a very nice fixture. He's been very fastidious in his build and made it as safe as possible. I won't expect to see his name in the obituaries because of his nice fixture - unless it falls on him. :lol:

I, wouldn't want to go to that much effort just so I can avoid some terminal blocks and paralleling. Without a doubt, 50V is below immediately fatal voltage levels to someone working on their fixture in the typical situation. Two hundred and thirty volts is most definitely not.
 
Thanks for the feedback brassmonkyballs. I will try a reading at 4 inches for comparison. And then factor in 30 % for the difference in our drive current levels. and see where things stand. I am ordering a full set of LED's for a 49 LED bank and then will be able to see the effect of a full bank and the mutual overlap.

I am thinking of going with Thomas Research supplies so that I can run straight series strings of approx 24 leds per string. Only downside is you have to pick a fixed current up front. 700 ma, 1050 ma or 1400 ma. I have been thinking of the 700 for the RB's and the 1050 for the NW's.

Regards,
Mark

The TRC drivers do seem nice but I'm not sure where to get them. Only place that has them listed is Nanotuners and have for a couple months with an availability of March sometime....and to be honest you could run Meanwells a lot cheaper... They show the TRC-120S070DT for $119.95. Voltage range is 68-114 and 700ma. Thats about 32 LED's in series. you can run 2 strings of 14 (which 28) at 700ma on a Meanwell ELN-60-48D for $30....I've done it.

I don't know about going over 50v on a string but there is something somewhere I saw but do not recall.
 
Thanks for the thoughts "TheFishMan65" and "Brassmonkeyballs".

FYI< Without optics and at about 3-4 in below the LED's I am getting PAR readings of about 500.

Kress's cautions on voltage etc are spot on.

I am considering the following.
1) I really do not like parallel strings and would prefer to avoid matching strings for current.
2) Meanwell supplies do not have power factor correction so it can be bad if you need to run a bunch of them off the same circuit. (the Thomas Research supplies are PF corrected).
3) My highest voltage is expected to be 80 v although the supply can go up to almost 100v.
4) The key to using the higher voltage supplies is to a) be sure that electrode spacing from ground is adequate. and b) to seal the connections so they can not be either contacted or subject to salt deposition.

I plan to RTV all of the exposed PC board and solder connections before the rig is anywhere near water. It is also important to thoroughly check out all wiring before applying power.

Final thought I have an MSEE so I think I can do this correctly. We will see (g).

Regards,
Mark
 
Only thing I can add is that aside from the LP and ELN series all the rest of the Meanwells are active PFC.....then again, you have to go parallel and I don't think they offer anything over 54V. What are the biggest concerns for going parallel?

Good luck.


Thanks for the thoughts "TheFishMan65" and "Brassmonkeyballs".

FYI< Without optics and at about 3-4 in below the LED's I am getting PAR readings of about 500.

Kress's cautions on voltage etc are spot on.

I am considering the following.
1) I really do not like parallel strings and would prefer to avoid matching strings for current.
2) Meanwell supplies do not have power factor correction so it can be bad if you need to run a bunch of them off the same circuit. (the Thomas Research supplies are PF corrected).
3) My highest voltage is expected to be 80 v although the supply can go up to almost 100v.
4) The key to using the higher voltage supplies is to a) be sure that electrode spacing from ground is adequate. and b) to seal the connections so they can not be either contacted or subject to salt deposition.

I plan to RTV all of the exposed PC board and solder connections before the rig is anywhere near water. It is also important to thoroughly check out all wiring before applying power.

Final thought I have an MSEE so I think I can do this correctly. We will see (g).

Regards,
Mark
 
Often nothing needs to be done.. It's more likely you might have to swap a few LEDs around if you have more than 2 strings in parallel. A few minutes work...
 
Status
Not open for further replies.
Back
Top