DIY LEDs - The write-up

[der_wille_zur_macht]

I'm always preaching to the disciples of the holy alter of TeeFIvery that you get what you pay for. What are they doing that they can sell them so cheap?

In addition to what widmer said above, it's worth noting that we're practically buying raw, individual semiconductors here; one or two steps from the factory. It's closer to a "raw material" commodity market than a consumer market, and there will definitely be fluctuations.

In the T5 world, even if you buy a lamp or ballast from an OEM instead of a reef-specific vendor, you're buying a conglomerate or parts that's passed through many more hands, which stabilizes the cost.

The good thing about buying raw LEDs is that you get what's on the order label. As long as the vendor shares the kit code or other info that lets you confirm the bins, you're good to go.

 

[TheFishMan65]

If you search the XR-E data sheet, R2 is red (dominant wavelength 620). I am confused. Skeptic are they red?

 

[der_wille_zur_macht]

FishMan, "R2" is used as a brightness bin code AND a color (dominant wavelength) bin code by Cree.

For the XR-E, it's used as a color bin for the "red" model. For other Cree products, it's used as a brightness bin code for white LEDs.

It fits the normal progression for white brightness bins. Cree uses a letter followed by a number and increments the letter every few numbers, usually starting the next letter at two. i.e.: P4, Q2, Q3, Q4, Q5, R2, R3, R4, etc.

So, if Cree had continued development on the XR-E product, the next logical brightness bin for the cool white color after Q5 would have been R2.

FWIW, I've seen the R2 brightness bin applied to XR-E cool white elsewhere on the web, just never in a documented, official manner.

 

[Skeptic_07]

If you search the XR-E data sheet, R2 is red (dominant wavelength 620). I am confused. Skeptic are they red?

no they are definitely not red.

 

[TheFishMan65]

Thanks all

 

[The Grim Reefer]

Has anyone tried the MC-C chip? I was thinking of using a 50/50 mix of Royal Blues and the Q5's with maybe 6 MC-E's with no optics on a driver would be interesting

Seems to me that the XP-G with no optics might not put as much light in the tank as the Q5 with no optics due the the Q5 having a tighter beam. Yes, No? Yno?

 

[der_wille_zur_macht]

Has anyone tried the MC-C chip? I was thinking of using a 50/50 mix of Royal Blues and the Q5's with maybe 6 MC-E's with no optics on a driver would be interesting

You've probably discovered this, but the MC-E is basically four XR-E emitters on a single die. Four times the light, four times the power, four times the heat, from a single point instead of four points. It's also roughly four times the cost. The current limit is lower (probably because of the difficulty the package would have dissipating heat at higher currents).

Personally, I'd make the call on an application-specific basis. I don't think I'd use MC-Es on a conventional build where you're trying to evenly spread LEDs across a large heatsink, but if you had some particular need to increase the density in a very small area, they'd be a good choice.

Seems to me that the XP-G with no optics might not put as much light in the tank as the Q5 with no optics due the the Q5 having a tighter beam. Yes, No? Yno?

It really depends. Mostly, on the ratio of "edge area" to "middle area" on your tank. On a small tank, ALL the LEDs are pretty much near the edges of the tank, and thus in danger of wasting light by projecting it outside the tank's boundaries. On a large tank, only a small portion of the LEDs are near the edges, and thus there's less waste.

Otherwise, it just breaks down to the incident angle of the light on the inhabitants in the tank. With a tighter beam, the light is far more directional. With a wider beam, you get a little more "sideways" light hitting the corals instead of just vertical light. This could or could not be a good thing, depending on your design and your goals.

I'm willing to bet that people with narrow optics and a conventional build where all the LEDs are spread evenly over the tank pointing straight down will see some weird growth patterns several months/years down the road.

 

[der_wille_zur_macht]

To add a thought to the above, I'm waiting for people to start making more use of angled LEDs to control "waste" outside the tank boundaries, instead of JUST using optics for that task.

 

[The Grim Reefer]

To add a thought to the above, I'm waiting for people to start making more use of angled LEDs to control "waste" outside the tank boundaries, instead of JUST using optics for that task.

I already picked up an angle tool for just that reason.

 

[hllywd]

DWZM... An idea I'm thinking about is using a ball swivel, speaker mount type setup with the intention of angling at least my end two modules to do just that. Someday... I have some optics coming with my order from the group buy I got into and was planning to play with a combination of the two, optics and angles. My catch will be the room I have in my canopy.

BTW - Got my DX Q5s in the mail today, tried one with a couple AAs and I'm still seeing spots! Not sure how to judge if they look good, or look bad, they look like the pic!

Tim

 

[der_wille_zur_macht]

DWZM... An idea I'm thinking about is using a ball swivel, speaker mount type setup

That's an innovative approach!

I have some optics coming with my order from the group buy I got into and was planning to play with a combination of the two, optics and angles

I built half a dozen nano-sized fixtures last summer to do just that. It's really cool to see the differences, I'd encourage anyone embarking on a large build to do one or two small fixtures (4 - 6 LEDs each) just to play with.

BTW - Got my DX Q5s in the mail today, tried one with a couple AAs and I'm still seeing spots! Not sure how to judge if they look good, or look bad, they look like the pic!

You can't tell other than measuring with tools a hobbyist won't have. Though to give the vendor credit, I really doubt they're pulling a fast one; people have reported predictable results with their LEDs.

FWIW two alkaline AA's will be ~3v with translates to around 150mA. When you drive them at the "proper" voltage on an actual driver, they'll be 3 - 4 times brighter than that. Which brings up a warning that bears repeating - DO NOT look directly at an HP LED when it's running at full current!

 

[The Grim Reefer]

That's an innovative approach!



I built half a dozen nano-sized fixtures last summer to do just that. It's really cool to see the differences, I'd encourage anyone embarking on a large build to do one or two small fixtures (4 - 6 LEDs each) just to play with.




You can't tell other than measuring with tools a hobbyist won't have. Though to give the vendor credit, I really doubt they're pulling a fast one; people have reported predictable results with their LEDs.

FWIW two alkaline AA's will be ~3v with translates to around 150mA. When you drive them at the "proper" voltage on an actual driver, they'll be 3 - 4 times brighter than that. Which brings up a warning that bears repeating - DO NOT look directly at an HP LED when it's running at full current!

I got a sneaking suspicion those running 40 degree optics might find this out the hard way with some of their corals.

I am going to remodel before I set up the tank so I got lots of time to experiment. Think I'll order a little kit from rapid and play.

 

[F35-Joint Strike Fighter]

Joint Strike Fighter -

It's not commonly recommended that you run the LEDs in parallel. You're best off driving the 6 LEDs in series, but you're going to have to have a 24 volt power supply. This is because as you add things in series, the voltage is additive (in contrast to the amperage, which is not additive when things are in series. Thus 6 leds X a little under 4 volts apiece = 24 volts.

If you were to run the LEDs in parallel, 5 volts would be OK to drive them, but then you would still need to regulate the current going into each one, and it is recommended at this point that each one has a fuse to protect it in case of excess current.

Thanks!

i will add the resistor so the voltage drops each LED ~ 4V, and for the current, it's variable so I think I will be OK.

 

[der_wille_zur_macht]

Can you post a link or datasheet for the power supply you're trying to use?

 

[der_wille_zur_macht]

I got a sneaking suspicion those running 40 degree optics might find this out the hard way with some of their corals.

To clarify, I don't think it'll inherently be a BAD reaction, it'll just be different growth forms than the coral might have had otherwise. The only case where I could imagine it being strongly negative would be if you had a very large, slow-growing colony in a tank with very poor reflectors (very diffuse light) and suddenly switched to an LED rig with tight optics.

 

[F35-Joint Strike Fighter]

Can you post a link or datasheet for the power supply you're trying to use?

It's Vicor.com

http://cdn.vicorpower.com/documents/datasheets/ds_flatpac.pdf

The model I will be using is dual output 5 and 28VDC

 

[der_wille_zur_macht]

It's Vicor.com

http://cdn.vicorpower.com/documents/datasheets/ds_flatpac.pdf

The model I will be using is dual output 5 and 28VDC

OK, maybe I'm missing something, but those look like (robust) commonplace fixed-voltage DC power supplies. I see where they have connections to trim the output voltage. But I don't see where they have any option for running in a constant current mode, which you'd need to run LEDs from them without some external current-limiting circuit (i.e. a buckpuck or other dc-dc constant current driver).

 

[kcress]

It's Vicor.com

http://cdn.vicorpower.com/documents/datasheets/ds_flatpac.pdf

The model I will be using is dual output 5 and 28VDC

This is not going to turn out well.. :hmm4:

 

[F35-Joint Strike Fighter]

OK, maybe I'm missing something, but those look like (robust) commonplace fixed-voltage DC power supplies. I see where they have connections to trim the output voltage. But I don't see where they have any option for running in a constant current mode, which you'd need to run LEDs from them without some external current-limiting circuit (i.e. a buckpuck or other dc-dc constant current driver).

It's fixed-voltage and autoranging current output.

it says: Vicor's FlatPAC is also available with a current controlled output using BatMod
converter modules of 12, 24, or 48 Vdc outputs

 

[TheFishMan65]

Speaking of the light angle issue and yes a little of topic (is there a better forum). I always wondered if current lighting was some what wrong anyway. Current lighting tends to flood the tank. In real life the sun rises in the east and slowly rises. So one side of the coral would get light, but at what point does it become useful. When it first rises I imagine the light is going through so much water that very little is useful to the coral. Then it sets with everything in reverse. So how much light does the north and south side get. So I pose the question might we get more realistic growth with LEDs directing the light straight down?

Or is all this wrong since the light is bounced of stuff in the water.