DIY LEDs - The write-up

[hllywd]

shizno,

Use your continuity function on the star pads, it will show you how they are connected.

Personally I'd try the iron you have before upgrading. If it doesn't do a suitable job then upgrade. There are plenty of opinions on irons that run contrary to one another. In my observation of this and other related threads it seems what works well for one may not be as good for another. My guess is you'll find your 30W isn't enough, if it does work, you won't be the first. It sounds like you have enough experience to recognize the difference.

BTW, you didn't mention flux. I learned at the OIT in Columbus, and as I remember it they taught to strictly use rosin core solder. I bota tin of flux because I was having som difficulty initially with the pads. A tiny bit of flux on each connection is the way to go.

 

[hllywd]

Also, if you're using a DVM it should have a diode test function that will cause the LEDs to glow slightly when the probes are touched to the pads on the star. No need for the AAs.

 

[der_wille_zur_macht]

Okay, so considering your thoughts on these...would you suggest the use of 2 or so of the 150-48 versions if I'm looking at doing a 290g with about 144 LED's?

That would be reasonable, but remember we're pretty much guessing at this point, since the devices have not been released and we don't have any real pricing info to know if it's a reasonable option or not.

Plus, while using a smaller number of higher-powered drivers has some advantages, it also has some potential disadvantages - if you screw something up in the configuration, you're putting a larger number of LEDs at risk. And, your "resolution" becomes poor - you won't have as many discrete units to dim/control separately.

How careful would I have to be about balancing bins and or keeping the current low? I would like to use XP-G's and XP-E's and push their current up to being useful rather than half.

Longer LED strings are actually more likely to "just balance out" in terms of having equal Vfs but IMHO it's still worth checking. I'd just randomly build your strings, run them, and check the Vfs across each. If they are vastly different, go down the string and check Vfs on individual LEDs and see if there's a simple way to swap one or two from one string to another to balance things out.

It's also worth repeating that different models/colors of LEDs have different Vfs, so you really can't blindly mix a string of XP-G with a string of XR-E on the same driver, for instance.

push their current up to being useful rather than half.

Keep in mind that this sentence depends on your definition of "useful." Any of these LEDs will produce more light in the absolute sense as you provide more current, but there's a diminishing return - doubling the current won't double the light, since efficiency drops off as current gets higher. Also, at higher currents, you'll have more heat to deal with. In theory, a unit built with X LEDs at Y current will be more expensive upfront but cheaper in the long run than a unit built with X/2 LEDs and 2*Y current.

 

[shizno]

hllywd, I tested about 5 or so and none of them light up or give a reading, when I use it on all my regular little LED lights on the x1K (horseshoe symbol) I get a reading across the regular LED, ony x10 I get a smaller reading but actually get a little light. On the Crees I dont get any light and no reading at all across any of the pads as if the circuit isnt complete...on any of them

 

[CJO]

I'd just randomly build your strings, run them, and check the Vfs across each. If they are vastly different, go down the string and check Vfs on individual LEDs and see if there's a simple way to swap one or two from one string to another to balance things out.

That's how I did it and it worked out well.

CJ

 

[Eltonka]

Do you guys think a strip of 6 leds(no lense) at 5.5 inches from the surface too low for a 55gal tank. They will be for adding shimmer to 4 t5ho(54w) hmm that is making wonder if that might also be too low foe the t5's.

 

[shizno]

This is killing me, I can't figure out why i cant get a good test across a Cree star, starting to feel like I wasted a lot of money...they can't all be bad??

 

[der_wille_zur_macht]

Do you guys think a strip of 6 leds(no lense) at 5.5 inches from the surface too low for a 55gal tank. They will be for adding shimmer to 4 t5ho(54w) hmm that is making wonder if that might also be too low foe the t5's.

Define "too low" - are you asking "will they not be bright enough?" or, "will they be too close to the water?"

This is killing me, I can't figure out why i cant get a good test across a Cree star, starting to feel like I wasted a lot of money...they can't all be bad??

Got a modern digital multimeter with a diode test function laying around? I can't see the dial on that thing clearly enough to know how/what you're actually testing. I really doubt all your LEDs are DOA.

Or try the battery test again. Put two AAs in series and connect the "+" of your battery pack to the "+" on the LED star (any of the positive pads), then touch the negative lead from the battery pack to a negative pad on the LED.

 

[supateddy]

It looks like your meter is lighting that LED in your picture. The ohm setting on the meter will not source enough current to make the LED very bright, but if it lit at all it should be fine.

 

[shizno]

der wille, AWESOME, it blinked a few times probably because of the poor connection with my taped down wires but thank you so much...so was that a no go on the clear flux? i know it will work on electronics i have used it before for resistors/mod chips/leds etc..

 

[TheFishMan65]

I had trouble with the battery test. I think it was oxidation on the pad. I had to move the leads around to kind of scratch the surface.

A few things you probably know but...
Make sure you are on a + and a -
If that does not work try reversing the battery leads.
Try other contacts I think I remember reading of a mismarked star
Just for fun try all combination (even +, + just not for very long). Yes with reversed polarity also.

Good luck

 

[shizno]

Yeah finally got it, why did it work with + to + and - to - for the batterys though? Shouldn't it ALWAYS be - to + ?

 

[TheFishMan65]

Sound like the star is marked wrong. With one start check the continuity (resistance) between the pads. If properly marked it should be zero from + to + and also zero from - to -. Sound like you will get zero for a +/- pair and zero for the other pair.

 

[Eltonka]

I meant to closw to the water. Not to sure whats the recommended height from the surface.

 

[der_wille_zur_macht]

Five or so inches is fine as long as you use a splash shield and clean it (of salt spray) regularly.

 

[Eltonka]

Okay thanks.

 

[kcress]

The star "+" is saying "hook plus here to make me work".
It's correct.


Clear flux? You should use rosin core solder and need no additional flux what-so-ever. Adding additional flux can make a mess, leave sticky residue, disagree with the solder's organic flux, or increase the chance of chemical etching or coating the LED's lens.

 

[skippyreef]

DOwnsizing to starphire rimless 48*24*22

DOwnsizing to starphire rimless 48*24*22

I have fixtures that were made for my 270 gallon tank. two of them are to long to be used on the new set up. I used the arctic epoxy to attach. How hard is it to get them off the heat sinks and attach them to the new 24" heat sinks? I hate to order more LEDS when I have a bunch right here.

 

[hllywd]

You should use rosin core solder and need no additional flux what-so-ever.

Until this project I would have agreed completely. That's the way I was taught. I couldn't get the solder to flow onto the pad though even if I roughed it a little. Once I used the flux my joints came together like they should have. I'm going to try a Kester Flux pen on the next 48 I solder unless you have a better solution. My solder is all a few years old, any chance the core goes bad?

I have fixtures that were made for my 270 gallon tank. two of them are to long to be used on the new set up. I used the arctic epoxy to attach. How hard is it to get them off the heat sinks and attach them to the new 24" heat sinks? I hate to order more LEDS when I have a bunch right here.

I tried to go outside the box on my first two units and don't care for the results. When I get time to revamp them, I though maybe a 3/4" wood chisel tapped under one side might pop them off. Otherwise if we devise some way to tap the star sideways the epoxy shouldn't have great shear strength. Those are a couple ideas I had, nothing tested yet. I do have a RB I knocked the lens off of, If I could find one to replace it I'd start there to see if either idea worked. Any body with an extra RB MR-E???

Tim

 

[westmoorenerd]

A chisel and a tap to the side of the stars take them off without too much fuss...

Also I've always been taught that you really can't have too much flux...