DIY LEDs - The write-up
- Thread starter Soundwave
- Start date
- Status
Soundwave
New member
TRASH????
Don't give up, I have cuts and burns and all kinds of stuff from doing things like this.
I almost cut my thumb off building a mini bike.
This one...
But I kept going and finished it.
The LEDs may have blown from heat. I think if you solder the connection for too long with a high power iron, you will melt the internals. I'm thinking that may have happened.
Do yourself a favor and pull it out of the trash. You'll get back on the horse.
Don't give up, I have cuts and burns and all kinds of stuff from doing things like this.
I almost cut my thumb off building a mini bike.
This one...
But I kept going and finished it.
The LEDs may have blown from heat. I think if you solder the connection for too long with a high power iron, you will melt the internals. I'm thinking that may have happened.
Do yourself a favor and pull it out of the trash. You'll get back on the horse.
james3370
Premium Member
yeah, i'm pretty sure i burned them up...i remember seeing 1 of the "frost" over while i was trying to get the wire to set right.
i only trashed the LEDs...kept the buckpucks, aluminum bar & wire & maybe when i have some $$ i'll order real LEDs from ledsupply.com instead of the ebay stuff. it's just not a priority at the moment
what blues & how many would you recommend just for moonlights? i have the 350mA dimable buckpuck & a 20volt/1amp wall wart
i only trashed the LEDs...kept the buckpucks, aluminum bar & wire & maybe when i have some $$ i'll order real LEDs from ledsupply.com instead of the ebay stuff. it's just not a priority at the moment
what blues & how many would you recommend just for moonlights? i have the 350mA dimable buckpuck & a 20volt/1amp wall wart
widmer
Drug Enthusiast
In order to test the LEDs, I would just take an old wall adapter for some random electronics like a cellphone, check to see if its safely within the voltage and amperage range for the LEDs, and touch the wires to the LED terminals. I know the charger for the bluetooth earpiece I lost puts out like 3.5v DC 100 mA.
But yea that might be a secondary benefit of those tabs which are supposed to be soldered to - it brings the heat of soldering a little further away from the die.
But yea that might be a secondary benefit of those tabs which are supposed to be soldered to - it brings the heat of soldering a little further away from the die.
james3370
Premium Member
29 gal....30x12x17
you mean these?? http://www.ledsupply.com/lxhl-mrrd.php
i guess the 20volt/1amp wall wart would be too much for 3-4 of them? or would it be ok since my 350mA buckpuck is dimable??
Last edited:
james3370
Premium Member
ok, just to give it a try, i cut all the strings apart, stripped each end back, laid (2) AAA batteries together & touched each to the respective ends of the batteries......all (10) LEDs fired up individually using this test method
so that means both buckpucks are defective??
james3370
Premium Member
& today's Idiot of the Day Award goes to......ME :wave:
earlier, when i checked the power out of the barrel plug the power supply plugs into w/ a pc fan, i didn't check that the fan ran backwards
apparantly, i misread the schematic on the packaging for the barrel plug as to which tab went to the center pin (which i knew from the wall wart schematic was +), so this whole time, i had the freaking power reversed out of the barrel plug into the LEDs....hence the reason they wouldn't fire up
earlier, when i checked the power out of the barrel plug the power supply plugs into w/ a pc fan, i didn't check that the fan ran backwards
apparantly, i misread the schematic on the packaging for the barrel plug as to which tab went to the center pin (which i knew from the wall wart schematic was +), so this whole time, i had the freaking power reversed out of the barrel plug into the LEDs....hence the reason they wouldn't fire up
ReefEnabler
Premium Member
<a href=showthread.php?s=&postid=14821073#post14821073 target=_blank>Originally posted</a> by james3370
doesn't appear that they spot is connected to where i soldered on the wire as seen in the broken one....2nd pic you can see it is soldered to wire coming off the LED
The Stars have some kind of internal circuitry that connects the positives and negatives of the LED to those solder pads on the edge. The solder pads are nice as Soundwave mentioned because they have a nice glob of solder on them already and are further from the LED. Also more room to work with.
Nothing wrong with what you did, thats how somebody WITHOUT a Star base would solder their LED... but you also might as well have just gotten an LED without the star (except the star also gives you the screw mounts).
Also if you didn't, its a good idea to have the stars mounted onto the heatsink while soldering, since the heatsink will take more heat than the LED just incase you do take too long to solder something. Looks like you had that part though.
I'm glad you found your problem
kcress
New member
widmer; Can you even find any for sale or are those vaporware? The ones you linked to in mouser are certainly not the same thing. They are rather paltry 70Lumen/W, 10W, 3V, DC HBLEDs.
The main problem I see with the AC ones is that they aren't very blue but are warm white, or less warm white, but not cool white. This would mean you still have to come up with some blue and they'll be low voltage.
Do you really want hobbyist line voltage LEDs?
The main problem I see with the AC ones is that they aren't very blue but are warm white, or less warm white, but not cool white. This would mean you still have to come up with some blue and they'll be low voltage.
Do you really want hobbyist line voltage LEDs?
widmer
Drug Enthusiast
I could have sworn that was the 10 watt, 90 lumen/watt LED I sourced.
http://www.mouser.com/search/refine...ium=featured+spotlight&utm_campaign=883797051
They also make pure blue AC LEDs but I don't have time to look those up now.
I'm not afraid of the line voltage I guess, it's the line amperage that kind of has me afraid. But I guess I think about these things purely from the perspective of how I'd like to inevitably have an array mounted on the ceiling with optics projecting it downward.
http://www.mouser.com/search/refine...ium=featured+spotlight&utm_campaign=883797051
They also make pure blue AC LEDs but I don't have time to look those up now.
I'm not afraid of the line voltage I guess, it's the line amperage that kind of has me afraid. But I guess I think about these things purely from the perspective of how I'd like to inevitably have an array mounted on the ceiling with optics projecting it downward.
TropTrea
New member
My big fear on a project like this right now is being able to get a balanced color spectrum.
Most of the white LED's I foiound are rated as aprox. 6,500K or lower. Some actualy give light spectrums but the all seem to lacking light in the 490nm range.
Now typicaly we use 10,000K bulbs or hiogher supmemented with atinic lighting. There are some noce 420nm LED's out there but not a lot to boast that 490nm Range.
The thing I keep thinking is from last studies I did led's produced a narow light spectrum. Now the different corals require different parts of the spectrum, and my fear is that some of there requirements in that 490 nm range especialy would be missing.
Generaly speaking with some experimentation I think this should work and be a very econoimical system to operate. But only time wil tell how the corals grow and color up under these systems.
Has anyone been able to plot out the actual frequency range of these LEDS rather then depending upon manufacturers data?
Dennis
Most of the white LED's I foiound are rated as aprox. 6,500K or lower. Some actualy give light spectrums but the all seem to lacking light in the 490nm range.
Now typicaly we use 10,000K bulbs or hiogher supmemented with atinic lighting. There are some noce 420nm LED's out there but not a lot to boast that 490nm Range.
The thing I keep thinking is from last studies I did led's produced a narow light spectrum. Now the different corals require different parts of the spectrum, and my fear is that some of there requirements in that 490 nm range especialy would be missing.
Generaly speaking with some experimentation I think this should work and be a very econoimical system to operate. But only time wil tell how the corals grow and color up under these systems.
Has anyone been able to plot out the actual frequency range of these LEDS rather then depending upon manufacturers data?
Dennis
Yossarian99
New member
Well, I tried my own build over the long weekend, and am quite happy with the result. A few notes of what I did differently:
- I went with all white LEDs. I wanted to get maximum light out of the setup, and I will continue to run some Actinic PCs. I'm using the Q5 Cree 107 lumen LEDs.
- I went with a single 24 V, 12.5A power supply.
- I wired it up so I could easily replace 1/2 of the white LEDs with blues if I wanted to change it up later.
- I built a driver module using 8 1000mA dimmable buckpucks, and used trim potentiometers to dial the current back to 750mA (as suggested earlier in the thread to lengthen the life of the bulbs).
- I added external pots so I can manually dim the lights.
- The driver module I built is split for two outputs (again so I can run blue/white later if I like). Each half is independently dimmable with a 5KOhm trim pot for maximum output, and a regular 5KOhm pot for manual adjustment. Each half runs 4 of the buckpucks.
- I used molex connectors between the LED strings and the driver circuit, enabling quick disconnect/reconnect for troubleshooting, configuration changes, etc. I highly recommend this step! The connectors are cheap, and the crimping tool isn't that expensive, either.
- I had originally planned to incorporate the LEDs mounted on the heatsinks into my coralife hood (taking out the MH light), but they wouldn't fit with the fans, and I would have also needed to run the power supply external. I found an old 48" PC hood from a freshwater tank I ran before, and it had plenty of room for all my electronics, plus a switch and power cord. So I mounted it all in there. I'm going to get an acrylic sheet to protect the internals from moisture, splashes, etc.
The end result: AWESOME! Since my coralife hood is still in pieces, I'm running with only the LEDs, and the light output is much higher. (as read by my photographic light meter - so no PAR or lumens, I'm afraid). I was running 2x96W 50/50 PCs, and a 150W MH. Granted, the MH is almost a year old, so is lower on its light output, but it's still impressive that I can put out more lights with about 1/4 the power.
Just by looking at the quality of the light, I'd estimate that it's not quite 10K, as advertised as the colour temperature in the CREE documentation, but it's close. I'll get some of the blue colour back when I get my 50/50 PCs back up and running.
I'm running the LEDs at 75% maximum power, and the light output is still fantastic. In fact, when dialed up to 1000mA, the difference is barely noticeable, which is borne out by the power curves in the data sheets for the LEDs. So I think it's definitely worth it to get a couple extra years of life by sacrificing an almost imperceptible amount of light output.
I'm also disappointed that the buckpucks do not have 1-10v dimming input. Instead it's a 1-5v, and to make it worse, the low end (1v) is maximum brightness and the high end (5v) is minimum. I still hope to be able to find a way of converting that (some kind of logic circuit?) to 1-10v, then I can get an ALC for my reefkeeper and have the dimming controllable by time of day.
I have some photos and extensive notes for the build, so when I have the time, I plan to fully document what I did. I don't have that time now, but I'd be interested to know what kind of interest there is - whether I should make the effort. In many ways, it's substantially similar to Soundwave's design anyways.
Thanks to Soundwave and all others who contributed such useful information to this thread. I would recommend it to anyone who has a reasonable amount of confidence with electronics projects!
- I went with all white LEDs. I wanted to get maximum light out of the setup, and I will continue to run some Actinic PCs. I'm using the Q5 Cree 107 lumen LEDs.
- I went with a single 24 V, 12.5A power supply.
- I wired it up so I could easily replace 1/2 of the white LEDs with blues if I wanted to change it up later.
- I built a driver module using 8 1000mA dimmable buckpucks, and used trim potentiometers to dial the current back to 750mA (as suggested earlier in the thread to lengthen the life of the bulbs).
- I added external pots so I can manually dim the lights.
- The driver module I built is split for two outputs (again so I can run blue/white later if I like). Each half is independently dimmable with a 5KOhm trim pot for maximum output, and a regular 5KOhm pot for manual adjustment. Each half runs 4 of the buckpucks.
- I used molex connectors between the LED strings and the driver circuit, enabling quick disconnect/reconnect for troubleshooting, configuration changes, etc. I highly recommend this step! The connectors are cheap, and the crimping tool isn't that expensive, either.
- I had originally planned to incorporate the LEDs mounted on the heatsinks into my coralife hood (taking out the MH light), but they wouldn't fit with the fans, and I would have also needed to run the power supply external. I found an old 48" PC hood from a freshwater tank I ran before, and it had plenty of room for all my electronics, plus a switch and power cord. So I mounted it all in there. I'm going to get an acrylic sheet to protect the internals from moisture, splashes, etc.
The end result: AWESOME! Since my coralife hood is still in pieces, I'm running with only the LEDs, and the light output is much higher. (as read by my photographic light meter - so no PAR or lumens, I'm afraid). I was running 2x96W 50/50 PCs, and a 150W MH. Granted, the MH is almost a year old, so is lower on its light output, but it's still impressive that I can put out more lights with about 1/4 the power.
Just by looking at the quality of the light, I'd estimate that it's not quite 10K, as advertised as the colour temperature in the CREE documentation, but it's close. I'll get some of the blue colour back when I get my 50/50 PCs back up and running.
I'm running the LEDs at 75% maximum power, and the light output is still fantastic. In fact, when dialed up to 1000mA, the difference is barely noticeable, which is borne out by the power curves in the data sheets for the LEDs. So I think it's definitely worth it to get a couple extra years of life by sacrificing an almost imperceptible amount of light output.
I'm also disappointed that the buckpucks do not have 1-10v dimming input. Instead it's a 1-5v, and to make it worse, the low end (1v) is maximum brightness and the high end (5v) is minimum. I still hope to be able to find a way of converting that (some kind of logic circuit?) to 1-10v, then I can get an ALC for my reefkeeper and have the dimming controllable by time of day.
I have some photos and extensive notes for the build, so when I have the time, I plan to fully document what I did. I don't have that time now, but I'd be interested to know what kind of interest there is - whether I should make the effort. In many ways, it's substantially similar to Soundwave's design anyways.
Thanks to Soundwave and all others who contributed such useful information to this thread. I would recommend it to anyone who has a reasonable amount of confidence with electronics projects!
Yossarian99
New member
Regarding the soldering questions above, I found a technique during my build that worked quite well.
First, I made sure I pre-tinned all my wire ends before soldering. Tedious, I know, but it makes a big difference.
Second, I used my iron to place a little daub of solder onto the pads of the LEDs. Just going from one to the next and it only takes a couple seconds of contact with a decent iron.
Then, I just placed the wire on the solder daubs, touched the iron to the wire, and within 1-2 seconds, the solder is molten again and the connection is solid.
This should help avoid problems with cold solder joints, too - a failure that may only become apparent after the unit has been running for a while.
First, I made sure I pre-tinned all my wire ends before soldering. Tedious, I know, but it makes a big difference.
Second, I used my iron to place a little daub of solder onto the pads of the LEDs. Just going from one to the next and it only takes a couple seconds of contact with a decent iron.
Then, I just placed the wire on the solder daubs, touched the iron to the wire, and within 1-2 seconds, the solder is molten again and the connection is solid.
This should help avoid problems with cold solder joints, too - a failure that may only become apparent after the unit has been running for a while.
- Status
Similar threads
