has anyone tried this for gluing stars to heatsinks? http://www.dealextreme.com/details.dx/sku.4579
Anyone have opinions on it?
Anyone have opinions on it?
DIY LEDs - The write-up
- Thread starter JohnL
- Start date
- Status
Reef_Matrix29
Water Lover
HndMafia, plenty of people have tried and been successful at gluing stars to heatsinks in contrast to screw and paste, it's up to you, but note that once they are on there....you can't remove them either at all or without possibly using some kind of solvent to dissolve the chemical bond...so plan wisely before you apply the thermal glue!!
marspeed
Member
FUJIK Silicone Thermal Glue interesting
der_wille_zur_macht
Team RC
To be clear, I've seen several people using thermal "epoxy" adhesives, I've never seen or heard of that "silicone" adhesive. I'd be interested to see a datasheet to get relevant information (thermal coefficient, expected lifetime, etc.) before using it.
der_wille_zur_macht
Team RC
Thanks. Behind the tank is a 1.5" wooden framebox with translucent blue acrylic covering it:
You are my HERO.
That looks like something out of an Amano-style planted FW tank. It's about time us marine hobbyists started doing interesting things with backgrounds.
Can we estimate PAR?
This spot is approximately 14" wide, just a bit more than 1 square foot.
It's made by 3 Cree XP-G 6500K and 3 Cree XP-E Royal Blue with optics running at 1000ma. Just considering the whites there's about 1050 lumens there. If the optics get 80% (?) into the spot there are around 800 lumens there + whatever the blues contribute.
This (http://www.thekrib.com/Plants/Tech/par-moles.html) seems to suggest that PAR is close to 0.02 x lux. So:
800 lumens/ft = 8000 lux = 160 PAR.
Seem reasonable, Can we guestimate PAR like this? How far off am I likely to be and am I under or over?
This spot is approximately 14" wide, just a bit more than 1 square foot.
It's made by 3 Cree XP-G 6500K and 3 Cree XP-E Royal Blue with optics running at 1000ma. Just considering the whites there's about 1050 lumens there. If the optics get 80% (?) into the spot there are around 800 lumens there + whatever the blues contribute.
This (http://www.thekrib.com/Plants/Tech/par-moles.html) seems to suggest that PAR is close to 0.02 x lux. So:
800 lumens/ft = 8000 lux = 160 PAR.
Seem reasonable, Can we guestimate PAR like this? How far off am I likely to be and am I under or over?
ReefEnabler
Premium Member
schnitm, very nice!
would be tricky to guess PAR numbers, but I think with a small tank you could easily be hitting the 300-400 par range in parts of the tank if you position that right. could be way off, but that's a guess based on what I saw from 4 XR-Es over my wifes tank at one point.
I'm sure if you measure in the air a few inches from that rig you'll see numbers near 2000 etc. but not really meaningful intank measurements.
would be tricky to guess PAR numbers, but I think with a small tank you could easily be hitting the 300-400 par range in parts of the tank if you position that right. could be way off, but that's a guess based on what I saw from 4 XR-Es over my wifes tank at one point.
I'm sure if you measure in the air a few inches from that rig you'll see numbers near 2000 etc. but not really meaningful intank measurements.
widmer
Drug Enthusiast
In my thread, I have a shot of the beam cast on the ceiling as well, and in the picture it looks just like the beam he's posted, but in real life there's not an absolutely perfect blend of the b/w beams. At least in my case, the picture is a minor misrepresentation...
I agree with that. Close but not perfect blue white mix.
Also, the "halo" appears blue. That's OK. I want it that way.
ReefEnabler
Premium Member
It's funny, after running the LEDs for a few hours, a bit of the smell did build up in my canopy. Wasn't that bad, to me it was "new computer smell" with a bit of an ozone smell. Is there a chance it was just a byproduct of running 200w of bran new LEDs? Didn't smell like anything was burning. heatsinks were barely warm to the touch.
I'm not really sure how to check the junction temp of my LEDs to make sure they're have a good thermal path to the heatsink. I have an IR thermometer but it isnt giving me a good reading. For example it tells me the heatsink itself is ~70 F when I can tell its around 105F. And I can't really get a reading of the star PCBs with the optics in the way. If I take a reading pointed at the optic, it goes way over 200F, which I'm assuming is due to emissivity.
maybe I need a temp sensor with a physical probe I can slip inside the optics next to the dies on the pcb.
Anyways the smell didn't bug her after running them a few hours last night, so I'm wondering if its something related to 200w of bran new LEDs maybe cooking off some kind of residue.
dnahas
Premium Member
Back to some wiring questions.
In non-electrical engineer terms, I guess rough %, what is the loss in efficiency if the drivers are mounted below the system and connected through an approximate 6' of wire opposed to in the canopy with an approximate 1' of wire?
Do we have an established "œbest practice" for the LED wiring in terms of series, parallel, or combination series and parallel?
In non-electrical engineer terms, I guess rough %, what is the loss in efficiency if the drivers are mounted below the system and connected through an approximate 6' of wire opposed to in the canopy with an approximate 1' of wire?
Do we have an established "œbest practice" for the LED wiring in terms of series, parallel, or combination series and parallel?
That's what worries me. I'm planning on 6 XP-Gs and 12 XR-E RBs in a spot the size in my picture. I think I want 300+ PAR but I don't really want 800 PAR do I?
I know, buy a meter.
der_wille_zur_macht
Team RC
dnahas, if you wanted to be precise, you could calculate the difference in efficiency, but it's not going to be much if you're using proper wire sizes. The bigger worry with longer cables is causing interference in the driver's operation, which is hard to measure or predict.
Wiring in series is, by a vast margin, the most common practice, but I think the "best" practice would depend on your personal criteria and the exact components you had to work with.
Wiring in series is, by a vast margin, the most common practice, but I think the "best" practice would depend on your personal criteria and the exact components you had to work with.
der_wille_zur_macht
Team RC
Most mixed reef tanks probably vary between 100 and 400 PAR.
AND dimmable drivers, so you can turn things down if it's too much. Actually, as fun as it is to know the number, IMHO it's not really important. We know you'll be in the correct range from a very rough perspective, so just do it and based on feedback from your corals, adjust the drive current if required.
TheFishMan65
New member
dnahas,
It won't matter about wiring. 6' will have a little more resistance (so a larger voltage drop across the wire), but the current will be the same (constant current source). If you are using a reasonable guage wire the drop on the wire in negligible. Oh you wanted simple - doesn't matter. IIRC, some drivers recommend an added capacitor.
I think everyone is doing series one string right now. I would avoid parallel for all rhe reason listed. My second choice would be the matrix (serial and parallel). With the matrix I would pick some safety factor (how many LEDs can you loose before it goes catastrophic and loose them all). I wrote something on that a while back (if you can't find it let me know and I will try and repeat it).
dwzm - types faster than I do. But now you can have 4 cents.
It won't matter about wiring. 6' will have a little more resistance (so a larger voltage drop across the wire), but the current will be the same (constant current source). If you are using a reasonable guage wire the drop on the wire in negligible. Oh you wanted simple - doesn't matter. IIRC, some drivers recommend an added capacitor.
I think everyone is doing series one string right now. I would avoid parallel for all rhe reason listed. My second choice would be the matrix (serial and parallel). With the matrix I would pick some safety factor (how many LEDs can you loose before it goes catastrophic and loose them all). I wrote something on that a while back (if you can't find it let me know and I will try and repeat it).
dwzm - types faster than I do. But now you can have 4 cents.
Obi-dad
Active member
Yes, I have some XP-Gs that are rated at 1.5A, so I wanted my 'play around with' buckpuck to be able to go to 1A.
- Status
