karimwassef
Active member
theatrus, was your LED white or blue?
The picture shows blue, but I don't know if that's the one you collected data for?
The picture shows blue, but I don't know if that's the one you collected data for?
Looking for a specific DC CC driver for LEDs
- Thread starter karimwassef
- Start date
karimwassef
Active member
Here are my conclusions so far (might change as I collect more data):
- Inferior parts exhibit a higher voltage at a given current due to a higher resistance and intercept V. Intercept V is like a "fixed" loss (P = I x V) while resistance gets much worse with current (P = I2 x R).
- Average parts were 1.2-2.4 Ohms in resistance & 26.5-29.5V in intercept voltage. The variance may be a function of the LED color?
- The bad LED was > 4.5 Ohms and >30V intercept.
- Cheaper $5 was inferior to the $10, but in the $10-$12, performance was comparable.
- White and theatrus were 2.4 Ohms, both blues were 1.2 Ohms.
- Inferior parts exhibit a higher voltage at a given current due to a higher resistance and intercept V. Intercept V is like a "fixed" loss (P = I x V) while resistance gets much worse with current (P = I2 x R).
- Average parts were 1.2-2.4 Ohms in resistance & 26.5-29.5V in intercept voltage. The variance may be a function of the LED color?
- The bad LED was > 4.5 Ohms and >30V intercept.
- Cheaper $5 was inferior to the $10, but in the $10-$12, performance was comparable.
- White and theatrus were 2.4 Ohms, both blues were 1.2 Ohms.
theatrus
100-mile-commuter
Mine was a Royal Blue.
What PAR measurement standard do we want? 12" center? I can use both an SQ-120 and SQ-500 from Apogee.
(I realize we're not trying to make an absolute measurement for PAR, but it would be nice if we could get closer - it can vary quite a bit for LEDs: http://www.reefcentral.com/forums/showpost.php?p=24378331&postcount=28 )
What PAR measurement standard do we want? 12" center? I can use both an SQ-120 and SQ-500 from Apogee.
(I realize we're not trying to make an absolute measurement for PAR, but it would be nice if we could get closer - it can vary quite a bit for LEDs: http://www.reefcentral.com/forums/showpost.php?p=24378331&postcount=28 )
karimwassef
Active member
Yes. I was thinking 12" center / no lens. I'm using the Apex PMK.
Royal Blue - so no correlation between color and the resistance.
I'm thinking of monitoring my LED current and voltage on an ongoing basis. The idea being that degradation will show an increasing voltage at the same current. I'm designing my fixture to easily replace LED chips (4 screws and a power clip) so it would be useful to have a gauge or basis for performance.
Royal Blue - so no correlation between color and the resistance.
I'm thinking of monitoring my LED current and voltage on an ongoing basis. The idea being that degradation will show an increasing voltage at the same current. I'm designing my fixture to easily replace LED chips (4 screws and a power clip) so it would be useful to have a gauge or basis for performance.
karimwassef
Active member
theatrus - I used your converter on the bad chip. I didn't know what to expect since I think you had it set to 3.0A? But since I had the input voltage limited with a 43V DC input source, it basically ran up to ~100% duty cycle with 42.9V. The current that can pass under that voltage was ~1.62A. So the chip was "self limiting" as long as the input voltage wasn't allowed to go above 43V.
Tomorrow, I'll push it up to 50Vin which should allow it to run to 48Vout and ~2.3A.
Observations... while I didn't run it hot (only went to 70W), the Sure Electronics converter was hot but touchable. Yours wasn't even warm. I think it's the massive toroid inductor you're using that's 2x the cross-section of the Sure inductor.
How expensive was it?
Tomorrow, I'll push it up to 50Vin which should allow it to run to 48Vout and ~2.3A.
Observations... while I didn't run it hot (only went to 70W), the Sure Electronics converter was hot but touchable. Yours wasn't even warm. I think it's the massive toroid inductor you're using that's 2x the cross-section of the Sure inductor.
How expensive was it?
theatrus
100-mile-commuter
Sounds about right. The converter is in dropout which means the switch is just held on and current is monitored.
About $1.50 @ Qty100. Or $2.30 in singles.
http://www.digikey.com/product-detail/en/bourns-inc/PM2110-470K-RC/M8764-ND/775303
karimwassef
Active member
It's a good inductor. Your solution is smaller too, about 2/3?
How do you mount the SMD parts? Hot air gun?
Do you have a hot plate (or hot skillet) to reflow? What solder paste do you use?
Going back to the question of CC boost vs buck...
I think that the buck is inherently safer here. The fact that an LED that is underperforming will demonstrate lower current at elevated voltage means that it is self limiting in terms of catastrophic failure. The light output will diminish, but a voltage and current sensor will capture that event (to be proven). I still prefer that kind of slow deterioration failure that I can spot and address vs. a Boost at CC of 3A that would drive up to 50V and drive a runaway thermal condition that ends in a puff of smoke (and hopefully nothing else).
On a side note- I was talking to a local reefer about this and I wonder if the whole debate of LED vs. MH isn't colored by the fact that these fixtures don't come with a PAR meter and assuming that LED wattage relates to PAR USA function of the chip's real performance -- both upfront and down the road with thermal management...
It may be that poor LED performance is a result of poor DIY design or other low cost "OEM" solutions that deliver poor light performance while still consuming the same (or more) power.
The question isn't "why is LED not working?"... It's more specifically "why is YOUR LED not working for you?"
In comparison, MH is mature. You buy the bulb and it works.
Having said that, I think that this investigation has some real potential in a smarter power supply for multichip LEDs. If it had a PMBUS or I2C interface, it could provide current and voltage data for continuous monitoring and alerts.
Combine that with an easy to replace bulb assembly and an I2C PAR meter... And you have a robust solution for LED, DIY or otherwise!
Thoughts?
How do you mount the SMD parts? Hot air gun?
Do you have a hot plate (or hot skillet) to reflow? What solder paste do you use?
Going back to the question of CC boost vs buck...
I think that the buck is inherently safer here. The fact that an LED that is underperforming will demonstrate lower current at elevated voltage means that it is self limiting in terms of catastrophic failure. The light output will diminish, but a voltage and current sensor will capture that event (to be proven). I still prefer that kind of slow deterioration failure that I can spot and address vs. a Boost at CC of 3A that would drive up to 50V and drive a runaway thermal condition that ends in a puff of smoke (and hopefully nothing else).
On a side note- I was talking to a local reefer about this and I wonder if the whole debate of LED vs. MH isn't colored by the fact that these fixtures don't come with a PAR meter and assuming that LED wattage relates to PAR USA function of the chip's real performance -- both upfront and down the road with thermal management...
It may be that poor LED performance is a result of poor DIY design or other low cost "OEM" solutions that deliver poor light performance while still consuming the same (or more) power.
The question isn't "why is LED not working?"... It's more specifically "why is YOUR LED not working for you?"
In comparison, MH is mature. You buy the bulb and it works.
Having said that, I think that this investigation has some real potential in a smarter power supply for multichip LEDs. If it had a PMBUS or I2C interface, it could provide current and voltage data for continuous monitoring and alerts.
Combine that with an easy to replace bulb assembly and an I2C PAR meter... And you have a robust solution for LED, DIY or otherwise!
Thoughts?
theatrus
100-mile-commuter
"Upgraded" toaster oven. Paste is either Kester EM907 (SC305 lead free) or EP256HA (lead, easier to work with). Using Kapton cut stencils.
I love monitoring
And I agree, a lot of reefing is akin to "put X over/in tank, hope it works". At least we're generally ok with maintaining basics, but there are a ton of things we are not controlling.
LEDs, especially DIY, is basically guessed by forum-thread-averaging ("I want 12 RB to 3 white"), and number of emitters. The trend to slam optics on all the LEDs also makes the problem worse, as they're notorious for hot-spots (but it does cut down costs).
Some of that knowledge is tricking into commercial fixtures (which are of course very price conscious - they need to pay the bills after all). PAR meters are still pretty expensive - I've looked at trying to find a good source for a PAR/PUR filtered photodiode (perhaps Hamamatsu) but don't have enough optical engineering experience to make and calibrate an assembly for this. Not having a PAR meter
As for measuring power, +1 to that. Even an inline I2C DC high current columb counter would be a great idea if the supply isn't managed (nearly guarantee no one runs one of those
). It would be nice to know Vf of each chain as well, but thats going to add some cost. I may be able to find enough pins on my "smart driver" concept controller (a super low cost PIC16F18323) to sense that.
karimwassef
Active member
Actually, I have a $6 75mOhm 100A shunt (overkill, I know) that I connect to that little $6 current and voltage meter. in theory, since we're reading mV for current and up to 50V for voltage, you can use an Arduino with some switches to create a ~multiplex read back. Basically, you use the same two ports, but query each voltage and current per chip in sequence. It reduces data acquisition rate, but who cares?? Just two Arduino ports and a $6 shunt for each 100W LED chip...
That seems feasible. Actually, I have 5 of those shunts and an available Mega to play with
As far as the PAR meter, I was think of oreo's thread where you're use the SQ-120 with a multimeter.. Convert to Arduino and you now have an external light reading.
If we can confirm that PAR is a function of current, not voltage, then this would allow an ongoing measure of light efficiency.
If we put all these ideas together...
That seems feasible. Actually, I have 5 of those shunts and an available Mega to play with
As far as the PAR meter, I was think of oreo's thread where you're use the SQ-120 with a multimeter.. Convert to Arduino and you now have an external light reading.
If we can confirm that PAR is a function of current, not voltage, then this would allow an ongoing measure of light efficiency.
If we put all these ideas together...
karimwassef
Active member
The little meter probably has a low pass RC filter to get average DC out of a switched voltage waveform, but I think Arduino code can do the integration just as well.
karimwassef
Active member
True. My architecture is a product of needing to deliver ~2000W worth of LED.
Actually I'm now looking more at ~1.6KW. The distribution losses, even at 48V are not insignificant. I'm probably going to run separate 34V lines to each LED though.
That's two fixtures at 23A 34V each!
Actually I'm now looking more at ~1.6KW. The distribution losses, even at 48V are not insignificant. I'm probably going to run separate 34V lines to each LED though.
That's two fixtures at 23A 34V each!
I love the thought of using voltage to monitor the output of the LEDs, never read of anyone implemnting that but it sure seems that it has some merit for judging the life of the array.
Karminwassef, I saw on one of theatrus' other threads that he's got a nice 16 up luxeon z board in the works, might be worth considering for your purposes, I think they are avialable in the wavelenghts you're after and despite the increased up front cost, the efficiency gain might be worth looking into, along with being able to mix more than one wavelength on the board, fewer emmiters, longer life span, less electicity, etc. with as many as you'd need you could I suspect justify the cost of fabrication compared to the high cost of the multichips your looking at.....at least for future consideration anyhow, last i looked the luxeon z violets were around $20 each however you'd be looking at buying a whole reel so that should push cost down into the realm of competitiveness?
Just a thought.
Great discussion guys, wish I had more to contribute to the electrical engineering side of this endeavor.
Karminwassef, I saw on one of theatrus' other threads that he's got a nice 16 up luxeon z board in the works, might be worth considering for your purposes, I think they are avialable in the wavelenghts you're after and despite the increased up front cost, the efficiency gain might be worth looking into, along with being able to mix more than one wavelength on the board, fewer emmiters, longer life span, less electicity, etc. with as many as you'd need you could I suspect justify the cost of fabrication compared to the high cost of the multichips your looking at.....at least for future consideration anyhow, last i looked the luxeon z violets were around $20 each however you'd be looking at buying a whole reel so that should push cost down into the realm of competitiveness?
Just a thought.
Great discussion guys, wish I had more to contribute to the electrical engineering side of this endeavor.
theatrus
100-mile-commuter
Reading the Vf of an array is relatively trivial if you know the rough range to expect. Super wide range becomes a bit more limiting in resolution.
As for the Luxeon UV emitters, they will work nicely. They're most likely a process generation or two ahead of the Chinese multichips, and give you a ton of variety of frequencies. It would be great to figure out a radiated power for the multichips to compare to the LHUV parts.
Cheapest one-off-parts distributor I know of for Lumileds is Future. A mid-bin 405-410nm part is about $13 in Qty 1: http://www.futureelectronics.com/en...rs-rev/Pages/8038749-LHUV-0400-0450.aspx?IM=0
Packing 16 into my Acrostar MegaZ would run you > $200 (for 50 watts in). The upside if you can blend 420nm, 385nm, etc parts as you see fit, and the efficiency is likely better.
As for the Luxeon UV emitters, they will work nicely. They're most likely a process generation or two ahead of the Chinese multichips, and give you a ton of variety of frequencies. It would be great to figure out a radiated power for the multichips to compare to the LHUV parts.
Cheapest one-off-parts distributor I know of for Lumileds is Future. A mid-bin 405-410nm part is about $13 in Qty 1: http://www.futureelectronics.com/en...rs-rev/Pages/8038749-LHUV-0400-0450.aspx?IM=0
Packing 16 into my Acrostar MegaZ would run you > $200 (for 50 watts in). The upside if you can blend 420nm, 385nm, etc parts as you see fit, and the efficiency is likely better.
karimwassef
Active member
So.. my biggest motivation for using the multichips is physical density.
I also don't have a full spectrum need. This is going to go over a solar-lit tank with the intention of supplementing with UV and royal blues. Target frequency is 365nm and 430nm. There may be some cool and warm whites to be able to see at night, but they're not intended for constant use.
<a href="http://s1062.photobucket.com/user/karimwassef/media/2_zpsxp5krsgt.png.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/2_zpsxp5krsgt.png" border="0" alt=" photo 2_zpsxp5krsgt.png"/></a>
Because this is solar-lit and is in a greenhouse setting, any shadowing by the LED light fixture is a major deficiency in design. So the cross-section of the fixture needs to be as thin as possible.
<a href="http://s1062.photobucket.com/user/karimwassef/media/Animation_zpsetmdy30e.gif.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Animation_zpsetmdy30e.gif" border="0" alt=" photo Animation_zpsetmdy30e.gif"/></a>
So, I'm creating two 1" square aluminum tubes that are 8' long and liquid chilled internally. The LED chips will need to sit on the tube lengthwise. You can see then on each side in that rendering, angled at about 45 degrees or so towards the tank.
<a href="http://s1062.photobucket.com/user/karimwassef/media/3_zpssichjr34.png.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/3_zpssichjr34.png" border="0" alt=" photo 3_zpssichjr34.png"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/91CAD15A-3728-4540-9895-29ACBFDB5CCF_zpseps8f5mu.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/91CAD15A-3728-4540-9895-29ACBFDB5CCF_zpseps8f5mu.jpg" border="0" alt=" photo 91CAD15A-3728-4540-9895-29ACBFDB5CCF_zpseps8f5mu.jpg"/></a>
I need about 800W on each 8' tube. When I looked through different design options, starting with single 3W UV stars (~1A @ 3V), there wasn't enough space on an 8' bar to get that many stars on board. I was assuming 20mm stars.
I've done this before, by the way - conventional but high density.
<a href="http://s1062.photobucket.com/user/karimwassef/media/0_zpstupuba7s.png.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/0_zpstupuba7s.png" border="0" alt=" photo 0_zpstupuba7s.png"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/22ADD796-4E3B-4CEA-9456-DC822BAD268D_zpso5axl1ri.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/22ADD796-4E3B-4CEA-9456-DC822BAD268D_zpso5axl1ri.jpg" border="0" alt=" photo 22ADD796-4E3B-4CEA-9456-DC822BAD268D_zpso5axl1ri.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/22C52CA5-7EE2-4EEA-AEFB-2C580A0C4E24_zps0leofpvq.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/22C52CA5-7EE2-4EEA-AEFB-2C580A0C4E24_zps0leofpvq.jpg" border="0" alt=" photo 22C52CA5-7EE2-4EEA-AEFB-2C580A0C4E24_zps0leofpvq.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/4C8C2913-25BC-4D1A-B2BF-27C33C4BB069_zpsmf5nq8px.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/4C8C2913-25BC-4D1A-B2BF-27C33C4BB069_zpsmf5nq8px.jpg" border="0" alt=" photo 4C8C2913-25BC-4D1A-B2BF-27C33C4BB069_zpsmf5nq8px.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/594C59C1-6144-41F0-A2D3-664FF15545AF_zps3vcfbzs8.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/594C59C1-6144-41F0-A2D3-664FF15545AF_zps3vcfbzs8.jpg" border="0" alt=" photo 594C59C1-6144-41F0-A2D3-664FF15545AF_zps3vcfbzs8.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/258FF03D-B8D2-4295-8B51-54C43E341093_zps70nj76l0.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/258FF03D-B8D2-4295-8B51-54C43E341093_zps70nj76l0.jpg" border="0" alt=" photo 258FF03D-B8D2-4295-8B51-54C43E341093_zps70nj76l0.jpg"/></a>
I also don't have a full spectrum need. This is going to go over a solar-lit tank with the intention of supplementing with UV and royal blues. Target frequency is 365nm and 430nm. There may be some cool and warm whites to be able to see at night, but they're not intended for constant use.
<a href="http://s1062.photobucket.com/user/karimwassef/media/2_zpsxp5krsgt.png.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/2_zpsxp5krsgt.png" border="0" alt=" photo 2_zpsxp5krsgt.png"/></a>
Because this is solar-lit and is in a greenhouse setting, any shadowing by the LED light fixture is a major deficiency in design. So the cross-section of the fixture needs to be as thin as possible.
<a href="http://s1062.photobucket.com/user/karimwassef/media/Animation_zpsetmdy30e.gif.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Animation_zpsetmdy30e.gif" border="0" alt=" photo Animation_zpsetmdy30e.gif"/></a>
So, I'm creating two 1" square aluminum tubes that are 8' long and liquid chilled internally. The LED chips will need to sit on the tube lengthwise. You can see then on each side in that rendering, angled at about 45 degrees or so towards the tank.
<a href="http://s1062.photobucket.com/user/karimwassef/media/3_zpssichjr34.png.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/3_zpssichjr34.png" border="0" alt=" photo 3_zpssichjr34.png"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/91CAD15A-3728-4540-9895-29ACBFDB5CCF_zpseps8f5mu.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/91CAD15A-3728-4540-9895-29ACBFDB5CCF_zpseps8f5mu.jpg" border="0" alt=" photo 91CAD15A-3728-4540-9895-29ACBFDB5CCF_zpseps8f5mu.jpg"/></a>
I need about 800W on each 8' tube. When I looked through different design options, starting with single 3W UV stars (~1A @ 3V), there wasn't enough space on an 8' bar to get that many stars on board. I was assuming 20mm stars.
I've done this before, by the way - conventional but high density.
<a href="http://s1062.photobucket.com/user/karimwassef/media/0_zpstupuba7s.png.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/0_zpstupuba7s.png" border="0" alt=" photo 0_zpstupuba7s.png"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/22ADD796-4E3B-4CEA-9456-DC822BAD268D_zpso5axl1ri.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/22ADD796-4E3B-4CEA-9456-DC822BAD268D_zpso5axl1ri.jpg" border="0" alt=" photo 22ADD796-4E3B-4CEA-9456-DC822BAD268D_zpso5axl1ri.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/22C52CA5-7EE2-4EEA-AEFB-2C580A0C4E24_zps0leofpvq.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/22C52CA5-7EE2-4EEA-AEFB-2C580A0C4E24_zps0leofpvq.jpg" border="0" alt=" photo 22C52CA5-7EE2-4EEA-AEFB-2C580A0C4E24_zps0leofpvq.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/4C8C2913-25BC-4D1A-B2BF-27C33C4BB069_zpsmf5nq8px.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/4C8C2913-25BC-4D1A-B2BF-27C33C4BB069_zpsmf5nq8px.jpg" border="0" alt=" photo 4C8C2913-25BC-4D1A-B2BF-27C33C4BB069_zpsmf5nq8px.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/594C59C1-6144-41F0-A2D3-664FF15545AF_zps3vcfbzs8.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/594C59C1-6144-41F0-A2D3-664FF15545AF_zps3vcfbzs8.jpg" border="0" alt=" photo 594C59C1-6144-41F0-A2D3-664FF15545AF_zps3vcfbzs8.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/258FF03D-B8D2-4295-8B51-54C43E341093_zps70nj76l0.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/258FF03D-B8D2-4295-8B51-54C43E341093_zps70nj76l0.jpg" border="0" alt=" photo 258FF03D-B8D2-4295-8B51-54C43E341093_zps70nj76l0.jpg"/></a>
karimwassef
Active member
The last generation was air cooled
<a href="http://s1062.photobucket.com/user/karimwassef/media/reef9_zps981db009.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/reef9_zps981db009.jpg" border="0" alt=" photo reef9_zps981db009.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/54FC0BFA-DD45-46AA-85B4-0D65EAD3AE1F_zpsivgvyotl.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/54FC0BFA-DD45-46AA-85B4-0D65EAD3AE1F_zpsivgvyotl.jpg" border="0" alt=" photo 54FC0BFA-DD45-46AA-85B4-0D65EAD3AE1F_zpsivgvyotl.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/F1CBE24B-891C-4EC3-A8C8-9236086A8601_zpsvrqifpxa.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/F1CBE24B-891C-4EC3-A8C8-9236086A8601_zpsvrqifpxa.jpg" border="0" alt=" photo F1CBE24B-891C-4EC3-A8C8-9236086A8601_zpsvrqifpxa.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/AA01D0DB-9EE2-4117-8803-1766CFCA9D94_zpsja6sjenk.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/AA01D0DB-9EE2-4117-8803-1766CFCA9D94_zpsja6sjenk.jpg" border="0" alt=" photo AA01D0DB-9EE2-4117-8803-1766CFCA9D94_zpsja6sjenk.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/BDF1797A-303A-4314-A8B6-114AE2C1FE96_zps6so54qja.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/BDF1797A-303A-4314-A8B6-114AE2C1FE96_zps6so54qja.jpg" border="0" alt=" photo BDF1797A-303A-4314-A8B6-114AE2C1FE96_zps6so54qja.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/73D44514-244D-4353-9BF6-E8E7A4110C16_zps9dvqnibq.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/73D44514-244D-4353-9BF6-E8E7A4110C16_zps9dvqnibq.jpg" border="0" alt=" photo 73D44514-244D-4353-9BF6-E8E7A4110C16_zps9dvqnibq.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/827217E2-6DED-4C75-A2AF-C25004C1B927_zpsiw3zjfgm.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/827217E2-6DED-4C75-A2AF-C25004C1B927_zpsiw3zjfgm.jpg" border="0" alt=" photo 827217E2-6DED-4C75-A2AF-C25004C1B927_zpsiw3zjfgm.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/5FF9D214-1FFB-4436-A47B-4A903C0257FA_zpsxyl7g8yv.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/5FF9D214-1FFB-4436-A47B-4A903C0257FA_zpsxyl7g8yv.jpg" border="0" alt=" photo 5FF9D214-1FFB-4436-A47B-4A903C0257FA_zpsxyl7g8yv.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/6102AF94-4DF2-4CBE-94D6-4A9F0788954F_zpsyoqnli6u.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/6102AF94-4DF2-4CBE-94D6-4A9F0788954F_zpsyoqnli6u.jpg" border="0" alt=" photo 6102AF94-4DF2-4CBE-94D6-4A9F0788954F_zpsyoqnli6u.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/96CBB6B3-4409-44AB-AF15-9EF45FB00CF4_zpsh06uwsbk.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/96CBB6B3-4409-44AB-AF15-9EF45FB00CF4_zpsh06uwsbk.jpg" border="0" alt=" photo 96CBB6B3-4409-44AB-AF15-9EF45FB00CF4_zpsh06uwsbk.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/reef9_zps981db009.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/reef9_zps981db009.jpg" border="0" alt=" photo reef9_zps981db009.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/54FC0BFA-DD45-46AA-85B4-0D65EAD3AE1F_zpsivgvyotl.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/54FC0BFA-DD45-46AA-85B4-0D65EAD3AE1F_zpsivgvyotl.jpg" border="0" alt=" photo 54FC0BFA-DD45-46AA-85B4-0D65EAD3AE1F_zpsivgvyotl.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/F1CBE24B-891C-4EC3-A8C8-9236086A8601_zpsvrqifpxa.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/F1CBE24B-891C-4EC3-A8C8-9236086A8601_zpsvrqifpxa.jpg" border="0" alt=" photo F1CBE24B-891C-4EC3-A8C8-9236086A8601_zpsvrqifpxa.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/AA01D0DB-9EE2-4117-8803-1766CFCA9D94_zpsja6sjenk.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/AA01D0DB-9EE2-4117-8803-1766CFCA9D94_zpsja6sjenk.jpg" border="0" alt=" photo AA01D0DB-9EE2-4117-8803-1766CFCA9D94_zpsja6sjenk.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/BDF1797A-303A-4314-A8B6-114AE2C1FE96_zps6so54qja.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/BDF1797A-303A-4314-A8B6-114AE2C1FE96_zps6so54qja.jpg" border="0" alt=" photo BDF1797A-303A-4314-A8B6-114AE2C1FE96_zps6so54qja.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/73D44514-244D-4353-9BF6-E8E7A4110C16_zps9dvqnibq.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/73D44514-244D-4353-9BF6-E8E7A4110C16_zps9dvqnibq.jpg" border="0" alt=" photo 73D44514-244D-4353-9BF6-E8E7A4110C16_zps9dvqnibq.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/827217E2-6DED-4C75-A2AF-C25004C1B927_zpsiw3zjfgm.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/827217E2-6DED-4C75-A2AF-C25004C1B927_zpsiw3zjfgm.jpg" border="0" alt=" photo 827217E2-6DED-4C75-A2AF-C25004C1B927_zpsiw3zjfgm.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/5FF9D214-1FFB-4436-A47B-4A903C0257FA_zpsxyl7g8yv.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/5FF9D214-1FFB-4436-A47B-4A903C0257FA_zpsxyl7g8yv.jpg" border="0" alt=" photo 5FF9D214-1FFB-4436-A47B-4A903C0257FA_zpsxyl7g8yv.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/6102AF94-4DF2-4CBE-94D6-4A9F0788954F_zpsyoqnli6u.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/6102AF94-4DF2-4CBE-94D6-4A9F0788954F_zpsyoqnli6u.jpg" border="0" alt=" photo 6102AF94-4DF2-4CBE-94D6-4A9F0788954F_zpsyoqnli6u.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/96CBB6B3-4409-44AB-AF15-9EF45FB00CF4_zpsh06uwsbk.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/96CBB6B3-4409-44AB-AF15-9EF45FB00CF4_zpsh06uwsbk.jpg" border="0" alt=" photo 96CBB6B3-4409-44AB-AF15-9EF45FB00CF4_zpsh06uwsbk.jpg"/></a>
karimwassef
Active member
I've even dabbled with light strips for alternatives and to get some data on light density options and tradeoffs in cost/performance/complexity:
<a href="http://s1062.photobucket.com/user/karimwassef/media/643635E4-334C-4366-B758-EBE159ACBD2F_zpstdhvm1sw.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/643635E4-334C-4366-B758-EBE159ACBD2F_zpstdhvm1sw.jpg" border="0" alt=" photo 643635E4-334C-4366-B758-EBE159ACBD2F_zpstdhvm1sw.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/AF07B500-5C05-4A29-A64F-F827427C6B70_zpsutuhugko.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/AF07B500-5C05-4A29-A64F-F827427C6B70_zpsutuhugko.jpg" border="0" alt=" photo AF07B500-5C05-4A29-A64F-F827427C6B70_zpsutuhugko.jpg"/></a>
This actually worked well for a diffuse moonlight effect that has had some interesting mating effects on my invertebrates
<a href="http://s1062.photobucket.com/user/karimwassef/media/643635E4-334C-4366-B758-EBE159ACBD2F_zpstdhvm1sw.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/643635E4-334C-4366-B758-EBE159ACBD2F_zpstdhvm1sw.jpg" border="0" alt=" photo 643635E4-334C-4366-B758-EBE159ACBD2F_zpstdhvm1sw.jpg"/></a>
<a href="http://s1062.photobucket.com/user/karimwassef/media/AF07B500-5C05-4A29-A64F-F827427C6B70_zpsutuhugko.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/AF07B500-5C05-4A29-A64F-F827427C6B70_zpsutuhugko.jpg" border="0" alt=" photo AF07B500-5C05-4A29-A64F-F827427C6B70_zpsutuhugko.jpg"/></a>
This actually worked well for a diffuse moonlight effect that has had some interesting mating effects on my invertebrates
Similar threads
