DIY LED light PAR Readings

[lynxvs]

Finally my Quantum sensor has arrived and I'm now able to do some PAR measurements on my LED light. If anyone wants to make some PAR measurements and doesn't want to invest $300.00 on a meter and you have a good multimeter lying around you can just buy the sensor. All you do is hook up the wires and measure the voltage. Then you take the voltage measurement and multiply it by the standard calibration. I purchased the Apogee Model SQ-120 with electric calibration for $139.00.

My DIY LED Light
http://reefcentral.com/gallery/data/500/200771Pendant.jpg

Quantum Sensor
http://reefcentral.com/gallery/data/500/200771PARsensorsmall.jpg

The first readings I took the Light is 10" from the water surface. I removed the egg crate and splashguard.

http://reefcentral.com/gallery/data/500/200771PARTank10_Inch.jpg

The second readings the light is 6" from surface.


http://reefcentral.com/gallery/data/500/200771PARTank6inch.jpg


From what I can see from the data the higher the light is mounted the better the light distribution on the edges with an overall decrease in PAR directly under the arrays, which makes perfect sense. I also found significant attenuation with a splashguard and egg crate. I'm not sure what spectral errors are associated for LED lighting, they do not list it in the documentation that came with the sensor. If I recall correctly from Dana Riddle's review of the Solaris he used Sunlight calibration, in that case my readings wouls be 8% low. How do these readings look? I'm new to PAR measurement and what are good PAR readings.

My DIY Build website
http://www.photiczoneled.com

[reefqu33n]

Very nice, atleast i know another sucessful story about these LED before moving onto these lights.
check out this other thread that I found here on RC about these lights but the reading are much higher in their par.
http://reefcentral.com/forums/showthread.php?threadid=1538174&perpage=25&pagenumber=1

[lynxvs]

Thanks, yea I think he used something like 100 CREE LEDs.... I didn't want to take out a second morgage to pay for it.

[areze]

that guy was a sham, notice his status... there is not logical answer to how he produced those par with those lights, as they are not CREE LEDs at all, rather appear to be 5mm LEDs, which doesnt add up. he never would answer the real question on what it used. 8w of lighting for ~150LEDs per bulb, does not produce that kind of par, and would certainly not be running 3w per LED as cree does.

there was another guy here that made a CREE array, though I cant remember his SN right now... he went ALL out, custom aluminum core PCBs, ect, he got like 1000 par at the sand bed or something absolutely ridiculous, he had issues with bleaching obviously, I think he just toned it down. whatever the case, I want to say he used like 50-70 cree LEDs per foot.

there is no question that LEDs can compete, they are more efficient, and they COULD be price competitive, but not with the outrageous markups that they currently have. the raw materials are competitive, but DIY is not for the weak of heart at that point.

[Ace1]

PAR SMAR, i think its all BS until you test LED's with some SPS for a prolong time period, then people can decide how to modify K temp to suite their own eye pleasure...

I have never seen any PAR information on all the different Halide and T5 bulbs...

[areze]

most conventional lighting shoots for 2-300 par at the sandbed. 1000 over the tank, LEDs are a bit skewed as at 10 or 6" the light wouldnt be diffused, so your reading will depend on if your in an LEDs beam pattern or not, here on T5 or MH obviously you dont have a "beam pattern" so you can get 1400par up there, and only 150 down at the bottom of the tank.

still you can support SPS in upper levels of that tank, and softies at the sand bed. maybe sps down bottom, but it would probably be a bit brown and slow growing.

[lynxvs]

areze
Not sure if you are talking about me or someone else...I am running with Luxeon Rebels 12 cool white and 13 Royal blue for a total of 25 LEDs per foot. The lumens output for the white is 130 Lumens at 700 mA campariable to the CREE LED. So I am assuming these are good PAR numbers... I thought they were a little weak I wasn't driving the array at max power when I made the measurements.

[tinnghe]

great work how long have you had the light for? do you have a part list?

[lynxvs]

Thanks tinnghe,
Here is a spread sheet of the parts and prices enjoy!

[areze]

there was another guy, had a car in his sig lol, cant remember his sn though. the par numbers are reasonable, but not blasting, I think most would want to see atleast 200 at the sandbed, if not 250 or 300. but when your DIY it doesnt matter what others might want, long as it gets you the par you want where you want it. the beauty of it.

[lynxvs]

<a href=showthread.php?s=&postid=14099939#post14099939 target=_blank>Originally posted</a> by areze
most conventional lighting shoots for 2-300 par at the sandbed. 1000 over the tank, LEDs are a bit skewed as at 10 or 6" the light wouldnt be diffused, so your reading will depend on if your in an LEDs beam pattern or not, here on T5 or MH obviously you dont have a "beam pattern" so you can get 1400par up there, and only 150 down at the bottom of the tank.

still you can support SPS in upper levels of that tank, and softies at the sand bed. maybe sps down bottom, but it would probably be a bit brown and slow growing.

areze
I'm not sure what you mean "here on T5 or MH obviously you dont have a "beam pattern" so you can get 1400par up there, and only 150 down at the bottom of the tank." ?

The 1400 PAR reading was centered directly under one of the arrays.

[lynxvs]

I have a new design in the works that should take care of that problem...

[areze]

I mean that LEDs output in a beam of light, so if you get too close to the actual array you will have "dark spots" between the beams, obviously once they get a little ways out, these beams all have spread and covered the whole area. with a T5 or MH the light is constant across its range.

thats the ussual story with LEDs anyway, solaris needs to be a bit off the tank or you see a spot light effect in the water at the top of the beams.

[eme]

I'd suggest using borosilicate glass for the lenses .. loads available with different spread patterns .

LEDs have already 'made it' but as others say there are new developments knocking around which will make them increasingly suitable for our applications.

[areze]

dont want it to spread though, LEDs have perfect optics as it is, near perfect vertical entry to the water. just with the requirement of sufficient clearance that the lights can disburse prior to entering the water. its a small price to pay though to have no inefficiencies of a lens, diffusor, or coilator.

[eme]

areze Each to their own, that's no problem! The loss through a b/s lens is minimal though compared to other challenges.

BTW - I should confess to part owning a small marine LED lighting company (www.lumishore.com) .. but, no, we dont manaufacture aguarium lights, indeed I'm looking to buy some myself now ! If only the manufacturers would actually respond to technical questions on their products ..

[cobra2326]

I just want to express my eternal appreciation for the parts list! I'm thinking about building a microcontrolled moonlight system with sunrise, sunset, etc. using a few of the luxeon LEDs. I might even use a few orange LEDs to simulate the color of sunset...

[lynxvs]

No problem I have quite a few Luxeon III stars that I used in my first design that might work good for moonlight. You wouldn't need to build a PCB to mount them like you would need with the Rebels just mount them on a heatsink You would still need an LED driver as well as your microcontroller.

[cobra2326]

I know liveforphysics said he used a computer power supply to run them and it worked fine. The big problem with them is the large current required and the heat that would be generated by using resistors, etc. to limit that. I'm still learning electronics, so it'll be a little while before I build one with the 1W LEDs. Right now I'm experimenting with an ATMega168 and some high-output standard 20mA LEDs.

[wayne in norway]

Wr.t. ace there is a mountain of PAR info on MH and T5 combos, (esp MH) so I just have to assume you are trolling......

Good to see some sensible numbers for LED coming too.

[eme]

cobra2326 Not sure what you're trying to achieve with the setup but 20mA isnt going to crack our lighting requirements in the aquarium world; with manufacturers producing 100+ l/watt and 200 round the corner the future will be about heat dissipation and drivers ... fluorescents and MH are now matched ( arguably surpassed) in terms of output with an inherent LED efficiency but you are going to have to use Luxeon Rebels (or similar) to achieve it.

Mind you I may have totally missed the point!

[areze]

I think he was more tinkering with the designing of an array with the comparetively cheap 5mm LEDs.

trouble with it is, their requirements are so different, pin mount vs surface mount, and obviously their current requirements, you can run 5mm LEDs off a potato(ok maybe not, no clue) the big dawgs require a little more finesse to power.

[cobra2326]

Agreed. Right now I'm just testing for moonlights. Basically I want to be able to see my tank at night, and 5mm LEDs work just fine for that.

I don't think I ever plan to use LEDs as my sole light source. T5s are cheap enough to accomplish my goals. I'll be using the LEDs to supplement the T5s for a more smooth on/off transition, sunrise and sunset, and moon lighting.

[lynxvs]

cobra2326
Will you be able to dim the LEDs with the microcontroller? LEDs can be dimmed by current or PWM. I know the microcontroller can supply the PWM but usually you have to have an LED driver. I'm not sure you can just feed a PWM signal to a standard LED and it would work.... I could be wrong I never tried it... I would go with some older Luxeon Stars or similar and a Buck Puck. If you really want to learn about electronics build your LED driver lots of LED Driver IC's out there. As far as power supplies I would buy a 24 V 5A or greater switching power supply.... they are very cheap I bought a 24 V 12.5 A power supply for $24.95. Lower current is even cheaper.

Now getting back to the original thread...
I'm surprised no one commented on the PAR sensor I used. I thought it was clever just to buy the sensor and use a common multimeter so you would have a PAR meter for $139.00.... oh well I guess it was common knowledge...
I'm also amazed that there were so few reponses to the PAR readings... good bad or otherwise......

So does anyone else have some PAR readings for an LED light ??
or some comparison to conventional lighting? I'm not trying to stir up the whole LED debate just trying to get a handle on PAR numbers whats good bad or otherwise.

[jimnrose]

Lynxvs, I also was shocked when I found out last week that the PAR meters cost $300 & noted your approach @ $139, but I'm new to the hhobby and haven't quite got used to even $139 for a meter. I'm still in the 'hardware phase' so can wait for both LED development as well as checking around for a used measuring devise. Thanks for your input. Jim

[evilc66]

I managed to hit 500 PAR at 14" and 200 PAR at 24" with 40 degree optics. I was pretty happy with that. I have more optics on the way to see how drastic the changes are going from 30 degree to 80 degree (and various points in the middle). This was using Cree XR-E Q5s @ 1A, and XR-E royal blues @ 700mA.

[eme]

lynxvs - The cost of your PAR meter is phenomenal! We dont own our own but our test costs per hour for lux/lumen calibration/test/certification are much more than that ! Congrats

PAR readings - I cant read them but you can always compare with Sanjay's site - he has control-tested aquarium lighting extensively.

[lynxvs]

The $139 is just for the Sensor... you still need a multimeter lying around. Thanks for the note.

[lynxvs]

evilc66
Your PAR readings sound good... i think that you will find the PAR readings with 80 degree lens much lower. and the 30 Degree not much higher with low spots on the edges. I'm not sure how you light is arranged but I have two arrays in a 24" fixture over a 30" tank. With 36 Degree lens I find I have to raise it 6" to 12" above the tank for complete coverage. I did a quik spreadsheet that calculates tank coverage, LUX and PAR.

[lynxvs]

I should note that some of the calculations i used came from the lens data sheet that I used and might not apply to all lens...The lux to PAR conversion came from the Apogee web site...

[cobra2326]

lynxvs
Your website link does not work. It's missing the 'h' in photic.

[cobra2326]

Also, did you use a metal-core PCB for the SMD LEDs? From what I'm reading, junction temps are crucial for longevity.

[lynxvs]

cobra2326
Thanks for the catch on the web site typo....

I don't use a metal-core PCB. The Luxeon Rebels can use a standard FR4 PCB as long as you follow the recommended layout.
It’s the thermal resistance number you worry about.... Yes thermal management is critical in the long life of an LED and a heat sink is critical but a good heat can be as simple as a flat aluminium plate as long as the right size is used.