Par vs LED Array Size

[marc111]

Based on all of the LED threads I figured I should build a prototype to see what color balance I liked before committing to a full size array.
My tank is 72x24x30 in deep. I have been figuring on approximately 48 LED's per 2 foot section.

As a proto I built up a 12 LED array. NW are XPG @ 1000ma and RB/B are XPE @ 700ma Spacing 2 to 2 1/2 inches between LED's.

I have tried 40 deg and 60 deg optics

NW RB NW
RB NW RB
RB B RB
NW RB NW

The array is 10 inches off of the nominal water surface. I bought an appogee sun calibrated PAR sensor and I am using a Fluke multimeter.

With the 40 deg optics in place the PAR at 10 inches is approximately 300.
At 40 inches below the LED's (the bottom of the tank) it is about 40.

This kind of surprised me. Looking at some of the fully built up arrays on the forum that have posted PAR numbers I expected the top number to be better than 1000.

It would seem that there is more par addup due to the overlap of the light cones than I was expecting.

Any thoughts would be appreciated.

Mark

[125G Reefer]

Interesting. Thanks for the numbers. Now this puts a damper on my build. Think I'll hold off now. Are those numbers with all your white and blue l.e.d's running?

[marc111]

Yes this was with the NW running at the full 1 amp and the RB's dialed down to about 500 ma for the color balance I liked.

Mark

[125G Reefer]

Wow. Those are the worst numbers I've seen yet. Yikes.

[Crosis]

Blue and Red spectrum's don't read well on the Apogee...

Here is part of an email correspondence between Apogee and I...

Hi,

The MQ-200 will be able to measure the PAR value of LED’s while submersed, however, there are a couple of caveats. The quantum sensor is considered a broadband sensor that ideally gives equal emphasis to all photons in the PAR region. With this in mind, and in addition to the imperfect spectral response of the quantum sensor (http://www.apogeeinstruments.com/quantum/spectralresponse.html), the MQ will underestimate the measurements of LED’s. To add to the difficulty of taking LED measurements, different setups will result in different accuracies for the MQ. This goes back to the imperfect spectral response, in that it underestimates the blue region (400-500nm) and drops off in the red region (650-700nm). So, the errors are much greater for a setup of Blue/Red LED lighting than say a setup of Blue/White LED lighting.

Because of the vastness of LED lighting systems these errors have not been quantified, but the MQ will still provide a good relative reference for your LED’s. The MQ will have a calibration setting for electric lights or sunlight, which is about a 10% difference in reading. Some users have mentioned that they will set the meter to sunlight calibration when measuring LED’s. In turn, this brings the PAR values closer to the actual values for LED’s.

Hopefully, that helps. Please let me know if I can be of any further assistance.

Regards,
Jacob Bingham
Applications Engineer

[marc111]

I have noted this response of Apogee's as well. I am running an essentially 50/50 mix of RB and NW. However it should be noted that I am really not comparing absolute readings but instead comparing my results to other LED results that have been posted on this board. These results have been from panels of LED's that range in the 36 to 68 count range with varying use of optics and anywhere from 2:1 to 1:1 RB to W LED's. My initial expectation was that a small subset of one of those arrays would at least in a localized area, perform at similar numbers to the full array. Since this is not the case, I can only conclude at the moment that overlap of LED illumination is a larger factor than I initially thought.

As a side note I put my sensor and multimeter under full sun at noon today (Apogee's calibration test) and got the expected approx. 2250 reading so I can say with confidence that my meter is not miss reading the sensor output in any significant fashion.

I am ordering the remainder of LED's for a 49 LED panel and will post once they are in and operating the next set of results.

Mark

[katchupoy]

Marc,

How did you arrange the 12 led build? Is it by rail? I mean...
1x12?
3x4?
4x3?

Thanks.

[katchupoy]

Wow. Those are the worst numbers I've seen yet. Yikes.

Compare to what?

[sfsuphysics]

IMO, those numbers sound about right. 12 LEDs will not get a PAR of 1000 10" from them. 300 at 10" doesn't sound terribly bad at all actually.

What's the PAR a couple inches away? This would be the good gauge to use, because at that distance it doesn't matter how many you have, the reading should be pretty uniform since it'll really only be reading one.

[Ram Jr]

Here is my results.

http://www.youtube.com/watch?v=aeFdcg995S4

[marc111]

To answer a couple of questions:
1) As shown in my original post it is a 3 x 4 array.
2) I will make a reading a couple of inches away tonight.

Thanks for the feedback,
Mark