DIY LEDs - The write-up

Status
Not open for further replies.
IMHO no difference as long as each string has protection (fuse or other). And once they are combined you are using heavy enough wire to carry the combined currents.
 
Thanks FishMan,
My plan was to use 22 gauge to link the LED's down the strings, and then 18 gauge to combine the strings at the ends and link back to the MW.
I was going to run resistors and fuses as per the drawing.
 
I would always use the terminal block method. It's neat and tidy and allows some flexibility. There's no reason you can't put the block anywhere that's convenient. You can put it next to the driver or you can put it in the fixture and run the one, bigger wire, back to the controller.
 
I remember reading somewhere that it is possible to use an Apogee sensor connected to a multimeter to calculate PAR readings ? Details ?

thanks
 
Yes, I ordered one and it appears to work. Waiting for access to a PAR meter to verify, but have no reason not to believe it works. There is a thread around here somewhere but all it really said was it can be done and there are 3 sensors. Two need power and just amplify the signal (and cost more).
SQ-110 Daylight is what I think you want (they also have electric reference). Multiply the millivolts by 5 to get PAR ($139 IIRC).
Then one as a gain of 2.5 so multiply by 1 ($189 IIRC).
And another has a gain 5 5 divide by 2 ($189 IIRC).
Standard shipping is $15.
 
If I Remember Correctly :)

They read about 10% different. Since we are trying to emulate the suns it made sense to me to get sunlight. Did you tell them the application or only that you wanted par from LEDs.

From http://www.apogeeinstruments.com/quantum/differentlight.html
This table shows the quantum sensor's output under varying light sources. The sensor's spectral response, the spectral output of electric lamps, and these errors are all constant -- allowing the user to calculate the correct output for each light source.

For example, if the MQ-100 Quantum Sensor set to electric lamps is used in sunlight, and a value of: 1,500 μmol m-2 s-1 is given, the actual light is 1,665 μmol m-2 s-1.

In a controlled environment with multiple lamp types (such as a mixture of MH and HPS), a more precise percent error can be determined when the amount (%) of each type of light present is known.

I am not sure it really matters. When folks here have bought PAR meters what sensor do they get? You should (I should have) gotten the same one. I didn't think about it at the time. I bought for our club and no one corrected me.

This thread also said sunlight - doesn't mean it is right or wrong. Just another point.
 
20 degree or less optics

20 degree or less optics

Can someone point me in the direction of anyone that has used 20 degree optics or less and mounted their LEDs 3-4 ft above the tank? or do these builds not exist yet?
 
Question to LED driver specialists:
I have installed a 30W 700mA NoName LED driver for a string of 11 XP-G and my Kill-a-watt reads 43W power consumption! Is is the fault of the Kill-a-watt or the driver is so inefficient?

Same setup with a MeanWell 350mA showed a 2W overhead.
 
I am looking to build an LED light for a new 115 gal tank. 48X 18 X 30. I planning on doing 48 LEDs. I want to do: 24 CREE XR-E Royal Blue 3W LED on star
12 CREE XR-E Q5 Cool White 3W LED on Star
12 CREE XP-G Neutral White 3W LED on Star
Has anyone used anything other than the traditional bulbs (cool white and royal blue) with good results? Any suggestions on patterns or even the use of the warm white?
Can I mix and match my white's on the same string or do they need to each be on their own string. I've also read that each color needs their own driver....can the white's be on the same? Also what is recommended as far as lenses for a tank 30 deep? I don't have anything in the tank at this time, I don't plan on having SPS but may in the future.

P.S. My electrical knowledge is very limited. My husband has an associates in electronic engineering so he is doing the actual work, I'm just attempting to learn and do the planning....This is also my first tank.

Dena
 
Has anyone used anything other than the traditional bulbs (cool white and royal blue) with good results?
I have tried - the results were not very good.
As far as your selection of LEDs - I would recommend XP-E royal blue and XP-G cool white. You can mix LEDs on 1 string, but you will lose the ability to control each color.

Adding neutral whites will lower the kelvin temperature. If you really want to do this, just reduce the number of royal blue and increase the number of cool whites.

Before ordering the cool whites check their exact color temperature - it can vary from 5000k to 8000k.
 
The cool whites I am looking at are on RapidLED.com. They say they are ~6,500K. If I am doing 2 sets of 24, 4 rows of 6...any ideas on how many extra whites I should do and how to pattern that.

Dena
 
MosMike, Measure the voltage of just the string and the current of the string. This willtell you the power the string is using. Everything else is assumed loss in the driver.

You could also hack an extension cord and measure the AC side (VERY CAREFULLY) the same rules apply current * voltage = power. Been a while there sis something about true RMS (Root mean square ??)measurements or a possible conversion. Maybe kcress will chime in or if you really want to go this far let me know and I can research it more.
 
Dena mike is shoewha correct. Changing the brightness of the strings will change the color. The problem i sthat if there is no red produced by the LED then even if you made the white string 100% and blue 0% you would still have no reds. IMHO the 6500 would work well.
 
Question to LED driver specialists:
I have installed a 30W 700mA NoName LED driver for a string of 11 XP-G and my Kill-a-watt reads 43W power consumption! Is is the fault of the Kill-a-watt or the driver is so inefficient?

Same setup with a MeanWell 350mA showed a 2W overhead.

hmmm 11 x 0.7 x 3.3 = 25.5W

I've not seen a bad Kill-a-watt, ever.

Yes that would be the efficiency result if you've actually measured the voltage across the string and the current thru it. It could be you're actually driving the string harder than you think too - if you haven't measured it.

You cannot measure the AC side current and multiply by the voltage to measure the driver's power consumption. The current won't be in phase with the voltage making the result invalid.
 
Status
Not open for further replies.
Back
Top