DIY LEDs - The write-up

[TheFishMan65]

Here is my thought. Feed the 10 volts into a 100k potentiometer. You can then set the maximum current. If you are worried about someone touching the pot add a parallel resistor to limit the range of the potentiometer. Anyone see a reason why this won't work?

 

[mtothet]

I have done alot of reading on these DIY threads and appreciate all the information in them. I think I am beginning to grasp some of the concepts. Please critique and give suggestions where needed. This plan is definitely still a work in progress.

Tank is 60x24x24, soft/LPS maybe a few SPS later on.

Plan:

2ea 24x8.5 heat sinks

LED's:
24 XP-E CW
24 XP-G NW
60 XP-E RB
12 XP-E B

60 degree optics

Drivers:
HLG 185-42B
Using fuses, resisters and terminal blocks (still need to research terminal blocks)


Question:

1. Which HLG Mean Well do I need? HLG 185-42B
2. These HLG's are hard to find, does anybody have a good place to find them?

Thanks

 

[TheFishMan65]

Sounds good, looks like you are planning for 12 LEDs per string so probably the 42 version that would allow 3.5 volts and maximize the number of strings. I would do 12 blue and 60 royal blue. Then each string can be the same and consist of 2 blue and 10 royal blues. How many HLGs are you planning?

 

[mtothet]

Working on that answer, Thanks for the reply.

Sounds good, looks like you are planning for 12 LEDs per string so probably the 42 version that would allow 3.5 volts and maximize the number of strings. I would do 12 blue and 60 royal blue. Then each string can be the same and consist of 2 blue and 10 royal blues. How many HLGs are you planning?

 

[TheFishMan65]

Dear [XXX]:
The D-Type power supply uses MCU to control the output current. You can reference the attached file which is the curve of Io vs time. You can define the curve with your requirement.

Best regards,
Willard
Application Engineer
Mean Well Taiwan

The enclosed PDF is a plot of time vs current at 0 time the current is 100% at 7 hours it drops to 70%, 9 down to 50%, 11 back to 70%, and 12 is 100%. I would like know how smooth you can make it and whether you can get it to zero. And of course the price.

 

[brassmonkyballs]

The enclosed PDF is a plot of time vs current at 0 time the current is 100% at 7 hours it drops to 70%, 9 down to 50%, 11 back to 70%, and 12 is 100%. I would like know how smooth you can make it and whether you can get it to zero. And of course the price.

A built in controller of sorts? I too would like answers to your questions plus if it's available.

 

[brassmonkyballs]

Here is my thought. Feed the 10 volts into a 100k potentiometer. You can then set the maximum current. If you are worried about someone touching the pot add a parallel resistor to limit the range of the potentiometer. Anyone see a reason why this won't work?

I like this approach.

 

[jkopp36]

Here is my thought. Feed the 10 volts into a 100k potentiometer. You can then set the maximum current. If you are worried about someone touching the pot add a parallel resistor to limit the range of the potentiometer. Anyone see a reason why this won't work?

Got it. So then my 0-100% dimming would go from zero-10v to zero-whatever the pot ends up needing to drop the max voltage down to right? Thanks for the info!

 

[TheFishMan65]

I will keep you posted on what I learn about the D version.

As for the pot do you agree with this wiring.
Outside legs of the pot get the 1-10 volt signal from the controller. Center and ground leg go to the HLG. A resistor across the center and ground (same as the APEX) would then limit the current. The resistance can be found from:

     100k * Rfixed
R = ---------------
     100k + Rfixed

So for 50% Rfixed = 100k
60% Rfixed = 150k
70% Rfixed = 233k
80% Rfixed = 400k
90% Rfixed = 900k
100% Rfixed = none

 

[brassmonkyballs]

I will keep you posted on what I learn about the D version.

As for the pot do you agree with this wiring.
Outside legs of the pot get the 1-10 volt signal from the controller. Center and ground leg go to the HLG. A resistor across the center and ground (same as the APEX) would then limit the current. The resistance can be found from:

     100k * Rfixed
R = ---------------
     100k + Rfixed

So for 50% Rfixed = 100k
60% Rfixed = 150k
70% Rfixed = 233k
80% Rfixed = 400k
90% Rfixed = 900k
100% Rfixed = none

Agreed

 

[brassmonkyballs]

For blue
34.2V * .45A * 4 = 61.56 watts
for white
31V * .9A * 3 = 83.7 watts (oops 50% so 41.85)
So 103.41 watts yep you need to check your math

Now do you by chance have par numbers with this set up?

I pulled out my PAR readings I had taken. When I posted the output I had them dimmed on the dimmers about right where they are now....at least very close. I also did not have optics on them at that time. I took readings with all LEDs, and then just blues, and just whites. Since I didn't take V and I readings at the time this isn't all that meaningful to correlate I to PAR but its indicative at least. I'll just give the center readings of each here. I don't own the meter so i can't take readings again and the person charges a nuisance fee for driving over and I'm not paying him again

FWIW here's the readings:

50RBs, 18CW, & 6 NW dimmed approximately as I mentioned above:
PAR at top = 420
PAR Center = 250
PAR bottom = 235

44RBs dimmed approximately as I mentioned above:
PAR at top = 310
PAR Center = 210
PAR bottom = 152

18CW, 6NW, & 6 RB dimmed approximately as I mentioned above:
PAR at top = 160
PAR Center = 100
PAR bottom = 80

 

[brassmonkyballs]

The enclosed PDF is a plot of time vs current at 0 time the current is 100% at 7 hours it drops to 70%, 9 down to 50%, 11 back to 70%, and 12 is 100%. I would like know how smooth you can make it and whether you can get it to zero. And of course the price.

Can you attached the PDF? Would like to take a look.

 

[TheFishMan65]

Well I think we can scratch the D off the list. I got this back:

Dear [XXX]:
I'm sorry we only provide four points of the time and five values of Io. The MCU can't completely turn off the output current. The program of the MCU will only be set by your requirement curve when we produce them. You can't change it by yourself. If you need D-type model, you can contact our global distributor(http://www.meanwell.com/contact_distributors.html).

Best regards,
Willard
Application Engineer
Mean Well Taiwan

after sending

Willard,

A few question please.

How smoothly can you drop the current? The graph shows large steps. Maybe a beter question would be how many step can be supported?

Can the MCU completely turn off the current?

How do you go about programming the MCU? What type is it?

Do you know availability and pricing?

Thanks

So in English, there goes our ability to dim. And they are Reallly BRIGHT moonlights. And still nothing on price or availabiliy, but this was tech support.
or in Pictures since we all like pictures:
:sad1: :headwalls: :mad2: :thumbdown: :headwallblue: :sad2: :uhoh3:
or one more shot
:ape::mtool:
But since BMB was nice enough to get me PAR readings

 

[TheFishMan65]

I had the pleasure of going out to Dr Mac's Pacific East Aquaculture over the weekend, boy I bet your jealous. But more to the point I asked him what he thought of LEDs. While I can't remember his exact words he was all for them. He felt they had good growth, light, and liked the energy efficiency. The only negative was the start up cost. It sounded like he would switch over if there was a cheaper way to get the fixtures.

 

[brassmonkyballs]

Measuring PAR

Measuring PAR

I contacted Apogee about which to quantum sensor to purchase for our application. (I hadn't realized I didn't need to spend $300 on a whole meter and could get this to use with my MM for alot less). I received such a good response I thought the information would be useful to post here.

"Hi xxxxx,

Yes, the SQ-120 (electric calibration) would be our best option for your application. These units are sealed and submersible up to 30m and are compatible with salt water. The SQ-100's are considered self-powered and will generate a mV output that you can measure with your voltmeter. We calibrate everything to a factor of 5.0 µmol/m2/s per mV, so just multiply your mV output by 5 to derive the photon flux value.

In regards to measuring LED's with the quantum sensor, there are some caveats to doing so. The following link shows the spectral response of our quantum sensor (http://www.apogeeinstruments.com/quantum/spectralresponse.html). As the graph shows, Apogee quantum sensors underweight blue light, and as a result, photon flux measurements for blue LEDs will be too low. They also overweight red light up to a wavelength of approximately 650 nm, above which they do not measure, and as a result, photon flux measurement for red LEDs will either be too high (if the LED output is all below 650 nm) or too low (if a non-negligible fraction of the LED output is above 650 nm). Additionally, LED's often have a very narrow spectral output, with a sharp peak of only a few nanometers. So, unless the quantum sensor has a perfectly flat spectral response, meaning it weights all wavelengths of light exactly the same, there will be errors. Electrically calibrated Apogee quantum sensors will likely provide a reasonable measurement for white LED's because they are broadband, and because electrically calibrated quantum sensors are calibrated under CWF lamps. However, for narrowband LEDs, like red and blue, Apogee quantum sensors will not provide an accurate measurement.

You can use the same spectral response graph (http://www.apogeeinstruments.com/quantum/spectralresponse.html) to get a relative idea of the error. For example, a 450nm blue LED will have a relative response of approximately 0.8. Therefore, you can figure that the photon flux reading from the sensor is probably reading approximately 20% low. Just remember, this approach is only relative so give yourself a wide margin of potential error. A blue/white configuration should give you reasonable accuracy, particularly from the broadband spectrum of the white.

Hopefully, that makes sense. Please let me know if there are any questions.

Regards,

Jacob Bingham
Applications Engineer
jacob.bingham@apogeeinstruments.com
www.apogeeinstruments.com
721 W 1800 N Logan, UT 84321
ph: 435.792.4700 fax: 435.787.8268"

 

[TheFishMan65]

Wish I had contacted them first. I got the sun reference version, but can pretty much verify everything else.

 

[brassmonkyballs]

Wish I had contacted them first. I got the sun reference version, but can pretty much verify everything else.

Well its just a 10% adjustment too high all the time if I understand their information in the website...and its always repeatable at least.

 

[The Grim Reefer]

Nice thing about the sun calibration is you could use the clearskies calculator Apogee has a link to on their website to use the sun to check your accuracy as it ages so you know when to recalibrate it.

 

[TheFishMan65]

+1 too both of you guys. Perhaps I should have qualified. I am not disappointed just had I known better would have gotten the other meter.

 

[The Grim Reefer]

I double checked and the 10% high in sunlight mode estimation is about right. I have a dual calibration meter. So for blue LED's subtract 10 then add back 20% or add 20% and subtract 10% LOL!