DiY LED - Cree CXA3070

Well I am running 16 LEDs at 60w a peice which is just under 1kw. Should I power factor correct it? Right now it shows I am running at 9 AC amps on my killawatt on my flower room alone.
 
9 amps should be no problem unless there are other things on that same circuit. Out of curiosity, what does your kill a watt say for pf?
 
Rooster99 said:
Holy crap!!!! 80???? You are building a monster!!! Its probably buried in the previous 47 pages but what is your proposed plan for them?
Click to expand...

Well, I did not decided yet. Wanted to grow with 8x or 6x 600 watt hps, but this thread convinced me that led is the better option for me. So now I am going for led, first 40 pcs @50 watts with active cooling. Did not decide how to build those 40 by the way.
 
CannaBare said:
I made contact with a chinese company who made me drivers with a custom current(1600mA) at 38 volts or something like that. It is perfect for the CXAs because they were basically made for them. I'll take a picture of the driver when I get a chance! Tested with a multimeter they put out 1550-1580mA at thermal equilibrium.
Click to expand...

I have noticed that the CXA3070 Z2 bins I have tested are just over 38V at 1450mA but the Z4s are 39.5vF at 1450mA. I wonder if that would be a problem for these drivers. Looks like the Z4s could be over 40vf at 1550mA. Maybe I can one from them for some testing.
 
I would assume they would run at the same voltage. It is just a more efficient system. They told me the drivers actually go to 40volts so it is not a problem. Why do you think the voltage increases? They don't ever mention that. And I have yet to have a problem

edit: according to pct.cree.com the forward voltage should be the same. I am going to go test my voltage on my driver
 
I checked my voltage and I am running at just under 38 volts. I think the reason your voltage is higher is because your drivers are remote and you might have voltage drop over that long distance( however long it is). My LEDs are also Z4.
 
stardustsailor said:
CXA3070 Z4 / Tc=45°C / Io= 2070mA : Vf=39.5 V measured at driver's output pins .
( > 6hours continuous operation. MeanWell HPL-80H-42 )
Click to expand...


Wow sds...45*c at 2a? My coolest one was 41c at 1a. The base of your heatsink looks a bit thicker than mine...I wonder if that is making the difference.

Could also be the way I tested the junction temp. I just used kapton tape to secure the probe. I think the probe might have been picking up the radiant heat from the cob
 
Positivity said:
Wow sds...45*c at 2a? My coolest one was 41c at 1a. The base of your heatsink looks a bit thicker than mine...I wonder if that is making the difference.

Could also be the way I tested the junction temp. I just used kapton tape to secure the probe. I think the probe might have been picking up the radiant heat from the cob
Click to expand...

Heatsink base thickness: 10 mm ( 1 cm )
+ 4x 120mm fans (+ noise ....)
(Ta= 27°C).
Indeed ,cooling of fixture is extraordinary...
Still,it takes >20W of fan power and 8 pounds (lbs ) of soft aluminium alloy heatsinks to reach those figures ..
Plus ,good heatsink surface treatment & good TIM conductivity (A.S.5 ) as well as thin TIM thickness (~0.2mm )



T case and not T junction ...:wink: ....
Tj = Tc + ( Rθj-c * Pheat )

At Io =2070mA ,if Tc =45°C ,then Tj= ~ 75°C
 
CannaBare said:
I checked my voltage and I am running at just under 38 volts. I think the reason your voltage is higher is because your drivers are remote and you might have voltage drop over that long distance( however long it is). My LEDs are also Z4.
Click to expand...

I did the testing on a bench and the driver was about 1 foot away from the LED. It may be that I just happened to select a COB with a particularly high vF. Also, that was at Tj25C so the vF would be about .25V lower once warmed up. I will test a few more Z4s and Z2s to see if that was just a fluke.
 
OH I just remembered. I was including the vF load that the ammeter adds to the circuit because I was testing driver efficiency. So when using a pair of multimeters, I check vF of the COB and ammeter combined. That explains the strangely high vF but it does not explain the large difference between the Z2 and Z4.
 
guod said:
View attachment 3162597
the note under the figure cleary states 3000K; 4000K; 5000K for the given graphs

View attachment 3162598
cree on the other hand gives us a range (3700K-5000K), but not say which colortemp it is exactly.

here is a graph of a warmwhite(2700K) with 17%(W) blue.
View attachment 3162600
white with a touch of pink in my eyes.
Click to expand...


Good eyes ,then ...
pink.JPG

Valoya AP67 – General Growth Spectrum
For vegetative and generative growth. Performs very well with a large variety of plants,
ranging from cucumber and tomato to lemon trees and even snow algae.
It provides strong vegetative growth and early flowering of plants.
With some leafy green plants elongation is achieved,
while with some flowers the growth result is a compact
plant. The AP67 spectrum has also been shown
to promote quick and good rooting. The AP67 is
strong on far-red (about 18 % of the total radiation) and has no uV radiation.
Suggested use:suitable for most vegetables e.g.
cucumber, flowers and tree seedlings. Also positive results with algae.
Appearance to humans:Pinkish light, quality
control possible, good working environment.
Click to expand...

Valoya uses two different types of (proprietary ) phosphor conversion leds (COBS ? ) in various combos .
Most probably one is a Cool White (yellow phosphor ) and the other one is ..pink ('orange' phosphor ) ...

(18% FR denotes a high CRI phosphor used =
most probably a peaking ~640-660 nm Nitride phoshor ...+blue die =
.....pink light )





http://www.valoya.com/document-repo...ument/horticultural-lighting-guide?format=raw
 
Last edited:
Sds..I thought I had a lot of fan power at 10w. Hmmm..my fans should have been on the ends right over the hotspots.
Still would have had plenty of air making it to the middle. Would have been easy to do shucks..

Thing I'll just roll with it for now
 
You must log in or register to reply here.
Back
Top