DiY LED - Cree CXA3070

stardustsailor

Well-Known Member
Imagine under 600 Watts ! ...


" How 12 pieces of CXA3070 AB 3000K 80 CRI ,at 1400mA,surpass any 1K HPS " ....

10 x AB at 1400mA.JPG
............................................................................
Normalised flux and radiant power specrums of the absolute flux/ power values of the 12 CXAs per (λ),to the corresponding standard.
In the background each standard's action curve ,normalised to max (100%)
RELATIVE to the absolute values of umols / (λ) or W / (λ ),of the CXA's .

din radpowr.JPG

PEF.JPG


..........................

Here the same amount of total radiant power (244,434 W ) of the 12x CXAs ,
compared with an "ideal" McCree and an "ideal" DIN5031-10 -neu , light sources of same total output light power.(same area under every curve).

12x abs rad pwr vs ideal mcc & din.JPG

Clearly,both standards suggest more evenly distributed power over the blue wls ,
while also both standards suggest more power at deep reds ...
Still the CXA 3000K 80 CRI ,can be considered as a superb solid state light source ,
for growing plants ,as it covers (falls into ) most of the Φ / (λ )* ,both standards proposing ...

*(Watts per wavelength )

Cheers.
 
Last edited:

stardustsailor

Well-Known Member
Well ...
My new LED-COB "bio analysis" tool ...

Now....


While the CXA 3000K is very efficient regarding it's photosynthetic capabilities and potentials ,
as a light source ...


(10 x CXA3070 AB 3000°K-80CRI ,1400mA,Tc=50C )

CXA3000Kbioessay1.JPG

Well ,ok ...
Photomorphogenically speaking ,ain't it's best point .....
(Lacks the deep red power,needed.... )
Still,it 'manages' pretty ok ....
CXA3000Kbioessay2.JPG

For reference ...
PHy absorption spectra.JPG
r_fr.JPG



--------------------------------------------------------------------------------------------------r_fr nature.JPG
--------------------------
Phytochrome and Light Control
of Plant Development

http://www.dls.ym.edu.tw/s46/(05-05-2008)Ch17.pdf

Cheers.
:peace:
 
Last edited:

Bueno Time

Well-Known Member
I realized that I forgot to hook up the Kill A Watt to my COBs after swapping drivers from MW LPC-60-1400s to the $13 JHX drivers.

CXA3070 3000K Z2 & Z4 bin one of each

w/ MW LPC-60-1400s 0.63 PF 121W 1.6A 191VA
w/ JHX "1500mA" drivers 1.00PF 119W 1.0A 119VA

Only 2 less watts even with what is supposed to be better efficiency drivers but a huge decrease in amperage pull with the power factor corrected drivers. I wonder if the current is a hair higher from the JHX drivers than the Mean Wells causing the wattage at the plug to be so close. I should hook up my multi-meter to see actual current output from drivers. Thing is I didnt test the current from the Mean Well but we can assume it is within spec since its Mean Well. For testing the current output of the driver, how do I do that properly just positive lead from MM on pos wire and negative lead from MM on neg wire from COB to test vF and positive lead from MM to pos from driver and negative lead from MM to pos on COB for amperage? Obviously switching the positive wire port for the amperage reading.

Ive used a multi-meter before for many things I just dont want to do something stupid and short the circuit out or fry something lol.
 

SupraSPL

Well-Known Member
Good data thanks BT! The current decreases a bit as the drivers warm up and the vF of the COBs decreases as they warm up and decreases as the driver warms up, so the readings can change a bit until the temps stabilize.

The JHX drivers put out 1420-1460mA once warmed up and are ~90% efficient. I have not tested a Mean Well LPC-60-1400 but the spec sheet says 85% efficient. I do have a Mean Well LPC-60-1050 and it puts out 1040mA at 88% efficiency (spec sheet claims 87%). So based on that, I suspect the JHX are driving the COBs slightly harder than the Mean Well.

Most multimeters have poor accuracy measuring current (and voltage and impedance for that matter) so don't be concerned if the measurements make no sense lol. What meter are you using? To measure current you can bug into the circuit at any point that is convenient and put your meter in series. Your description is correct.

There is one thing to watch out for that can confuse driver efficiency measurements. The multimeter adds some vF load to the circuit while measuring current, which can slightly change the drive current of the driver and changes the power draw, potentially skewing your calculations. If you check the kill-a-watt with and without the multimeter attached you can see this effect (if it affects the drivers drive current). For example, this is especially noticeable on the Fasttech "50W" drivers where a slight change in vF makes a huge change in driver current because the vF range is maxed out with a cxa3070 connected.
 

JavaCo

Well-Known Member
I realized that I forgot to hook up the Kill A Watt to my COBs after swapping drivers from MW LPC-60-1400s to the $13 JHX drivers.

CXA3070 3000K Z2 & Z4 bin one of each

w/ MW LPC-60-1400s 0.63 PF 121W 1.6A 191VA
w/ JHX "1500mA" drivers 1.00PF 119W 1.0A 119VA

Only 2 less watts even with what is supposed to be better efficiency drivers but a huge decrease in amperage pull with the power factor corrected drivers. I wonder if the current is a hair higher from the JHX drivers than the Mean Wells causing the wattage at the plug to be so close. I should hook up my multi-meter to see actual current output from drivers. Thing is I didnt test the current from the Mean Well but we can assume it is within spec since its Mean Well. For testing the current output of the driver, how do I do that properly just positive lead from MM on pos wire and negative lead from MM on neg wire from COB to test vF and positive lead from MM to pos from driver and negative lead from MM to pos on COB for amperage? Obviously switching the positive wire port for the amperage reading.

Ive used a multi-meter before for many things I just dont want to do something stupid and short the circuit out or fry something lol.
No that would be a short circuit on the amp test. You have to run either the positive side or negitive side threw the meter. When you plug into the amp jack on the MM you would take the positive from driver and run it to one lead then the other lead would go to the positive on the LED. Or you can do that on the negative side also. But never touch both positive and negative with the MM in amp mode you will cause a short circuit. Having said that testing CC drivers can be done but if you slip off while the driver is powered up it will damage or fry it. So it is quite risky testing current with a MM unless you can attach the wires so you know they wont come off.
 

Bueno Time

Well-Known Member
Good data thanks BT! The current decreases a bit as the drivers warm up and the vF of the COBs decreases as they warm up and decreases as the driver warms up, so the readings can change a bit until the temps stabilize.

The JHX drivers put out 1420-1460mA once warmed up and are ~90% efficient. I have not tested a Mean Well LPC-60-1400 but the spec sheet says 85% efficient. I do have a Mean Well LPC-60-1050 and it puts out 1040mA at 88% efficiency (spec sheet claims 87%). So based on that, I suspect the JHX are driving the COBs slightly harder than the Mean Well.

Most multimeters have poor accuracy measuring current (and voltage and impedance for that matter) so don't be concerned if the measurements make no sense lol. What meter are you using? To measure current you can bug into the circuit at any point that is convenient and put your meter in series. Your description is correct.

There is one thing to watch out for that can confuse driver efficiency measurements. The multimeter adds some vF load to the circuit while measuring current, which can slightly change the drive current of the driver and changes the power draw, potentially skewing your calculations. If you check the kill-a-watt with and without the multimeter attached you can see this effect (if it affects the drivers drive current). For example, this is especially noticeable on the Fasttech "50W" drivers where a slight change in vF makes a huge change in driver current because the vF range is maxed out with a cxa3070 connected.
Thanks for the info. I suspected the JHX drivers run the COBs just a hair harder than the LPC-60-1400s.

I just have a cheap multimeter it was like $20 ETEK 10709

http://www.amazon.com/Digital-Voltage-Current-Testing-Temperature/dp/B004LKDCY6
 

Bueno Time

Well-Known Member
No that would be a short circuit on the amp test. You have to run either the positive side or negitive side threw the meter. When you plug into the amp jack on the MM you would take the positive from driver and run it to one lead then the other lead would go to the positive on the LED. Or you can do that on the negative side also. But never touch both positive and negative with the MM in amp mode you will cause a short circuit. Having said that testing CC drivers can be done but if you slip off while the driver is powered up it will damage or fry it. So it is quite risky testing current with a MM unless you can attach the wires so you know they wont come off.
I thought that is what I typed. I meant pos of MM to pos driver an neg MM to neg driver for vF measurment and pos MM to pos driver and neg MM to pos COB for current.

Thanks for the warning, I might pass on the measurements. Supra has tested enough of them (JHX drivers) already anyway.
 

JavaCo

Well-Known Member
Java, I love that driver but I am having a problem with the seller. They sent me a shipment of drivers that are slightly longer than the original I tested. These new drivers were defective. Vf range only goes up to 25, they only put out 700mA @ 37vF and the operate at 83% efficient. To be honest I am angry at the seller. Normally I just eat it when I get a dud, but I felt tricked in this case and it was a whole batch so I pursued a refund. They made me return the drivers to China at my cost to prove my complaint, despite photos of the multimeters and I have been working with them for over a month to try and get (half) my money back. Hopefully other DIYers dont get anything from that defective batch.
View attachment 3177178

This driver on the other hand, has been performing great, puts out over 1400mA and runs up to 40vF. So far not a single dud out of 52 drivers, although they have yet to be tested for reliability.
Thanks for the heads up. Checked my 2nd batch out and they are longer then the first batch. Still have to test them out just been too damn busy here lately. But i need to do it soon you wait longer then 45 days on ebay and your SOL.
 

SupraSPL

Well-Known Member
I am still fighting with them for a refund. They told me their engineers checked out the drivers and that they work fine and offerred me $25 leaving me out $55 and nothing to show for it. I explained that they output 1400mA under 25vF but not above. They replied that their engineers will check them out again. Normally I get great service from a 99.5% eBayer.

Hopefully yours are not from that batch (this batch shipped quickly from the US, completely unprotected inside an envelope) If you have a power meter you will know if it is a bad batch if they are drawing ~35W each. If they are working correctly they should be drawing about 63W each.
 
Last edited:

DiyKindaGuy

Active Member
Out of 15 or so psu from ebay... ive blown two by accident and one died after three days and one thats running under powered...ive bought all my psu from ebay and ive had some psu for almost two years with no issues

★★★kushed_out★★★
 

JavaCo

Well-Known Member
I am still fighting with them for a refund. They told me their engineers checked out the drivers and that they work fine and offerred me $25 leaving me out $55 and nothing to show for it. I explained that they output 1400mA under 25vF but not above. They replied that their engineers will check them out again. Normally I get great service from a 99.5% eBayer.

Hopefully yours are not from that batch (this batch shipped quickly from the US, completely unprotected inside an envelope) If you have a power meter you will know if it is a bad batch if they are drawing ~35W each. If they are working correctly they should be drawing about 63W each.
Well doesn't look good, because that is a carbon copy of how my second order went. First Order took right at a month to get here and the second only took a week.

Going to make time today and test these out. I have a digital panel mount amp & voltage meter that goes up to 10 amps 100 volts dc built in shunt so I will wire that into the neg side. Haven't seen the kill-a-watt meter since the move two years ago. Beginning to lose hope on that one. If they test bad i will file a case with ebay. Might help you out if they have more they one person saying these drivers are mislabled or bad.
 

JavaCo

Well-Known Member
I was Robbed,....... tested two out so far. First one put out 35 vF 800 mA 2nd one put out 35 vF 650 mA. Then i hooked up one of the short ones from the first batch it put out 37 vF and was bumping between 1.4 amps and 1.2 amps. Something wrong with that one too. So i am fucked pretty much. If that fool ass seller wants me to ship this shit back to china i am going to grind this crap into tiny little pieces then take a shit in the box and send it .lol

On that note any discounts or sales on meanwell drivers anybody know about , please share.
 

SupraSPL

Well-Known Member
Updated flower room #1 to all passive cooled, 536W dissipation. All the COBs are CXA3070 3000K AB bin and some are running at 700mA some are running at 900mA. Some of the RWB modules are still using 3000K XTEs which will be swapped out for COB eventually.
IMG_0134b.jpg
 
Last edited:

churchhaze

Well-Known Member
One thing I've noticed is just about all the DIY led panel builders grow in dirt, and swear by it.

DWC would give so much more consistent (and likely better) results. It just strikes me as a bit odd that the majority of DIY's techies prefer dirt over hydro.

I can't wait to eventually try out the cxas or veros with a DWC.
 
Top