DiY LEDs - How to Power Them

[churchhaze]

DC power is calculated as volts times amps and can be easily measured with two multimeters. A high-resolution multimeter measures the current delivered to the load and a standard multimeter measures the power supply’s output voltage.

AC input power can’t be calculated simply as RMS input voltage times RMS input current because of the differences in phasing between the voltage and the current in an AC system, designated as ϕ and seen in figure 1. If we did this, it would yield a result that’s too high because the value would include both the power converted by the power supply (P) and the power that’s returned to the AC source (Q). (...)

For a power supply that claims 0.99 power factor (on 115VAC) you can assume the phase lag/lead between wall voltage and current into the driver is negligible and can basically just multiple the AC voltage and current treating the the driver/cob combo in total as a resistive load.

Or am I wrong about this? (that's always what I assumed)

 

[SupraSPL]

@sds Lets assume you are correct and there is a systemic error in my measurement technique (RMS Watt meter). If that is so, my numbers could not be used for absolute figures, but they could still be used for relative comparisons. So that argument does not invalidate my data.

Also worth mentioning, my measurements agree closely with the Mean Well data sheets. I have no reason to suspect that it is inaccurate. The Watt meter measures power factor. When it returns a Watt figure, power factor is already applied to that figure, it is not returning a VA figure.

One example to pick on, the Mean Well LPC-60-1400. Perfect voltage match for the CXA3070 at 1.4A. No power factor correction, 85% efficient and cost $18.

The generic counterpart puts out 1.45A, IS power factor corrected, 89-91% efficiency and cost $13-$15. Not one has failed so far that I can recall. Mean Well ought to copy the generic design and do us a favor, no?

 

[alesh]

@sds Lets assume you are correct and there is a systemic error in my measurement technique (RMS Watt meter). If that is so, my numbers could not be used for absolute figures, but they could still be used for relative comparisons. So that argument does not invalidate my data.

Also worth mentioning, my measurements agree closely with the Mean Well data sheets. I have no reason to suspect that it is inaccurate. The Watt meter measures power factor. When it returns a Watt figure, power factor is already applied to that figure, it is not returning a VA figure.

One example to pick on, the Mean Well LPC-60-1400. Perfect voltage match for the CXA3070 at 1.4A. No power factor correction, 85% efficient and cost $18.

The generic counterpart puts out 1.45A, IS power factor corrected, 89-91% efficiency and cost $13-$15. Not one has failed so far that I can recall. Mean Well ought to copy the generic design and do us a favor, no?

Which Watt meter are you using?

 

[SupraSPL]

The Ensupra, not my company just a coincidence
http://www.amazon.com/Ensupra-Electricity-Monitor-Reduce-Energy/dp/B007KX2YXQ

I have a pair of them and they agree very closely. Many interesting things turned up, surge protectors that take energy, timers, wireless outlet controller, even GFCI outlets consume some energy. The night lights that turn them selves off during the day, they take just as much electricity when they are on as they do when they are turned off. Motion light sensors and dusk/dawn sensors constantly consume energy as well.

 

[Jeeyah]

@SupraSPL I could use your help when you get a second.

I want to build a veg light. Does this match up OK?

Meanwell HLG-120H-C700A
(4) Vero 18 5000K

Can you recommend a good power supply for the fans? I used an old cell phone charger last time. Thank you in advance.

 

[Actionbone]

@SupraSPL I could use your help when you get a second.

I want to build a veg light. Does this match up OK?

Meanwell HLG-120H-C700A
(4) Vero 18 5000K

Can you recommend a good power supply for the fans? I used an old cell phone charger last time. Thank you in advance.

for 4 vero 18 you need hgl-80h-c700a

 

[Jeeyah]

for 4 vero 18 you need hgl-80h-c700a

Thank you so much. This has to with adding up the voltage in series right? The voltage is too low with 4 on the HLG120? I haven't messed with electronics in 15 years. I need to go back to ohms law or some shit. Lol

 

[Jeeyah]

I went back to ohms law. It looks like the voltage and power have a direct relationship. If I use your example, I take the 28.35 volts x 4= 113.4. (P=VI). 113.4x.7=79.38 watts. 80 watt driver pushing 79.38 watts. If I match a driver/COB's with a lower voltage, the power drops. Clear as mud!

 

[EfficientWatt]

I is for current.

So I = 700mA = 0.7A.

V = 4 Vero @ 28.35V = 113.4

Now you can use P = I * V

P = 0.7 * 113.4
P = 79.38 W

Welcome back to school

 

[alesh]

Pretty interesting drivers:

Spec. Link
KLA-030 series: http://www.isweek.com/product/30w-led-driver_677.html
KLA-045A1 series: http://www.isweek.com/product/45w-led-driver_683.html
KLA-060A1 series: http://www.isweek.com/product/60w-led-driver_684.html
KLA-080 series: http://www.isweek.com/product/80w-led-driver_687.html
KLA-100 series: http://www.isweek.com/product/100w-led-driver_689.html
KLA-120 series: http://www.isweek.com/product/120w-led-driver_704.html
KLA-150 series: http://www.isweek.com/product/150w-led-driver_705.html

 

[UKpeanuts]

http://www.gzkaisen.com/en/disproduct.asp?fid=27&id=292

KLA-120 The Internet of things series --

that suggests I can talk to it via, ethernet or wirelessly. that would be really neat.

 

[testiclees]

Folks I have an optic vero 360. It uses one driver for each cob. So there are 4 vero 29 cobs and 4 meanwell drivers.

Learning about the sweet lights you guys built is contagious. I've got the bug. I would like to plan my own diy. I'd like to power 6 vero 29s to get about 70-80 dissipated watts each.
Is it possible to power this setup with just two drivers or is my best and safest bet to use a discrete driver for each cob? A dimming function would be desirable but not essential.

I have read through this entire thread 1x. I am not in a rush and I have untill Autumn to complete my project. My budget is not super tight.

COMPLETELY OPEN TO MORE INFORMED APPROACHES OR ALTERNATIVES.

THANKS FOR SHARING EXPERIENCE & EXPERTISE

 

[Jeeyah]

Folks I have an optic vero 360. It uses one driver for each cob. So there are 4 vero 29 cobs and 4 meanwell drivers.

Learning about the sweet lights you guys built is contagious. I've got the bug. I would like to plan my own diy. I'd like to power 6 vero 29s to get about 70-80 dissipated watts each.
Is it possible to power this setup with just two drivers or is my best and safest bet to use a discrete driver for each cob? A dimming function would be desirable but not essential.

I have read through this entire thread 1x. I am not in a rush and I've it untill Autumn to complete my project. My budget is not super tight.

Check out these charts. https://www.rollitup.org/t/cob-efficiency-spreadsheets.865238/

I'm pretty sure the LPF-90 will work. I don't think the drivers that run multiple COB's will be able to run at 2.1 amps. I'm just a newbie though. Lol

I actually want to get the new Vero 29's. But, I don't know if they're available. It seems moving forward we should aim for 40+% efficiency. I noticed the old 5000k is super efficient.

 

[SomeGuy]

draz, I am currently building a veg /mother light with vero 10's and bought this led driver https://www.fasttech.com/products/1714902 from fasttech.
It is powering 4 ea vero 10's in series at 297ma. Not dimmable but works great. took about 4 weeks to receive them but the price is right. Gl OG

I built two veg lights. Each with five 5000k vero 10 and one of those drivers you linked each. I did notice these drivers have a noticeable pulse in the peripheral vision. Its fine but may be something to note if you would be exposed to the fixture for extended periods.

 

[SupraSPL]

Learning about the sweet lights you guys built is contagious. I've got the bug. I would like to plan my own diy. I'd like to power 6 vero 29s to get about 70-80 dissipated watts each.
Is it possible to power this setup with just two drivers or is my best and safest bet to use a discrete driver for each cob? A dimming function would be desirable but not essential.

3 dimmable options. The HLN is 91% efficient and the HLG are 94% efficient:

Vero29 3K V2.0 @ 2.1A on 1 HLN-80H-42, you would get 79W ea at 40-41%.
CXB3590 3K CB @ 1.05A, you can run 2 CXB3590s on 1 HLG-120H-C1050 and you would get 76W/COB at 46%.
CXB3590 3K CB @ .7A, you can run 4 CXB3590s on 1 HLG-185H-C700 and you would get 49W/COB at 52%

 

[SupraSPL]

I built two veg lights. Each with five 5000k vero 10 and one of those drivers you linked each. I did notice these drivers have a noticeable pulse in the peripheral vision. Its fine but may be something to note if you would be exposed to the fixture for extended periods.

I noticed it on that driver also when I was walking past the lights it is noticeable. Glad I am not crazy LOL
https://www.fasttech.com/products/1714902

 

[testiclees]

Thanks S,

I prefer 4k for the vero 29. Will HLN-80H-42 be cool for 3 of those?


EDIT
I've got the data sheet for the newer vero29. It looks to me that there is no difference between driver requirements 3 & 4k..So I believe I'm good to good. Please give a heads up if I missed something.

Many thanks.

 

[SupraSPL]

Thats correct, you can choose either version or color temp and will be compatible with that driver.

I have to retract my speculation about the color temp shift caused by glass lenses. I have been searching for data to back up that observation and found nothing. So I think it may be an illusion caused by the higher intensity on the lens side. When I move the unlensed light much closer to the target the color temps match more closely. Not sure if that has any effect on your color temp choice, the 4000K may prove to be better or equal, I am still unsure about that.

 

[ShyGuru]

I'm planning my cob build and would like to ask the advice of the experts before I start buying parts. My plan so far is to use 8 cxb3070 on rosewill rcx-z1 coolers set up in a 3-2-3 staggered pattern over a 36x22 grow space with 54 inches of total height. I was considering running them in series on two mean well hgl-185-1400 which comes to 400 watts. This is an arbitrary number since I was going to run a 400hps in the space. But at just shy of 60w per sq ft is that overkill? I was wondering what you guys would advise. Thanks in advance!

 

[alesh]

I'm planning my cob build and would like to ask the advice of the experts before I start buying parts. My plan so far is to use 8 cxb3070 on rosewill rcx-z1 coolers set up in a 3-2-3 staggered pattern over a 36x22 grow space with 54 inches of total height. I was considering running them in series on two mean well hgl-185-1400 which comes to 400 watts. This is an arbitrary number since I was going to run a 400hps in the space. But at just shy of 60w per sq ft is that overkill? I was wondering what you guys would advise. Thanks in advance!

It's actually 400W over 5.5 sqft = 72.8W/sqft. It's totally overkill with such highly efficient LEDs.
edit: If you are highly space-restricted, you'd get maximum out of your space although not from the LEDs.