Introducing CobKits.com - specializing in DIY and Citizen COBs

speaking of which i just added 200 cxm22 3500k 90 cri (was out of stock)

also updated inventory on gen5 citi 3000k 90, 3500k 90, and 4000k 90 which were out of stock
 
Great so you suggest me to buy 50w diodes and dim them @ 35 right?
they are not 50W, they are 50V (nominal, actually 48-54V based on current), they can run in excess of 100W each so 35-50W is very efficient

for both veg and flower 3500k 80 or 3500k 90
 
they are not 50W, they are 50V (nominal, actually 48-54V based on current), they can run in excess of 100W each so 35-50W is very efficient

for both veg and flower 3500k 80 or 3500k 90
Awesome! Thanks! I will send an e-mail to the address on your website so we can keep up the conversation private and I don't flood this topic with posts that may seem useless to other members:roll:
 
One question: is the Meanwell HLG480H-C1400 a good option? I've read around that constant current drivers might be a good option considering the life of the LED diode. What do you think?
 
your diodes do not know whether you're using a constant current or constant voltage driver. As long as the current is in the appropriate range it doesn't make a difference

Driver draw question.

Does it matter with cc driver wether you are driving one cob or multiple cobs? I have 3 of the same driver (meanwell owa-60U-42) two seem to draw around 65 watts from the wall when driving a single cob, the other I have connected to 5 cobs in parallel - it draws only 55 watts.
20171027_122355.jpg
Although there is a cpu chiller involved, it dosent seem to have any connection to why the driver is drawing less watts from the wall.

Any insight?
 
Does it matter with cc driver wether you are driving one cob or multiple cobs? I have 3 of the same driver (meanwell owa-60U-42) two seem to draw around 65 watts from the wall when driving a single cob, the other I have connected to 5 cobs in parallel - it draws only 55 watts.
at same total current, more cobs = lower current per cob = lower voltage per cob = lower driver voltage = less watts
 
at same total current, more cobs = lower current per cob = lower voltage per cob = lower driver voltage = less watts
So I'm trying to let this digest....

Since the driver is not a constant voltage (I thought it was and your response made me go back and look) the chips or whatever is hooked to it will only draw the voltage needed resulting in a lower wattage?

Does this mean that if using a driver with an adjustable but constant drive voltage (hlg-320h) it would be more efficient (both with power consumption and light generation) to drive at the lowest possible voltage to ignite the led? I know there are some stats in the cree data sheets but it seemed lumen generation was more or less tied to the ability to maintain a lower operating temperature and pushing the chip harder, not necessarily more efficiently.

edit: to answer my own question (sort of) since I am not able (or at least confident enough) to measure dc side voltage there seems to be an intensity/power consumption plateau. Voltage under plateau - less power consumption but less led output. Sweet spot is somewhere, hoping @CobKits already knows where.
 
Last edited:
Since the driver is not a constant voltage (I thought it was and your response made me go back and look) the chips or whatever is hooked to it will only draw the voltage needed resulting in a lower wattage?
they are constant voltage/constant current. if your load can take all the current, voltage will drop according to array demand (see chart on datasheet). cobs at a given current can ONLY operate at the corresponding voltage per datasheet

Does this mean that if using a driver with an adjustable but constant drive voltage (hlg-320h)
i assume you mean something like hlg-320h-48A, etc as opposed to HLG-320h-C1400A

it would be more efficient (both with power consumption and light generation) to drive at the lowest possible voltage to ignite the led?
LEDs are *always* more efficient than run softer. most driver efficiencies are pretty flat down to 50% load so even if you reduce the load a bit, driver efficiency loss should be imperceptible (92 vs 94% or whatever). the cob efficiency bump will trump that by a significant margin.

I know there are some stats in the cree data sheets but it seemed lumen generation was more or less tied to the ability to maintain a lower operating temperature and pushing the chip harder, not necessarily more efficiently.
every cob data sheet shows relative flux vs current. and they are all a curve, meaning at lower currents the lumen/watt (PAR/watt) increases. thats why people dont attempt to run cobs full tilt

but you are right, its all about temperature. even if you keep your heatsink at teh same temp, at higher currents there is more heat for the cob to move from dies to chip to heatsink so current is a major factor

edit: to answer my own question (sort of) since I am not able (or at least confident enough) to measure dc side voltage

thats a $10 voltmeter measuring a relatively safe circuit. i would consider the ability to measure DC voltage fundamental to DIYing cobs. there are tons of vids on youtube demonstrating how to use a voltmeter and you can always ask here

Sweet spot is somewhere, hoping @CobKits already knows where.
sweet spot is determined by your budget. for a given total rig wattage, the more chips you use, the lower the current per chip, the more efficient the rig is

700-1750 mA are popular currents for growers
 
i assume you mean something like hlg-320h-48A, etc as opposed to HLG-320h-C1400A

You are an insightful mofo. Yes a hlg-320h-36a.
That graph of current vs. voltage makes more sense to me now, thanks.

I originally was voing with the 42v drivers thinking there would be an easy effecient way to drive both the led and heat managment (i.e. fans) but that was a futile effort.
When I last looked at the cree data sheets they listed incomplete info, some had typical voltages, others had maximum but no minimum, this is why I did not see measuring the voltage as critical being that the adjustzble range was within the operational spec of the led chips I was using at the time (cxa/b3070), upgraded to 3570s since original build.

I've just stoner momemnted morenthan once and done thing like crossing the voltmeter probe across two terminals, letting the smoke out and causing more harm than good....just my own nurocees getting the best of me.

Thanks for the insight
 
Back
Top