Drivers - What is the Knowledge!

How complex it is

  • I can do it with my eyes closed

    Votes: 3 42.9%

  • Kinda tricky

    Votes: 3 42.9%

  • I just guess generally

    Votes: 0 0.0%

  • Whoaaa! Get an electronics engineering degree first, sonny

    Votes: 1 14.3%
  • Total voters
    7
beepotron

beepotron

Active Member
Hello all

Firstly, thanks to everyone who freely offers up their knowledge and time in this particular forum, I've spent hours fascinated and becoming duly educated (if slowly) about the construction of LED systems.

I've done a bit of searching, but I can't seem to find a thread where this is covered, so I'm hoping someone can share what is probably very straight forward once you "get it".

How, in the most basic form, do you calculate the specs of the drivers you need? I'm looking at single COBs with single drivers (perhaps rudimentary and the easiest to work out for a novice), but also series, and theoretically, parallel arrangements... Systems with the minimum number of drivers seem cleaner to me. I'm particularly interested in high output COBs, but hoping the same principal applies to strings of 3W LEDs, etc. too.

I know there's loads of questions along the lines of "I want to power xyz arrangement of COBs, what meanwell driver do I need?", and these get answered every time with good information, but I'm after the knowledge to fish and feed my family every day without asking piecemeal each time.
 
Meinolf

Meinolf

Well-Known Member
You decide for the particular current you want to drive your particular LED-chip with. You look into the LED's datasheet to determine the needed voltage to accomplish that current. By multiplying the two values you get your chip's dissipation wattage. The driver you want to find needs to suit 3 variables then: It has to have the correct constant current, it has to provide it in a voltage region that contains the estimated voltage of your chip and the multiple of those two variables should be as close as possible to its max rated power, because that helps for efficiency. Use google then.

Sometimes you can trick around with series or parallel circuitry. In case of a series circuit you need to provide (higher) voltage that is then shared by the chips (same current then). In case of a parallel circuit you need to provide (higher) current, that is then shared by the chips (same voltage then).

Does this answer your questions? I can tell you, it can become a pain in the ass to tetris around with datasheets. I'd suggest you save a lot of time and money and go straight for one of those pre-tested combinations: https://www.rollitup.org/t/diy-leds-how-to-power-them.801554/page-74#post-11629572
 
JimmyIndica

JimmyIndica

Well-Known Member
beepotron said:
Hello all

Firstly, thanks to everyone who freely offers up their knowledge and time in this particular forum, I've spent hours fascinated and becoming duly educated (if slowly) about the construction of LED systems.

I've done a bit of searching, but I can't seem to find a thread where this is covered, so I'm hoping someone can share what is probably very straight forward once you "get it".

How, in the most basic form, do you calculate the specs of the drivers you need? I'm looking at single COBs with single drivers (perhaps rudimentary and the easiest to work out for a novice), but also series, and theoretically, parallel arrangements... Systems with the minimum number of drivers seem cleaner to me. I'm particularly interested in high output COBs, but hoping the same principal applies to strings of 3W LEDs, etc. too.

I know there's loads of questions along the lines of "I want to power xyz arrangement of COBs, what meanwell driver do I need?", and these get answered every time with good information, but I'm after the knowledge to fish and feed my family every day without asking piecemeal each time.
Click to expand...
Chip data sheet will tell ya don't run it below this< Don't run it above this? and we recommend u run it here! Best thing to do is run em towards the topend and change chips after a year with something better!
 
JimmyIndica

JimmyIndica

Well-Known Member
Don't drive em soft because your yield will suffer! Unless your stacking underpowered chips over canopy 6 inches? run em hard and put em away wet!
 
SupraSPL

SupraSPL

Well-Known Member
You can run them as soft as you like without decreasing yield, as long as you use more COBs. CXAs and CXBs respond well to that approach. If you want to run hard Veros might be the best bet. You just have to find the value point that suits your situation and there is more to the story than electrical savings, although some locations have very high electrical cost. The more efficient you are, the less drivers and heatsinks it will take to do the same job. If running at higher efficiency reduces or eliminates your AC costs that can make a big difference. Most importantly of all, if running at high efficiency increases the yield in your space it pays for itself in days/weeks.

current droop.png

Vero 18 29 V2.png
 
Last edited:
salmonetin

salmonetin

Well-Known Member
....Dissipation Power...
...for example, the Power Dissipation of a 100 W LED Driver is 11.1 W if the Driver is 90% Efficient, but increases to 25 W if the Driver is only 80% Efficient.

...from...

http://www.astrodynetdi.com/power_supplies/guides/led-drivers/faq/

...link is only for examples or ideas...

:peace:

...maybe we need more info about protections on the driver side...

overcurrent, overvoltage, overtemp, emifiltered, opencircuit protections, etc

saludos
 
Last edited:
Greengenes707

Greengenes707

Well-Known Member
The concept of intensity needs to be understood here.

More intense light does not move faster or with with any more force/power. Intensity is more like crowding photons into an area than being "more powerful". All photons move at the same speed(speed of light)
Which is why running any type of LED softer will give you more yield per the same wall wattage but obviously more chips are needed to get the wattage...and no loss in penetration. Actually gains it per watt. There are more photons per watt...and you crowd enough light/cob/single dies to get to the desired intensity.

Which also brings up another point/issue. And that is the inverse square law/penetration based on distance from the source. Light doesn't get less intense per say...just less dense/crowded(over massive universe distances will lose some energy). But if you confine the space and reflect most of it...distance from the source is not an issue or a penetration figure to use. In a high reflective case...the overall PPF will represent the intensity factor and correlate to a penetration. Whether it's 5umols of green light, or 500umols...about the same percentage of is still transmitted(passes through) and will "penetrate".


To stay on topic...
Series is positive to negative and I accumulates Vf and doesn't change the current of the string no matter how long it is, as long as your voltage will cover he series.
Parallel is all positives together, and all negs together. In the way, current is divided by the number of series strings you have and doesn't require more Vf when adding more strings in parallel. Just more current because it keeps getting divided per.

The two methods can be combined. Running multiple series strings in parallel.
 
Last edited:
beepotron

beepotron

Active Member
It's all getting a bit confusing, with this talk of massive universe distances, dark matter, and super massive black holes...
 
You must log in or register to reply here.
Top