DIY led primer? Where do i start?

[h2so4hurts]

I run my blues at 75% and white at 65%. COBs can be challenging to get a good color mix, but they're much more convenient than single puck LEDs

 

[oreo54]

Power supply is a 54v Meanwell (HEP-600-54 ). High efficiency and plenty beefy.

Recommended top voltage on the driver chip is 48V...
Seems to work for you but def. not recommended

 

[Sayn3ver]

If you wire them in parallel then you're only limited by the output wattage and the total voltage per string.

Is it common for this type of lighting application to wire them in parallel?

In text book examples of parallel circuirts with wall supply power sources, all loads will see the same voltage, but current varies. How does that work out with a current limiting driver?

I have to imagine parallel wiring presents alot more wire to manage and dress up as well

 

[oreo54]

However, The driver will limit what voltages it can take?

Yes.. it will change the voltage on a sting till mA setpiont is reached. Then will regulate from there.

Their dual driver chips seem to only allow up to 48v input and after their voltage drop out out 45v.

Call it driver losses.. Needs a voltage differential of 3V..or so..depending on the driver..
15 leds that are at 3V (f) at say your target current of 1A is max. So 45V.
14 would make more sense and it does depend on the ACTUAL diode.

Add one more LED and the driver will just top out at 45V and underpower the string.. i.e won't reach 1A in general

So in theory I could put 3x channel 1's on a single driver. I get 2 drivers per chip so if I wanted all 4 channels separated for dimming purposes I would need 2x of their dual driver chips for three full power 48w acrostar leds off a single 48v power supply (assuming I can find a large enough ampacity psu.

Not that's a bit tricky.. Easiest way is just add up the mA of your drivers to guess at the current capacity needed.

To get exact would require measuring every string voltage at the drive current of the driver in a working array..
Calculating watts and them matching..

I suppose the next question then is, they note max efficiency is when input and output voltages almost match.

More of a minor issue..more important if the driver uses linear chips that suck upp the extra voltage the old way i.e heat..
Steves drivers use such chips.. and it means the string should match the power supply closely or the driver chip can over heat..
Running a 48V ps and lighting one 3V diode would burn out a linear driver (well probably)since it has to dissipate 45V's
W/ Meanwells and I assume Bluacros only issue is a loss of efficiency..
Driver chip just turns off/on w/ more off than on creating an average voltage of 3V. Less heating.

So if i needed to go to 4x 48w acrostars I basically would need four of their dual driver chips if I'm running all four channels separated?

And then at that point my efficiency is then going to be going down as each driver is then less loaded?

ehhh. usually it's more of a power supply thing than a driver thing though they both do that a bit.
Again, personally, splitting hairs here..

If I'm understanding correctly then it looks except channel 4 @ 9v, the rest benefit from being groups of 3.

Although their description says a pair of dual driver chips can power 5 of these? I guess they combine channels?

you can do whatever you want as long as the V(f) adds to be under max voltage of the driver at that current..

Am i reading the driver description correctly as well. I can tune each individual drivers current output with an onboard potentiometer?

Yes, driver has a variable set point based on the pot setting.

 

[oreo54]

Is it common for this type of lighting application to wire them in parallel?

In text book examples of parallel circuirts with wall supply power sources, all loads will see the same voltage, but current varies. How does that work out with a current limiting driver?

I have to imagine parallel wiring presents alot more wire to manage and dress up as well

No, and not usually recommended w/ constant current.
Now constant voltage arrays are run series/parallel.. but few do DIY that way.
See w/ parallel strings one driver will split the current between the 2 (or more) strings..
If one of the strings fails open the remaining string gets added current.

you can also look up current balancing circuits..
Let's just say NOT recommended in general..

Instead of 2 strings on one driver you get 2 drivers 2 strings..at an added cost of about $8..

Sort of correct way to run LEd's w/ current balancing circuit.

 

[bblumberg]

Recommended top voltage on the driver chip is 48V...
Seems to work for you but def. not recommended

Hmm. My understanding is 9-56v input and 2-52v output for these drivers

https://www.meanwell.com/Upload/PDF/LDD-H/LDD-H-SPEC.PDF

which is why i selected that particular power supply.... Am I missing something here?

 

[oreo54]

Hmm. My understanding is 9-56v input for these drivers

https://www.meanwell.com/Upload/PDF/LDD-H/LDD-H-SPEC.PDF

which is why i selected that particular power supply.... Am I missing something here?

Sorry, thought you were referring to the bluacro ones..which the op was considering..
Just skimmed your post
My bad..

 

[bblumberg]

Sorry, thought you were referring to the bluacro ones..which the op was considering..
Just skimmed your post
My bad..

Oh, OK. I thought for a minute I had picked the wrong PS here.

 

[Sayn3ver]

Meanwell makes the ldh line of drivers that can step voltage up.

At what point do you fry leds? I realize most are using/need 3.3v. i know these volt drops add in series. So if you theoretically need 120v for the string, doesn't the first led in series see the full 120v? Or due to the nature of the LEDs, they take their 3.3v and pass the rest along since they are not operating like a resistor?

There still has to be some kind of safe upper voltage limit due to insulation/isolation of the circuit pathways within the diode and on the chip.

 

[oreo54]

Meanwell makes the ldh line of drivers that can step voltage up.

At what point do you fry leds? I realize most are using/need 3.3v. i know these volt drops add in series. So if you theoretically need 120v for the string, doesn't the first led in series see the full 120v? Or due to the nature of the LEDs, they take their 3.3v and pass the rest along since they are not operating like a resistor?

There still has to be some kind of safe upper voltage limit due to insulation/isolation of the circuit pathways within the diode and on the chip.

you fry them when you raise th potential (voltage) high enough that they draw too much current and over heat .
See thermal runaway..

Actually if you can cool them then that point gets higher.. Like a diode in Liq N..
I suppose eventually you get photon damage to the crystal lattice..

When a junction gets warmer, the current through it at a given voltage will increase. The increased current in turn heats the junction further, and the problem gets worse. Eventually, if nothing limits the current, the junction will fail due to the heat. Because of thermal runaway, it’s important to use some current limiting circuit even with a regulated voltage source. A current limiting resistor is the typical solution for most hobbyists, while circuit designers may prefer to use a current source based on one or more transistors.

I'm sure it's more complicated than that but the bottom line again is how much current it can take is dependent on how well you cool it. Keep in mind the diode design itself can limit this i.e good or poor heat transfer built in.


Need to sort of look at it like you have an upper current limit..

You can have a 1000V power supply but as long as you limit the current to the diodes properties.. not an issue...