Genuine Chinese Sellers

[Futuregrow]

No, a sphere test isn't needed. You can quite readily establish the facts without one. There is zero confusing the lm561b and c, I'll upload the info later

[Futuregrow]

Yes a sphere test is what is needed. The b+ diodes and lower bins perform great as well. You just don't get the same Effeciency.

Here we go man, here's a screencap I took from a Samsung presentation video on the topic (I'll put the youtube link at the end), You can see the size of the emitter in a C diode vs a B, it's noticeably bigger, and can be seen with the naked eye:

Mine have LM561C, it's easy to tell they have the large die emitter, here's a couple of pics I took through the loupe:

On this one I moved the light closer, zoomed in, and used HDR mode, so it went a little grainy, but the contrast improved and shows the outline of the emitter better.

Test conditions as follows, measured with multimeter:

250 x 250 PCB
320 pcs Samsung LM561C S6 3000K LEDs, 40 parallel strings of 8 diodes.

uPowerTek BLD-050-C210-ENS driver - 50 W constant current driver (the 50 W driver is not as efficient as the 400 W drivers, which are around 92-94% efficiency rated), I just have this one for testing individual boards.

54.2 W draw from the wall, 2.11 A driver output, 21.8 V measured across the board.

2.11 x 21.8 = 45.998W

The board is running at 46 Watts.

2110 mA / 40 = 52.75 mA per string of 8 diodes.
21.8 V / 8 = 2.725 Volts per diode

Now the datasheet specifies the A1 bin to be 2.8 to 2.9 volts (IF = 65 mA), and the AZ bin to be 2.7 to 2.8 volts... however in my case IF = 52.75 mA, which is going to give a lower forward voltage, and 2.725 V seems about right for the either bin, depending on where within that voltage range the diode itself falls (due to manufacturing variation).

If I use the Samsung component calculator using the values I have measured, and choose A1 bin, I get the following result, which states 2.82 V and 47.6 W for 320 diodes:

Now I'm not sure if there are any losses associated with the copper of the board itself, or to what scale they'd have an effect, however if I switch to AZ bin I get pretty much exactly what I've calculated from my measurements taken from the test board... 2.72V and 45.9 W:

Now in either case, I have the following output values:
9286 lm total output from the board, and efficacies as follows assuming either of the bins:
A1: 195.1 lm/W
AZ: 202.2 lm/W

Now if I refer to the PAR multiplier thread on RIU forums, as stated by Stephen from HLG himself, the par multiplier for Samsung LM561C 80 CRI diodes (derived from results out of the integrating sphere) is 0.014505

So what this means is you take lm/watt and multiply it by 0.014505 to get the efficiency in uMol/j

So for my numbers established above, here's the maths:

A1: 195.1 x 0.014505 = 2.83 uMol/J
AZ: 202.2 x 0.014505 = 2.93 uMol/J

The supplier said these are A1 bin diodes and that would make sense assuming a lower forward voltage at lower forward current... however the measured numbers make it appear to be exactly AZ bin, but those calculations haven't been modified to account for any variance due to copper on the board or anything... which is unlikely to have a big effect imo.

So I'll leave it up to you guys to decide. Worst case scenario, A1 bin, 2.83uMol/J (very efficient), or AZ bin, 2.93 uMol/J (very, very efficient).

I took a video of this board testing and filmed each part of it to show the driver, measurements, etc... happy to post it up if anybody doesn't believe me.

[CrazyFool]

Thanks for the insightful comments guys.
So a visual examination can tell you if its an lm561c.
A voltmeter can tell you the voltage binning.

What about if its s6, s5, s4 etc?
Do we need a sphere for that?

[Grower]

What about if its s6, s5, s4 etc?
Do we need a sphere for that?

that is the binning.
Imho you can't be sure of the binning by testing the whole board, you should desolder a led and test it

[CrazyFool]

I can easily desolder an led and test it with a voltmeter
Any noob friendly step by step instructions any where?
So basically we run the diode at voltage x and if the current across the diode is y then we know its genuine?
What about overlap? Surely there'll be many diodes/brands/generations that run around similar parameters?


Where can I find a chart that compares current vs voltage the lm301b/lm561c data sheets only have 1 chart. Surely there should be a separate graph for s4, s5,s6 etc?

Also, A1 AY and AZ aren't separate products? Is it just a voltage range? Every LED can be run as either a1 ay or az?

Thanks again

[Grower]

I can easily desolder an led and test it with a voltmeter
Any noob friendly step by step instructions any where?
So basically we run the diode at voltage x and if the current across the diode is y then we know its genuine?
What about overlap? Surely there'll be many diodes/brands/generations that run around similar parameters?


Where can I find a chart that compares current vs voltage the lm301b/lm561c data sheets only have 1 chart. Surely there should be a separate graph for s4, s5,s6 etc?

Also, A1 AY and AZ aren't separate products? Is it just a voltage range? Every LED can be run as either a1 ay or az?

Thanks again

there is the Samsung online calculator

[CrazyFool]

Thanks for all the help thus far guys.
this is making much better sense.

I've been doing some more reading and apparently, it is very easy to identify s5 or s6 with a loupe or microscope. the black dot is still slightly visible on s5 whereas it shouldn't be visible at all on s6.

This makes me a lot more comfortable.

So basically there are separate LEDs that are A1 AY and AZ. You can't run every LED at a lower AY voltage range. I have to specifically ask for AY diodes. These can be verified with a voltmeter.

Is my understanding correct?

Thanks again

[Futuregrow]

Yeah they're sold by voltage bin. Check the attached pic:

[Grower]

You could also get a genuine Samsung strip off digikey or arrow and compare its diodes to your boards

[CrazyFool]

Thanks for the reply fella.

Yeah they're sold by voltage bin. Check the attached pic:

So there's no way of knowing exactly which voltage bin you're getting, you're only choice is XA and XK?

You could also get a genuine Samsung strip off digikey or arrow and compare its diodes to your boards

Yeah, tbh there's plenty of microscopic photos people have posted of genuine s5/6 diodes.
Its an easy comparison. I'm feeling pretty good about this.

I'm almost ready to place my order.
I just need advice about drivers. And in what orientation I should have my diodes wired - series/ paralel etc.
I'm mainly looking to have lots of small strips for decorative lighting. Each strip (8-12 diodes) delivering roughly 250 lumens or 1-2w. How do I drive these small runs?
I will also eventually run longer strips for growing my strawberries/mint/microgreens. I imagine that will be a lot simpler.

Thanks again

Crazy