F Series driver recommendation

[kony brado]

Yeah, he reminded me that this week Arrow makes a $ 50 off sale on purchases over $ 200. Discount Code: fifty
Shopping Cart is already filled.
There are still a few XP-E HEs, but I get them somewhere else.

@kony brado
Many thanks for the hint!

Glad it worked for you random.
ill keep posting sales from all the companies on my profile page,can save folks some money so im happy

 

[kony brado]

From the data sheet, "Note 7) Temperature at which performance is specified."
The data sheet does not say the temperature will be 65C° at 1.12A, it is saying what the specs are at an operating temperature of 65C°

Yes ,they say Temperature at which performance is specified measured at tc point(measuring point ).
so they drive 1.12 A ,and get 65c at the measuring point ,that sounds logical (i don't think they put a fan on the stripe) .most companies measure at ambient temp of 25c(i don't think samsung specify this).sorry maybe i did not get your point as English is not my language ,i did not think they have a garenty for the temp,its just their measurement .do i make sense ?pleas correct my if i got something wrong
thanks

 

[Randomblame]

Yeah,
all in all it saves me 75 bucks compared to digikey.Thats a lot on a 325$ bill...23% less!!
Many thanks, mate! Such offers are allways much appreciated!

I found a cheap source for Cree XP-E HE C01 top bin photoreds on 20mm stars(19,90$/10pcs) but have to verify if it's really top bin diodes. I've already ordered a few and if they are brighter as my current XP-E's(901 bin) I will let you all know where to get them.
I only have to wait until they arrived... economy shipping from china... can take a while, LOL!

 

[kony brado]

sound good,yes
Does the xp-e have some marking on them for bin verification?

 

[Randomblame]

Nope, unfortunately not! But I will use my lux meter and a multimeter to compare both bins at 1w and 2w.
C01 bin should be ~20+% brighter as 901(350mW vs. 425mW). Let's see!

 

[1212ham]

Yes ,they say Temperature at which performance is specified measured at tc point(measuring point ).
so they drive 1.12 A ,and get 65c at the measuring point ,that sounds logical (i don't think they put a fan on the stripe) .most companies measure at ambient temp of 25c(i don't think samsung specify this).sorry maybe i did not get your point as English is not my language ,i did not think they have a garenty for the temp,its just their measurement .do i make sense ?pleas correct my if i got something wrong
thanks

I believe they run it at 1.12 amps, regulate the temperature to 65C, and measure the other specs.
I hope I make sense. This might be a good time for one of the experts to jump in.

 

[Rider509]

I tend to think you're right, 1212ham. I too have wondered if they spec'd it at likely operating temps inside a fixture rather than free air temps since one of the intended applications is for retrofitting existing fluorescent light fixtures.

edit: def not one of the experts! LOL

 

[Randomblame]

I believe they run it at 1.12 amps, regulate the temperature to 65C, and measure the other specs.
I hope I make sense. This might be a good time for one of the experts to jump in.

Nope,
they run them at 1120mA and the typical temps after stabilization is 65°C on the messuring point. There is no way to regulate the temps without changing the current.
If you run them at 720mA(each diode gets 80mA like on the H-Series strip) the stabilized temps would be 50°C(maybe 2 or 3° more, because of more diodes on the F-Series strips).

 

[Randomblame]

Like you, my head was about to explode when I started looking at all this. But after a while, it all seemed to make sense. Have a look at the Forward Voltage vs Forward Current graph on page 21 of the LM561c spec sheet: http://www.samsung.com/global/business/business-images/led/file/product/lighting/201511/Data_Sheet_LM561C_Rev.3.2.pdf

LEDs are resistors, so the more current you put through them, the more resistance they offer.

According to the graph, when you put 80ma through each diode (40% of their rated power), their individual voltage drops 2.8v. For four strips in series, that would be (2.8v x 16s = 44.8v) x 4 boards = 179.2v total, and you would use (0.080a x 9p) 0.72a to power them for a total wattage of 129.02w.

Have another look at the graph where the line intersects 160ma and roughly 2.94v. Do the maths and you get 47.04v for each board (188.16v total in series) @ 1.44a for 270.9504w.

So do you see how the more current you put through each board, the higher the voltage drop?

It's not linear - you would at first expect that double the current equals double the wattage, but in fact the wattage has gone up by more than 2x because the individual voltage has also increased.

For the sake of drivers (and hopefully I won't get my figures mixed up this time!), a HLG-240H-C1400a would drive the first example (it would be 0.2v over the 179v max, so would likely just handle it), but not the second example - because even though it can provide 1.4a, it can't provide it at 188.16v. You need to find another driver. Also, the total wattage (271w) is a giveaway, because the 240 series drivers are only rated to 240-250w, though I understand they will actually put out a little more.

Each strip would only drop 48.4v if you were running them at max 1.8a, but at a more conservative (recommended) 1.12a, they drop about 46v - that's where the 1.12a/46v figures come from in the Samsung spec sheet for each strip.

I chose a 240 series driver to power four strips for three reasons: cost-effective (2x 240s appear to be cheaper than 1x 480); a smaller driver is more efficient than a larger driver powering the same number of boards (from my understanding); and lastly, I didn't really want to drive my boards much higher than 70%, so the 240H-48 in parallel fits the bill perfectly. In fact, I'll probably only be driving my boards at about 60%, so I could have gone with the 240H-C1050A in series, but the 48 gives me the option of a bit more available power if I need it, even though I will dim it down.

My suggestion? Run 4x strips on a 240 series, then when you buy another four strips, buy another 240 series driver - cheaper than 8x strips/1x480 with a lot more flexibility. You could run two channels (different CCT strips) etc or whatever you want.

I hope that all makes sense. I'm pretty sure my numbers are right this time

Good calculations, I also like the 48v CV/CC drivers for it's safety and
because one can make better use of their performance.
Only two things I want to mention.
All HLG-drivers above HLG-120 should have the same efficience within 1%. 240 - 600 CC are 94-95%, and CV/CC are 93-94%.
And your voltage calculations are at 25°C, it's 1% less at 50° and 1,4% less at 65°C(from datasheet, vf vs. temps). But the voltage applies at the start untill the temp getting stabilized.
The lumen min./max. ratings are also temps dependend, max. is at 25°C, min. at 95°C. For example, at 25°C the vf is 24,2v and at 1,12A one strips use 27,1w. At 95°C the same strip needs only 21,9v at 1,12A and deliver only 24,5w. Efficiency goes up to 190lm/w(3500k) at 25° and down to 154lm/w at 95°C(this all within 7% messuring tolerance).

S6 flux and A1 voltage bins are used on the F-series strips(the only combination which results in values from the datasheet, samsung calculater), if the differences at 65°C were really so high(154-190lm/w @65°C and 23v), they would have to use different bins on the strips, which would also result in a much wider dispersion in the forward voltage. They could use different voltage bins but this would increase the effort involved in fitting the strips so that they all work on the same voltage at the end. We can therefore assume that only the best flux bin and the worst voltage bins are used. The difference in one flux bin is max. ±3% or with other words, max. 6% between the lowest and the brightest chip.

 

[1212ham]

Nope,
they run them at 1120mA and the typical temps after stabilization is 65°C on the messuring point. There is no way to regulate the temps without changing the current.

May I ask the source of your information? At what ambient air temperature and density do the F-strips stabilize at 65C? Is the test done with the strip mounted to something or suspended "naked" in the air? Temperature certainly can be regulated, I've done it on cobs by using a water block heat sink and varying the water temperature. If temperature can't be regulated, how can manufacturers provide graphs of voltage and flux etc. vs temperature?

 

[nfhiggs]

May I ask the source of your information? At what ambient air temperature and density do the F-strips stabilize at 65C? Is the test done with the strip mounted to something or suspended "naked" in the air? Temperature certainly can be regulated, I've done it on cobs by using a water block heat sink and varying the water temperature. If temperature can't be regulated, how can manufacturers provide graphs of voltage and flux etc. vs temperature?

You are absolutely on point - for TESTING purposes the manufacturer absolutely uses a sinking method to control the temps and set them to what they want to test them at, whether its 65C, 85C or 25C.. Not every strip is going to stabilize to the same temp under the same conditions at the same current.

 

[Prawn Connery]

Here's a question about the constant current drivers. If you are under the total voltage, can the driver go over the rated amperage?

For example, if you have a HLG-240H-C1050A and four F series strips (as in my case) that need around 184v total in series, can the driver (rated at up to 238v) be adjusted via the inbuilt pot to deliver more than 1050ma? Will it ever go over 1050ma if it is a 1050ma CC driver?

 

[Rider509]

I just scoured everything I could find in MeanWell's website and I don't think exceeding the rated current output is possible.

 

[Prawn Connery]

Actually, I think they do. I just went over to Led Gardener's YouTube channel and watched his Quantum Board torture test video again, and towards the end he's running a 304 board on a HLG-240H-C2100A at 119.5v and just over 2.19a for a total of 262w, which is about 12w - or almost 5% - over its total rating (119v @ 2.1a). So it looks like they pack a bit of juice. I'm not sure how many more milliamps it would have driven the board if it hadn't been on the 119v maximum voltage rating.

 

[Rider509]

I should have qualified my answer by saying not significantly outside of stated tolerances.
Current is spec'ed at ±5%.

 

[kony brado]

Here's a question about the constant current drivers. If you are under the total voltage, can the driver go over the rated amperage?

For example, if you have a HLG-240H-C1050A and four F series strips (as in my case) that need around 184v total in series, can the driver (rated at up to 238v) be adjusted via the inbuilt pot to deliver more than 1050ma? Will it ever go over 1050ma if it is a 1050ma CC driver?

hi sir connery
meanwell A type drivers (with build in pot) will go over rated current (we love them for this).
go to meanwell site ,to your selected driver,and see the "report" section. wach the current adj. range for the real max current on your chosen driver(the 240h-c2100 for exemple go to 2.444~a)it has nothing to do with voltage. as long you are in the correct voltage rang the driver will give same current,it has to do with A type meanwells.

 

[Rider509]

I just watch an LED Gardener vid (on multimeters) where he showed what you just mentioned, kony brado. He took a 185H-C1750A driver to over 2000mA.
I like that!

I just ordered two HLG-600H-48A drivers for testing. I've got drivers coming out my ears but the leftovers will find a home in my tattoo shop lighting.

 

[kony brado]

I just watch an LED Gardener vid (on multimeters) where he showed what you just mentioned, kony brado. He took a 185H-C1750A driver to over 2000mA.
I like that!

I just ordered two HLG-600H-48A drivers for testing. I've got drivers coming out my ears but the leftovers will find a home in my tattoo shop lighting.

as you can see in the report 600h-48a ,A type will go to 14.1amps instead of 12.5amp in the specs .B driver will go to 12.5amps. in the cc.cv A meanwells the voltage is also adjustable and in your case (600h-48a) it will go to 50.4 volts,so you have the potential of 50.4V*14.1A=710.64 watts!!!
if you run f strips at~ 46v you can get about 648 watts.
Hope that helps.

 

[nc208]

Here's a question about the constant current drivers. If you are under the total voltage, can the driver go over the rated amperage?

For example, if you have a HLG-240H-C1050A and four F series strips (as in my case) that need around 184v total in series, can the driver (rated at up to 238v) be adjusted via the inbuilt pot to deliver more than 1050ma? Will it ever go over 1050ma if it is a 1050ma CC driver?

Short answer is no. The unused voltages will just be unused.

 

[Prawn Connery]

Thanks for the replies.

Next question: I've got 4m (12') of insulated 14awg/300v line to go from my driver to my light frame for a remote driver set-up. I want to connect the four strips in parallel and can see a number of ways of doing this:

1. Run 14awg positive/negative wires from driver to one board, then run 18/20awg positive/negative wires from that board to the others, linking the same connector on each board, a bit like below below (except instead of T junctions, there will be one wire going into each connector and one wire coming out of the same connector to the next board to create a parallel string) :

2. Run separate 18wg p/n wires from each board and solder them directly to the driver p/n wires as below:

3. Run separate 18wg wires from each board to a wago style 5-pole connector, with the driver wire plugged into the fifth pole as below:

What would you guys suggest? I'm thinking of running 18 or 20awg strand wire to/from/between boards, as I want a little bit of flexibility in the wiring just in case I want to move things around later. The wagos would probably be the most flexible set-up, too, yes? Any pros/cons to connectors vs direct solder?