Needing some help with designing my led light.

Randomblame

Well-Known Member
I've read a few of these threads by folks like rider509 and graying.geek regarding the led strips and personally I think that it's better for me and my application over the traditional cobs. I like the fact that the light is more easily spread out by default vs cobs, you don't need reflectors, you don't need traditional expensive heatsinks, they run cooler which is a huge plus and what heat is generated is spread out instead of hot spots. I also like the fact I can get the lights closer to the canopy since they run cooler which also increases efficiency for what we're using them for. Thanks stone_free for the introduction to these strips!! I've looked at both the Samsung F series and the Bridgelux EB series. I can see the Sammy F series are more efficient but If I'm looking at the spec charts right the lumens per watt between them and the EB series doesn't seem that much at not least to justify a double price tag. I've used a few calculators to play around with the volts/amps that are in the spec charts for these two lights and what I can see is that the Samsungs do offer the ability to produce more at a higher amp than the EB series can at there max amps...the sammy's can be driven a little harder. But comparing the different loads they can handle and imputing the volts/amps into the calculator I doesn't seem to be such a huge factor to pay double. Are the F series just that much better as I see lots of people really like them, am I missing something about them over the EB? I want a good light and don't mind paying for them but they just seem to not be worth double which seems to be the conclusions of a few others. With the EB series I seem to be able to get close to what I wanted, here's what I've been looking at...I'm just posting everything that I was considering for conversation. What I'm considering is 8 of the below EB series @3000K for my 4Wx2D space. I already have one HLG-185H-C1050A driver that never got used and it can drive four strips all the way to 1A. With another driver and the other four strips, according to the calcs, I should be able to get close to the 400 watts I wanted. Each strip running at 1040mA / 45.3V gives=47.112 watts. That's 377 watt potential which as pointed out to me with the proper growing techniques should be enough. If I go this route it looks pretty close to greying geeks which seems to be doing awesome with his build. I really liked the " LT-FB24B" dual row leds but dam they are so expensive and I figured I'd get better spread with more strips so once again the EB series seems to shine on price point. Does this sound on point or are there better options I missing?

1. https://www.digikey.com/product-detail/en/bridgelux/BXEB-L1120Z-30E4000-C-A3/976-1502-ND/6236288

2. https://www.digikey.com/product-detail/en/samsung-semiconductor-inc/SI-B8V521B20WW/1510-2225-ND/6624009?WT.srch=1&gclid=CjwKCAjwssvPBRBBEiwASFoVd7ow9fYnB1gKVyjaAupjFtFBVn_vaiyhhlbVFP0Mu9eJ1Jb0KUyiqRoCAYgQAvD_BwE

3. https://www.digikey.com/product-detail/en/samsung-semiconductor-inc/SI-B8VZ91B20WW/1510-2229-ND/6676693?WT.srch=1&gclid=CjwKCAjwssvPBRBBEiwASFoVdzMTK4JZ0ks5jWXU2sEN0LbKRlrziDlne2MeYVAfRNPNA2T-5ANYNxoCw0UQAvD_BwE

Here are the tech sheets I found on them.

http://www.bridgelux.com/sites/default/files/resource_media/DS130 EB Series Datasheet Rev A_0.pdf

http://www.samsung.com/global/business/business-images/led/file/product/products/201704/Data_Sheet_F_Series_G3_Rev.1.0.pdf

Here are the calculators I used for the basics.

https://www.google.com/search?q=convert+milliamps+to+amps&oq=convert+mill&aqs=chrome.1.69i57j0l5.5663j0j7&sourceid=chrome&ie=UTF-8

https://www.supercircuits.com/resources/tools/volts-watts-amps-converter

You need to compare the chips not the strips.
EB Series diodes runs at 87,5mA for 159lm/w,
but LM561c diodes with the same 87,5mA is 177lm/w and there are more diodes, 72 vs 48 on the 2footers.

If you take EB Series make sure to take the new gen.2 which is 175lm/w at 87,5mA.
56 diodes on 2ft., 112pcs on 4' strips.
 

suthrngrwr

Well-Known Member
You need to compare the chips not the strips.
EB Series diodes runs at 87,5mA for 159lm/w,
but LM561c diodes with the same 87,5mA is 177lm/w and there are more diodes, 72 vs 48 on the 2footers.

If you take EB Series make sure to take the new gen.2 which is 175lm/w at 87,5mA.
56 diodes on 2ft., 112pcs on 4' strips.
Seems like the Gen.2 strips aren't in stock anywhere. Has anyone been able to source the new Gen.2 strips yet?
 

taproot

Well-Known Member
constant voltage drivers in parallel might be better suited for you

do some reading and ask some questions
First, thanks so much for chiming in...I really appreciate your time. This might explain why I've seen some guys that run their setup lower say at 300 watt say I can turn this setup to 600 watts for ex...but then when I look at their driver It's forwarding voltage isn't enough for the amount of leds they have. This is negated in parallel from my reading but also from said reading it seems parallel isn't considered efficient, also isn't it said that certain leds in the parallel circuit will get more amps than others thus for ex. ones in the front might be brighter or putting out more lumens than the ones in the back? To be honest I'd rather have a single driver that was capable of driving up to about 600 watt if should I ever expand but also capable of running at 300-400 for now.

This is what I was reading on the wiring differences.

http://www.ledsupply.com/blog/wiring-leds-correctly-series-parallel-circuits-explained/

This was the sort of driver I was looking at ..it's max volts are 48V which isn't good for series wiring but it should work for parallel wiring; it has 12.5 amps and rated at 600W. It says it has dimming which is good..but my question is how do we know how low they can dim? What if I only find out my current setup can handle 350W how do I know the dimming on this driver can reduce the current enough to get that low? Also, on these drivers when it list one voltage such as this one at 48 does the led have to be the exact same..what if the led only need 35 or so can these drivers lower the volts down past what they are listed?


https://www.mouser.com/ProductDetail/Mean-Well/HLG-600H-48A/?qs=5pLaOnqdcwMTG4LuGsfZDA==&gclid=Cj0KCQjw1dDPBRC_ARIsAJZrQfpk_Q6mvTltE-cW6U4FfBmCZtS4DMy-7ucAVXyfdWxEM7SPEb5swdMaAq23EALw_wcB

I'm not sure how to interpret the Data specs on that driver. The specific model I made an example of above show it might have a vol adjustable range from 24-48 but then down a little lower is shows 40.8 ~ 50.4V. I also shows the current range 6.2 ~ 12.5A ...so I assume there's no way to get it down to what I was looking for which would have been 900mA to 1.5 or so

http://www.mouser.com/ds/2/260/HLG-600H-SPEC-1131635.pdf
 
Last edited:

suthrngrwr

Well-Known Member
http://www.ledsupply.com/blog/wiring-leds-correctly-series-parallel-circuits-explained/

This was the sort of driver I was looking at ..it's max volts are 48V which isn't good for series wiring but it should work for parallel wiring; it has 12.5 amps and rated at 600W. It says it has dimming which is good..but my question is how do we know how low they can dim? What if I only find out my current setup can handle 350W how do I know the dimming on this driver can reduce the current enough to get that low? Also, on these drivers when it list one voltage such as this one at 48 does the led have to be the exact same..what if the led only need 35 or so can these drivers lower the volts down past what they are listed?


https://www.mouser.com/ProductDetail/Mean-Well/HLG-600H-48A/?qs=5pLaOnqdcwMTG4LuGsfZDA==&gclid=Cj0KCQjw1dDPBRC_ARIsAJZrQfpk_Q6mvTltE-cW6U4FfBmCZtS4DMy-7ucAVXyfdWxEM7SPEb5swdMaAq23EALw_wcB

I'm not sure how to interpret the Data specs on that driver. The specific model I made an example of above show it might have a vol adjustable range from 24-48 but then down a little lower is shows 40.8 ~ 50.4V. I also shows the current range 6.2 ~ 12.5A ...so I assume there's no way to get it down to what I was looking for which would have been 900mA to 1.5 or so

http://www.mouser.com/ds/2/260/HLG-600H-SPEC-1131635.pdf
Depending on the size of your light, I would split the strips in half and power each half on their own power supply. That way you can achieve efficiency and proper dimming control.

For example, if you wanted a 600 watt light, source 2 300watt dimmable power supplies.

This would also achieve the requirement of proper current supplied to a series circuit. If for instance, you wished to power a series of strips at 1050ma, but wanted the option of dropping down to 700ma and up to 1400ma, go with a 1400ma constant current driver.
 
Last edited:

taproot

Well-Known Member
Depending on the size of your light, I would split the strips in half and power each half on their own power supply. That way you can achieve efficiency and proper dimming control.

For example, if you wanted a 600 watt light, source 2 300watt dimmable power supplies.

This would also achieve the requirement of proper current supplied to a series circuit. If for instance, you wished to power a series of strips at 1050ma, but wanted the option of dropping down to 700ma and up to 1400ma, go with a 1400ma constant current driver.
Right, which is what I was looking at doing ..see my previous post about doing exactly this. Parallel was mentioned and I see some guys using is thus the interest. I'd like to see the pros and cons on it etc.
 
Last edited:

taproot

Well-Known Member
The two biggest issues I see with parallel is that you can have current hogging some one led/strip might take more than it's fair share. The voltage is the same across the board but the total available current is divided by the amount of devices in the circuit; is this correct? And, after reading more about it that driver, and assuming the latter is correct, I mentioned above that can scale from 6A to 12A if I was say to scale the driver back to 6A and had 6x 1A strips and one strip decided to fail the other 5 leds would receive the extra current so now they are 1/5 over driven than before. Is this correct? Seems like that could be an issue under certain circumstances.
 

nfhiggs

Well-Known Member
The two biggest issues I see with parallel is that you can have current hogging some one led/strip might take more than it's fair share. The voltage is the same across the board but the total available current is divided by the amount of devices in the circuit; is this correct? And, after reading more about it that driver, and assuming the latter is correct, I mentioned above that can scale from 6A to 12A if I was say to scale the driver back to 6A and had 6x 1A strips and one strip decided to fail the other 5 leds would receive the extra current so now they are 1/5 over driven than before. Is this correct? Seems like that could be an issue under certain circumstances.
The whole "current hogging" thing is just WAY overstated. Its only a risk if you are running your power supply at the very limit of its current capacity.
 

nfhiggs

Well-Known Member
The two biggest issues I see with parallel is that you can have current hogging some one led/strip might take more than it's fair share. The voltage is the same across the board but the total available current is divided by the amount of devices in the circuit; is this correct? And, after reading more about it that driver, and assuming the latter is correct, I mentioned above that can scale from 6A to 12A if I was say to scale the driver back to 6A and had 6x 1A strips and one strip decided to fail the other 5 leds would receive the extra current so now they are 1/5 over driven than before. Is this correct? Seems like that could be an issue under certain circumstances.
And no, if a strip fails, the others don't receive more current, unless you are overloading the power supply to begin with.
 

Dave455

Well-Known Member
You need to compare the chips not the strips.
EB Series diodes runs at 87,5mA for 159lm/w,
but LM561c diodes with the same 87,5mA is 177lm/w and there are more diodes, 72 vs 48 on the 2footers.

If you take EB Series make sure to take the new gen.2 which is 175lm/w at 87,5mA.
56 diodes on 2ft., 112pcs on 4' strips.
How about the 2 foot Samsung double F strips... 144 diodes each..8700 lumens at 168 l/w ?
 

Randomblame

Well-Known Member
How about the 2 foot Samsung double F strips... 144 diodes each..8700 lumens at 168 l/w ?
Same diodes as on the single row strips. At 1120mA each diode gets 124mA for 168lm/w but if you run the strip lower at 720mA, they would get 80mA which correspondents to 180lm/w.
In my next DIY project the strips gets only 533mA(188lm/w) but 12 2ft. single row strips per HLG-150H-48A. 6p2s circuits would need ~45v, driver has 3,2A : 6=533mA per strip.
 

taproot

Well-Known Member
And no, if a strip fails, the others don't receive more current, unless you are overloading the power supply to begin with.
Are you sure about that? I think it's about capacity in a parallel circuit; if a participant on the circuit goes out the available capacity for the other remaining participants has increased. Whatever is on the circuit, in this case leds, would have to have a governor built into them or they would continue to consume until the excess has been absorbed. The limited research on a parallel circuit that I've done seems to indicate this.
 

Randomblame

Well-Known Member
@taproot
The first row with 24-48v means their constant current region, because it's a CV/CC driver.

Only if you set the voltage higher, the driver changes to the CV mode and regulates the current down until it remains within its specifications.

The 2nd 43-53v is the range for the voltage regulation with the build-in voltage poti. Those amperage values means the same, you can regulate the current within this min./max. values via current poti.

The structure of the circuit is dictated by the driver. That means, if you use a 600w CV driver, then you must also hang LED, which can handle 600w. And do not worry about thermal runaways, today's LEDs are subject to very small fluctuations. If you want to be quite sure, check the VF with a multimeter before you use the strips. Differences of a few millivolts does not matter. You can avoid thermal runaways also by adjusting the voltage poti to the exact the same vf as the strips need at max. driver current. If one strips fails the voltage limit prevents the current from drifting higher.
These datasets are not optimal, but if you study them thoroughly, you will find all the information you need.
 

Randomblame

Well-Known Member
The HLG-600H-48, A-Series has a voltage and a current regulater and is dimmable via current poti within 50 and 100%(~105% max.)
The B-Series of the same driver is dimmable between 0-100%, but there is no voltage regulator. This means, you can only use them in their CC range(24-48V), higher voltage up to 53v is not possible.
 
Last edited:

nfhiggs

Well-Known Member
Are you sure about that? I think it's about capacity in a parallel circuit; if a participant on the circuit goes out the available capacity for the other remaining participants has increased. Whatever is on the circuit, in this case leds, would have to have a governor built into them or they would continue to consume until the excess has been absorbed. The limited research on a parallel circuit that I've done seems to indicate this.
No, it absolutely does NOT work that way with a constant voltage driver. I'm a professional electronics tech, been working on this kind of stuff for over thirty years.

What you are talking about ONLY occurs when using a constant *current* driver with parallel loads, or using a constant voltage driver with too many loads in parallel
 

Randomblame

Well-Known Member
No, it absolutely does NOT work that way with a constant voltage driver. I'm a professional electronics tech, been working on this kind of stuff for over thirty years.

What you are talking about ONLY occurs when using a constant *current* driver with parallel loads, or using a constant voltage driver with too many loads in parallel
I think he means, when in a circuit with 6 parallel strips one strip would have a forwards voltage which is 1v lower than the others, it would run at much higher current than the other 5 strips.
This problem occurs when there is a significant difference between the vf of the used strips. Therefor I adviced him to check the fv of each strip with a cheap multimeter and 10w/300mA driver(12,50 + 3,50$) and only to use the strips with the same fv(within a few millivolts).
As long as he always uses the same driver to test the individual strips, he should find the faulty strips easily.
However, I do not think he will find any with more than 0.1v difference.(without heatsink, 5 min. to stabilizise)
 

taproot

Well-Known Member
I was talking about the load as in current. All the research I did seemed to suggest that the load was divided equally in a parallel circuit thus if one ,or more, participant failed the total capable circuit load would now be available to the remaining participants and once again divided equally. A 4A Parallel circuit with 4 leds on it would give each led 1A, take one out and you should now have the remaining 3 running at 1.33A. Or, you could add another to make 5 and all five should have 0.8A. That's the way the article I read on it seemed to suggest so the warning that you could potentially have a overload was pointed out. nfhiggs says he does this for a living so I'll have to assume I'm misunderstanding what I read and do more research when I can. I honestly appreciate both you guys taking time to comment and point these things out..thanks.
 

nfhiggs

Well-Known Member
I was talking about the load as in current. All the research I did seemed to suggest that the load was divided equally in a parallel circuit thus if one ,or more, participant failed the total capable circuit load would now be available to the remaining participants and once again divided equally. A 4A Parallel circuit with 4 leds on it would give each led 1A, take one out and you should now have the remaining 3 running at 1.33A. Or, you could add another to make 5 and all five should have 0.8A. That's the way the article I read on it seemed to suggest so the warning that you could potentially have a overload was pointed out. nfhiggs says he does this for a living so I'll have to assume I'm misunderstanding what I read and do more research when I can. I honestly appreciate both you guys taking time to comment and point these things out..thanks.
It simply does not work that way with Constant VOLTAGE supplies - as long as you are not operating at max current load of the supply. Where problems CAN occur is a situation where you add parallel loads until the max current is reached.

As an example - lets you have a 36V, 10 amp (360W) Constant Voltage supply and you put five 36 ohm resistors in parallel on it. Each resistor will have one amp of current going through it. The supply will be putting out 180W of power and 5 amps at 36V - leaving 5 amps of "headroom" - unused power. You can add or remove a resistor and nothing will change on the others - they will all have 36V and 1 amp through them - because the current through a load is determined by the voltage applied across the load - and the voltage has not changed.

Now if you add resistors until you have ten 1 amp loads, you are now in the supply's "Constant Current" operating region - any increase in the amount of load resistors results in a voltage decrease to maintain the 10 amp maximum current. Voltage will drop to about 32.7V and each resistor will have .909 Amps of current through it. Now at this point, if you remove a resistor, the current though the others WILL increase, back up to 1 amp. But if you remove another one, there is no change, because that moves you out of the power supply's constant current region. As a general rule Constant Voltage supplies are best utilized at 80% of max current load. Running them in their constant current region tends to cause them to overheat.

So, to sum up - Constant Voltage supplies are best used with parallel loads, and Constant Current supplies should be used with series loads.
 
Last edited:
Top