What led strips to consider in 2024?

taproot said:

What I'm confused about is there was a mention that the samsung strips could have a run away event leading to a fire. Voltage is electrical pressure, I get that. I also understand that the lower voltage 46 vs 48 like the other strips means its resistance is slightly less and as a result will consume slightly more current. All the strips, including the samsungs, are rated for 60W...it's just the voltage is different. I figured the samsungs would just consume a little more current and run slightly hotter. For example, if I ran the entire array at half it's capacity ..based on a killowatt I'd assume the 48V strips would be close to 30W and the 46V samsungs 32 or 33W. I know they won't be consuming current evenly but what I'm not understanding is where is the run away event that poses so much danger?

Regarding the pot...I was just saying I thought in a runaway like the light shorted out the pot would prevent the driver from giving full current.

Click to expand...

I dont know how much of a danger thermal runaway is in practice. Ill have a crack at explaining it but im abit shy since theres electricians on this thread.
From what i understood and im happily corrected if im wrong: the lower voltage/resistance strings is where you will see a bit more current. So these strips will heat up a bit more. With higher temps comes lower voltage. Which then means higher current. This means that there is a positive feedback loop, the more they heat up the more current they draw which means heating up more.

In practice the higher current in the low voltage strips would also mean the voltage goes up. So im not sure if theres a point where it will balance itself out.

I think this issue was more prominent with older leds where voltage would vary very much between one piece and another and voltage binning wasnt very tight.

If you go ahead with this id be very curious of your experience with it, please tag or dm me, but i cant really say much more about this issue. To me its as simple as not building in a security risk into my build, even though it seems to work fine id eventually be stressing out about it. Also with a build that isnt really by the book could convert a minor failure into a mayor one. Too unpredictable imo. Somebody talked about the solution of adding resistors, which works but youre adding more points of failure.

Id get two separate drivers and be done with it. But im also interested in what would happen. I have sometimes toyed with the idea in of a build which would have unbalanced string but that was by about 1% difference in voltage. Never came to fruition but it wasnt due to this.

Rocket Soul said:

With higher temps comes lower voltage. Which then means higher current. This means that there is a positive feedback loop, the more they heat up the more current they draw which means heating up more.

Click to expand...

I understand now what's being said. With that..I'd wager to say the other strips would be heating up to and thus lowering voltage. With starting voltages being so close I don't see the samsungs being so hotter than the others that they get stuck in that loop and run away. But, I can see this happening in much larger variances. Even if they were all the same voltage, same brand same make and model...what in theory keeps them from lowering voltage as they get hotter which feeds more current and hotter and more current etc? I'd assume the driver would be the brake at this point either in its overall capacity or its set limit with the pot?

What I might do is get a few strips and individually power one of each, one 48v and a 46v @ 30W per strip and measure the differential temp of each.

Then power one of each together in parallel at 60W max out from the driver and measure the differential temp. And, I can at this point use a multi-meter to measure the dc amps each led to see which one is hogging more of the overall 60W output from the driver.

Anyways, thanks for the input gents.

taproot said:

What I'm confused about is there was a mention that the samsung strips could have a run away event leading to a fire. Voltage is electrical pressure, I get that. I also understand that the lower voltage 46 vs 48 like the other strips means its resistance is slightly less and as a result will consume slightly more current. All the strips, including the samsungs, are rated for 60W...it's just the voltage is different. I figured the samsungs would just consume a little more current and run slightly hotter. For example, if I ran the entire array at half it's capacity ..based on a killowatt I'd assume the 48V strips would be close to 30W and the 46V samsungs 32 or 33W. I know they won't be consuming current evenly but what I'm not understanding is where is the run away event that poses so much danger?

Regarding the pot...I was just saying I thought in a runaway like the light shorted out the pot would prevent the driver from giving full current.

Click to expand...

"It all starts with the profile of silicon. The electrical resistance of silicon increases with temperature up to about 160 Celsius, then starts decreasing. When resistance decreases, it allows more current to enter through the overheated areas, in turn causing yet more compounding heat, eventually leading to total failure of the device and hence the term Thermal Runaway.

With adequate thermal management and the use of a constant current driver, an LED will not exceed its maximum operating temperature and will never approach Thermal Runaway."

Cheated this off of google.
But as long as you have the constant current to prevent it from hogging too much current in the event it would go over 160celcius,
there is no risk of thermal runaway.

cage said:

"It all starts with the profile of silicon. The electrical resistance of silicon increases with temperature up to about 160 Celsius, then starts decreasing. When resistance decreases, it allows more current to enter through the overheated areas, in turn causing yet more compounding heat, eventually leading to total failure of the device and hence the term Thermal Runaway.

With adequate thermal management and the use of a constant current driver, an LED will not exceed its maximum operating temperature and will never approach Thermal Runaway."

Cheated this off of google.
But as long as you have the constant current to prevent it from hogging too much current in the event it would go over 160celcius,
there is no risk of thermal runaway.

Click to expand...

Right, the HLG driver I listed is CC and CV which is why I thought I'd be ok in the scenario we've been conversing about. Due to the following:

A: Such a small gap between the two voltages

B: A driver that limits the amount of requested current.

Would I be correct in saying, with this driver being used, that they can get hot and request all the current they want but the driver isn't going to oblige?

I can always get a few more of the same strips..I'm just more interested in the theoretical principal now.

taproot said:

Right, the HLG driver I listed is CC and CV which is why I thought I'd be ok in the scenario we've been conversing about. Due to the following:

A: Such a small gap between the two voltages

B: A driver that limits the amount of requested current.

Would I be correct in saying, with this driver being used, that they can get hot and request all the current they want but the driver isn't going to oblige?

I can always get a few more of the same strips..I'm just more interested in the theoretical principal now.

Click to expand...

Yes, you would. The power supply won't give them more current that has been set.
This is why it's called constant current. Though adjustable, but constant.

In my understanding the driver dimming can control the total current coming out of the driver, however even in CC mode, it doesnt control the distribution of that current over different voltage strings. Once the current is out of the driver the it just follows laws of physics which means the different voltage/resistance strings will draw different current.

Rocket Soul said:

In my understanding the driver dimming can control the total current coming out of the driver, however even in CC mode, it doesnt control the distribution of that current over different voltage strings. Once the current is out of the driver the it just follows laws of physics which means the different voltage/resistance strings will draw different current.

Click to expand...

You'd be right too.
And the balance of the strips could shift due the different diode resistance, different level of heatsinking etc.
But it won't go in thermal runaway because the driver controls the amount of current that comes out.
And as in my post it says the thermal runaway only comes into effect after 160celcius, after which the resistance starts dropping in the diode and it would like to draw more current.
So in theory if I insulate the other strip thermally to overshoot the 160celcius, then it could go in thermal runaway and hog the current from the other circuits.

But running them soft enough/heatsinked enough to avoid 160celcius

edit: I mean the samsung lm301b chips is specced for up to 85celcius, absolute maximum temp 110c.
But you should also plan for the eventuality of one of the strips failing which depending on how they are internally wired could drop part of the strip
or the whole strip. So have some thermal buffer.
But the 160celcius and thermal runaways is pretty off the specs so you'd have broken leds long before too.

taproot said:

Then power one of each together in parallel at 60W max out from the driver and measure the differential temp. And, I can at this point use a multi-meter to measure the dc amps each led to see which one is hogging more of the overall 60W output from the driver.

Anyways, thanks for the input gents.

Click to expand...

The 46v strip will draw most of the current. That assumes the voltage numbers are accurate, and that's a huge assumption if they aren't from the manufacturer's data sheet. Study post #72, a 1 volt increase nearly doubles the current...

Seems you guys are stuck on the driver and its current output from some reason.
A runaway board/diode takes current from the other boards/diodes...not from the driver.

Example(again, pretty worst case, but thats how you easiy show a concept)...
-drivers max current = 2.5amp
-each boards max = 1.4amps
-in a perfect parallel system each board should get ~1.25amps.
-In a unbalance/runaway situation...one could be getting 1.4a and the other only 1.1a...then runaway keeps going...and you're at 1.6 and .9...an so on. Over one boards max current and in the danger zone. All the while the driver never went over 2.5a output total. The driver is not saving anyone in the situation. Heatsinking is your friend in the situation...not a CC driver.

Now if your board can handle the whole current of the driver if all else fails...then youre in the "it can't happen to me" zone if the worst case happens.

cage said:

You'd be right too.
And the balance of the strips could shift due the different diode resistance, different level of heatsinking etc.
But it won't go in thermal runaway because the driver controls the amount of current that comes out.
And as in my post it says the thermal runaway only comes into effect after 160celcius, after which the resistance starts dropping in the diode and it would like to draw more current.
So in theory if I insulate the other strip thermally to overshoot the 160celcius, then it could go in thermal runaway and hog the current from the other circuits.

But running them soft enough/heatsinked enough to avoid 160celcius

edit: I mean the samsung lm301b chips is specced for up to 85celcius, absolute maximum temp 110c.
But you should also plan for the eventuality of one of the strips failing which depending on how they are internally wired could drop part of the strip
or the whole strip. So have some thermal buffer.
But the 160celcius and thermal runaways is pretty off the specs so you'd have broken leds long before too.

Click to expand...

Frankindank said:

Seems you guys are stuck on the driver and its current output from some reason.
A runaway board/diode takes current from the other boards/diodes...not from the driver.

Example(again, pretty worst case, but thats how you easiy show a concept)...
-drivers max current = 2.5amp
-each boards max = 1.4amps
-in a perfect parallel system each board should get ~1.25amps.
-In a unbalance/runaway situation...one could be getting 1.4a and the other only 1.1a...then runaway keeps going...and you're at 1.6 and .9...an so on. Over one boards max current and in the danger zone. All the while the driver never went over 2.5a output total. The driver is not saving anyone in the situation. Heatsinking is your friend in the situation...not a CC driver.

Now if your board can handle the whole current of the driver if all else fails...then youre in the "it can't happen to me" zone if the worst case happens.

Click to expand...

Seems like that can happen even if all strips are the same and one has some issue. In my scenario I'll have 16 strips at 48V and 8-10 at 46V so there's a buffer there. Everything is so dam expensive these days..Aluminum bar used to be cheap.

I've never got a answer on this question in the past so I'll ask again here while we're on this topic. Has anybody ever removed the thermal tape from strips. I used thermal tape on my last light and I'd take it off and reuse the flat bar if I thought it wouldn't destroy the strips but that shit is like gorilla glue.

This here makes great sinks..just pricey now. Still cheaper than a professional sink.

taproot said:

Has anybody ever removed the thermal tape from strips. I

Click to expand...

I've removed strips that have been taped to aluminum bar, but haven't actually taken the tape off the strips themselves (the tape stayed on the strip, not the aluminum). Getting them hot made it easier.

Best place to get aluminum is from a local metal supplier, they usually deliver for around 50$ (they come in 16-24' lengths)

taproot said:

Seems like that can happen even if all strips are the same and one has some issue. In my scenario I'll have 16 strips at 48V and 8-10 at 46V so there's a buffer there.

Click to expand...

You can protect against that by using drivers with adjustable voltage limit. For Meanwells its drivers with A or AB type dimming. Set max voltage slightly higher than needed for max desired strip current and all are safe. Even with just one strip connected, it will be protected by the volt limit.

1212ham said:

You can protect against that by using drivers with adjustable voltage limit. For Meanwells its drivers with A or AB type dimming. Set max voltage slightly higher than needed for max desired strip current and all are safe. Even with just one strip connected, it will be protected by the volt limit.

Click to expand...

That's the exact driver I stated I was going to use above..the AB version that caps the voltage and current which is why I still don't see a thermal runaway from a slight variance in voltage. I'm still interested in seeing how much one pulls over the other just for curiosity. If that driver can protect a runaway when all the strips are matching but one having a issue how can it not protect when a slight differential is present?

In your post about the 1 degree doubles the current, is that evident in the drawing? I wonder how bridgelux came up with that as I don't see that for samsungs or am I missing it? From what another poster stated silicon doesn't exhibit those properties until it gets fairly hot..I think it was like 180F.

Thanks for taking the time to converse about this..it's very interesting.

In any event anybody got a resource for some cheap < $20 decent strips 2700k - 3500k running @48V 22 inches in length other than aliexpress?

taproot said:

That's the exact driver I stated I was going to use above..the AB version that caps the voltage and current which is why I still don't see a thermal runaway from a slight variance in voltage. I'm still interested in seeing how much one pulls over the other just for curiosity. If that driver can protect a runaway when all the strips are matching but one having a issue how can it not protect when a slight differential is present?

In your post about the 1 degree doubles the current, is that evident in the drawing? I wonder how bridgelux came up with that as I don't see that for samsungs or am I missing it? From what another poster stated silicon doesn't exhibit those properties until it gets fairly hot..I think it was like 180F.

Thanks for taking the time to converse about this..it's very interesting.

Click to expand...

I said one VOLT, haven't said anything about runaway

1212ham said:

I said one VOLT, haven't said anything about runaway

Click to expand...

"Study post #72, a 1 volt increase nearly doubles the current"

It sounded to me you we're describing a runaway, which is why I asked for an expanded explanation.

Anyways, thanks for your input!

From what I've seen, both Bridgelux Gen3 and Samsung strips are solid choices. As for color temp, 3500K is still popular for its balanced warmth.

When it comes to drivers, the difference between constant current and constant power can affect efficiency and performance, but it depends on your setup and preferences. Some swear by heatsinks for better longevity, but others manage fine without them.

From what I've seen, both Bridgelux Gen3 and Samsung strips are solid choices. As for color temp, 3500K is still popular for its balanced warmth.

When it comes to drivers, the difference between constant current and constant power can affect efficiency and performance, but it depends on your setup and preferences. Some swear by heatsinks for better longevity, but others manage fine without them.

I haven’t seen Gen 3 strips available other than from unauthorized brokers/distributors, which I’m not willing to gamble on.

560pc in stock at digikey


See a few authorized disti's here...