Recent LED Strip Comparisons?

I have been researching LED strips. It seems they come in two varieties: cheap and expensive. I read a post on another site a while back where someone bought a bunch of comparable strips at a variety of price points and reached the conclusion that they were all identical (produced at the same factory anyhow) so buy the cheapest. Unfortunately he did not compare 5050s to 3528s nor did he include any of the "higher end" products such as the FlexFire Ultra Industrial series which claim 730 lumens per foot. My questions:

• Has anybody done actual side-by-side comparisons of lights from various manufacturers (from Hitlights at $19 per reel to FlexFire at $335 per reel?)
• Are manufacturers claims of output and efficiency generally reliable?
• Is CRI info left off of strip light specs because it is horrible? Is bright white better than warm white?

I am trying to make some purchasing decisions and don't feel that I have enough verifiable information to make an informed choice. I many have to buy a few brands to compare but spending $350 on a spool of FlexFires just to experiment with seems a bit extreme!

Thanks!

Don't know anything about the rest, but RE: CRI, CRI has only started to become a focal point for LEDs relatively recently, so unless the LED is tested specifically for it then the data wouldn't have existed. But yeah (very) generally speaking unless the company was focusing on higher CRI then the CRI is probably less than or around say 70?

As for "bright white better than warm white", it depends entirely on what your purpose and preference is. Generally, "cooler" white usually has a higher lumen rating, but for a lot of situations many people prefer a warmer white at the expense of some brightness.

Actually I have been comparing light strips from various stores "mainly ebay stores" as well as the different "cool white, neutral white and warm white" strips. This is not being done in any "professional way" but it's just me comparing different strips to find one that has a nice warm white output around 3000K.

So far I can say that I have not been able to find any strips using the 5050 package LEDs that have good warm white output, they are either too cool or way to yellow. I have found one decent semi warm white strip using 3528's but when I slapped it on an aluminum panel it was a little off, looked a lot better when I added a short strip of the too yellow 5050 strip I had. I have more strips on the way to play around with....

As far as the price difference, while only 1-2 online stores I have dealt with are not ebay stores they where still stores that had products coming from china. I can't tell the difference from the cheap ones and the more expensive ones they look like the same quality to me?

I have had a small strip of the 3528 warm white I found funning 24-7 for almost a month now to test the longevity of these strips, so far it still is bright as ever and all the LEDs are still lit up. This was one of the cheaper strips from an ebay store..

I just bought some 3528's on Amazon from LEDwholesalers. I put them under a workbench/cabinet in my garage with a LED switch. It looks GREAT! No more digging around in the dark to find whatever I'm looking for. They were the warm white 3100k and give off a great tint - not yellow, very bright!

alpg88 said:

I used 5050 and 3528 strips in kitchens and closets, hallways..I used to glue them to wood, and latter to aluminum profiles. The result been the same, at 12v input they burn out within months, those on aluminum took longer, but all dimmed. And burned to the point where you could see the dots burned thru phosphorus layer, when they were off.

The only thing that helped is doubling amount of strip and lowering voltage to around 10. Even than I use them in places that get occasional use, like closets, and inside cabinets. They also work well when sawn onto a dog's collar\harness,and fed by 6aaa. still bright enough to see dog at night from a far, but not too bright.
After using around half a dozen of rolls in various applications, I came to conclusion, to use them only if it is no practical way to use anything else,and even than you need to consider their limitations.

Click to expand...

Curious what the runtime was on yours until degradation?

I did some 3528's I sourced locally from LED City taped directly to kitchen cabinet surfaces. They're looking like new after only tens of hours, although I generally only use them for minutes at a time. I was going to use up the rest of them in closets and utility spaces but never got around to it.

I also installed some 5050 RGB's in aluminum channels over the garage door as holiday lights - those have many tens of hours runtime without obvious degradation.

i ran them long, some were on 24\7 others half a day at least, first month or two there was no noticeable change but later i could tell by eye they got dimmer, and from there they kept loosing brightness. in 6 moth or so they only had 1\3 or 1\4 of initial output, not measured, but what my eye saw. on 5050 strips even copper traces got darker, especially positive. here is a pic. you can see dark traces and 3 dots burnes in, the other piece is same one that has not been running long.

when i said burned out i didn't mean burned to the point they did not work no more, they did but a much lower output. so if you use strip for not long, don't let them heat up much, than you might use them long time and be happy with them, but if you plan on running them continually , my advice try something else, they are not the best solution. unless you willing to change them, every several month. since they are not expensive, their not long lifespan might not be much of concern.

Interesting. I guess the thermal limits of the die and 5050 package are being exceeded. Seems that 3528's might do a little better since there's only one LED in each of those - although they may not have some of the thermal features of the 5050.

I was never as gung-ho on LED tape as some - seemed too cheap, never made by a name brand, and always sold by hobby- or second-tier resellers. What I've got is likely destined for low-frequency/low-duration uses ... under-cabinet lighting, utility closets for water heater/furnace, a rarely-used guest room closet, and some safes.

alpg88 said:

i ran them long, some were on 24\7 others half a day at least, first month or two there was no noticeable change but later i could tell by eye they got dimmer, and from there they kept loosing brightness. in 6 moth or so they only had 1\3 or 1\4 of initial output, not measured, but what my eye saw. on 5050 strips even copper traces got darker, especially positive. here is a pic. you can see dark traces and 3 dots burnes in, the other piece is same one that has not been running long.

Click to expand...

The most obvious thing on your photo is the flux puddles forming on the solder joints of the leds. Only way that happens is extreme heat, u need way more heatsinking

There are significantly more choices out there is the conclusion I have come too. The issue of cost is pretty straight forward, if it sounds cheap, it will be. You can buy strip now in HI CRI, single double triple row, 120,240,360 leds per metre and using 3528, 5050, 5630, 5710 and more. best thing with higher led count is closer pitch which means with diffusers on your extrusion you do away with the ugly led dot effect

WeLight said:

The most obvious thing on your photo is the flux puddles forming on the solder joints of the leds. Only way that happens is extreme heat, u need way more heatsinking

Click to expand...

1 please quote so my posts don't look like yours. 2, you need to look more carefully at photo, those are not flux puddles, but darkened traces. the whole reason i post it, is to show heatsinking wont help, the choke point is flawed design.

alpg88 said:

those are not flux puddles, but darkened traces. the whole reason i post it, is to show heatsinking wont help, the choke point is flawed design.

Click to expand...

Those look like the "water proof" kind covered with a silicon type product, the ones I have been messing around with are open to air flow. I wonder if the silicon is trapping the heat more than the non water proof kind?

Thanks for the input everyone. Based on these responses and additional research I have done I have reached the following conclusions:

• Basic characteristics of a strip light are determined by the diode used (5050, 3528 etc. - see table below) and the spacing
• There will be minor differences based on manufacturer (efficiency, color consistency, heat dissipation)
• Some of the above may be due to better brands requiring better bins of same diodes
• Generally most warm white diodes have a CRI of 75 while most cool white diodes have a CRI of about 80.
• CRIs are available as high as 93 with some diodes, but the highest I have found for strips labeled "high CRI" are 85

Basic Diode Properties (from manufacturer specs - China price shown)
Note: Per foot/reel numbers are for densest single-wide diode spacing as shown in LED/5M column

So, a glance at the table shows:

• 3528 are dim and inefficient but relatively inexpensive and tightly spaced
• 5050 produce the most light per $, but inefficient and broadly spaced
• 3014 produce the most light per foot, with very tight spacing and high cost
• 2833 most efficient at over 100 lumens/watt
• 5630 perhaps a good efficient upgrade from the 5050 in brightness and efficency

Does all that look about right?

stevemayman said:

Thanks for the input everyone. Based on these responses and additional research I have done I have reached the following conclusions:

• Basic characteristics of a strip light are determined by the diode used (5050, 3528 etc. - see table below) and the spacing
• There will be minor differences based on manufacturer (efficiency, color consistency, heat dissipation)
• Some of the above may be due to better brands requiring better bins of same diodes
• Generally most warm white diodes have a CRI of 75 while most cool white diodes have a CRI of about 80.
• CRIs are available as high as 93 with some diodes, but the highest I have found for strips labeled "high CRI" are 85

Basic Diode Properties (from manufacturer specs - China price shown)

Diode	Lumens	Lum/Watt	Max Lum/FT	$/5M	LED/5M	Lum-ft/$
3528	3.6	44	130	27	600	79
5050	12.6	55	240	34	300	116
3014	9.2	68	630	240	1020	43
5630	25.6	85	470	110	300	70
2833	9.3	102	340	70	600	80

So, a glance at the table shows:

• 3528 are dim and inefficient but relatively inexpensive and tightly spaced
• 5050 produce the most light per $, but inefficient and broadly spaced
• 3014 produce the most light per foot, with very tight spacing and high cost
• 2833 most efficient at over 100 lumens/watt
• 5630 perhaps a good efficient upgrade from the 5050 in brightness and efficency

Does all that look about right?

Click to expand...

I'm familiar with 3528 and 5050 packages since I've experimented with LED tape utilizing them. Tapes with 3528 and 5050 packages are everywhere, often available in double-density for a price premium over standard density (one local vendor sells triple-density 3528 on a 3x width tape). From what I've seen, 3528's are single-die and 5050's are triple-die. Both seem to be limited to around 20mA per die.

As far as I know, the packages are fairly generic - short of whatever optical and power limitations they possess, the manufacturer can insert chips and phosphors of whatever quality they can produce/source.

The other packages look to be available in different densities to accommodate different voltages and up to whatever the limit is of the traces on the PCB (be it flexible tape, FR4, or traces on a MCPCB).

Doing some quick research, I'm seeing 3020 and 5630 on some rigid bars (presumably MCPCB) and a handful of 2833's and 3014's on flex tape. If the package dissipates much power at all, mounting it to MCPCB strip should extend its life if it has provisions for effective heatsinking. I believe that similar to 5050 and 3528 packages, efficiency is mostly going to be a function of die/phosphor quality.

Your analysis looks very good for the samples you include. Unfortunately, the conclusions that can be drawn from it cannot be extrapolated to other strips using the same LED packages. It's possible to either put an efficient, inexpensive die in a 3528 package, and pack them tightly on the strip, or to put an inefficient, expensive die in the same package and space them widely on a strip. The results would be radically different, even though the package is the same. In other words, 'quality' cannot be determined by looking at the LED package type.

Lumen maintenance, the opposite of the 'burn-out' described by alpg88, is definitely affected by temperature. But keeping the LEDs cool does NOT ensure good lumen maintenance. It only ensures it will be better than if the LEDs are allowed to run hot. I have a nite-lite that has a few 5mm white leds in it. They certainly run cool, but after a year (8760 hours) running almost 24/7, it was notably less bright, which means probably 50% maintenance. Good LEDs like Cree or Philips have 70% maintenance after 50,000 to 100,000 hours. Now, after 6 years (around 50,000 hours), it barely puts out any light at all. You can see the light, but it doesn't illuminate anything more the a foot or two away. That's probably less than 5% of the original output, or 5% lumen maintenance at 50,000 hours, where a good LED would be 70% at 50,000 hours.

I have no doubt that what you have said below is true, but what can we use to determine the performance or LED strips or otherwise make purchasing decisions? I was hoping that someone would share their first-hand experience with actual products. In lieu of actual experience I was comparing the manufacturers specifications for various packages and they were remarkably similar for a given package.

For example, for the (7) 3528 packages I looked at the stated lumens/diode only varied from 3.1 to 3.9 and the efficiency only varied from 39 to 48 lumens/watt. Much of this variation was due to color temperature. This seams to suggest that 3528 packages perform very similarly to each other. Is it possible that Cree has produced a 10 lumen/diode 150 lumen/watt 3528 diode in the lab? Sure, but you don't see a lot of them when shopping. You see chips that claim to produce about 3.5 lumens per diode and this is largely independent of price.

I went in expecting to see significant variations in output, efficiency, and lifespan based on the quality of the die but I just didn't find them. In fact the goal of my original post was to suss out such differences if they exist, but I have been hard pressed to find any significant variation in specs other than by package. If anybody has any examples of products that significantly outperform the package averages I would be very interested as I am looking for a product that performs well and am willing to pay a reasonable premium for it.

I didn't mean to come off as defensive, and I do appreciate you comments but perhaps I am seeing very similar specs because I am looking in the wrong places? I simply Googled phrases like "3528 LED strip" and clicked through to as many web sites as seemed reputable. Are there specific vendors out there that sell high quality/ high performance strips from name-brand manufacturers? It seems to me that the world of strips is a world of similarly performing generics...

Thanks!

DIWdiver said:

Your analysis looks very good for the samples you include. Unfortunately, the conclusions that can be drawn from it cannot be extrapolated to other strips using the same LED packages. It's possible to either put an efficient, inexpensive die in a 3528 package, and pack them tightly on the strip, or to put an inefficient, expensive die in the same package and space them widely on a strip. The results would be radically different, even though the package is the same. In other words, 'quality' cannot be determined by looking at the LED package type.

Click to expand...

I wonder if maybe all the leds in a particular package come from the same manufacturer. Each mfr has their own package, and that explains why all samples of a particular package look similar? I wouldn't have thought that was the case, but maybe it is. That's why it's great to have people with different biases talking - my bias would never have lead me down the path to discover this.

That is my theory. My guess is that all the diodes in the strips <$50 come either from one or two similar generic immitators racing to the bottom or that they come from the same or similar "reject" (or at least less desirable) bins from a name manufacturer that won't put their name on it. The big boys have probably all moved on to more lucrative markets.

The only branded diodes I saw were several strips that claimed samsung LEDs in a high CRI 5630 package.

DIWdiver said:

I wonder if maybe all the leds in a particular package come from the same manufacturer. Each mfr has their own package, and that explains why all samples of a particular package look similar?

Click to expand...

stevemayman said:

That is my theory. My guess is that all the diodes in the strips <$50 come either from one or two similar generic immitators racing to the bottom or that they come from the same or similar "reject" (or at least less desirable) bins from a name manufacturer that won't put their name on it. The big boys have probably all moved on to more lucrative markets.

The only branded diodes I saw were several strips that claimed samsung LEDs in a high CRI 5630 package.

Click to expand...

There's also the possibility that the strip manufacturers are simply regurgitating the spec sheets from the LED supplier, which may simply be doing a back-of-the-napkin calculation based on spec sheets from their LED die and phosphor manufacturers. You'll notice that the only name brand on the flex strips seems to be 3M (who supplied the adhesive tape), and they're generally sold from hobbyist / small-scale supply houses, who make little if any mention of where the tape comes from ... ripe for a little inflation or unthinking copypasta of data from the upstream step in the chain.

I suspect the motivation for high quality is fundamentally low. The tape suppliers likely change their public face all the time by changing sales agents. The resellers likely deal with quality claims through replacement stock out of their no-doubt appreciable margins or byzantine RMA procedures that discourage returns. Customers probably use them in lower-demand locations and perhaps have a natural tendency to use enough LED's that significant lumen depreciation is not all that noticeable.

My only quibble with your table were with your "Max Lum/FT", "$/5M", "LED/5M", and "Lum-ft/$" calculations, as those are not specific to the diode type with 1x, 2x, 3x, and 4x densities available in many of those LED types (especially 3528's). If you're comparing whatever passes for "standard density", then my quibble pretty much vanishes.