Author Topic: Explain to me the (de?)-evolution of LED lightbulb technology  (Read 7712 times)

0 Members and 1 Guest are viewing this topic.

Offline Rick Law

  • Super Contributor
  • ***
  • Posts: 3490
  • Country: us
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #25 on: March 23, 2021, 08:43:06 pm »
It is to be expected.  CFL light bulbs also went through this expensive-to-cheap phase.

Nearly 30 year ago, I got a 4 pack of GE CFL, 60W equivalent lumans at about $15 each.  Two of them are still working!   The home HomeDepot/Walmart CFL's bulbs these days dies left and right at a youthful age.  I've a couple of those ~3 year old CFL with so much brightness lost it is kept as emergency - in case I needed a bulb when stores are closed.  Most of the others hardly last past 3 years old.

I expect the HomeDepot/Walmart LED lights (60W equivalent lumans) to be dying of young age as well.
 

Offline rsjsouza

  • Super Contributor
  • ***
  • Posts: 6105
  • Country: us
  • Eternally curious
    • Vbe - vídeo blog eletrônico
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #26 on: March 23, 2021, 08:59:14 pm »
I wonder if this race to the bottom price-wise results in LED bulbs that generate lots of RF noise? I wouldn't be surprised at all if it did.

So now we have a product that generates two kinds of pollution: RF pollution and light pollution.

I recently bought some new LEDs to replace some early ones that were starting to fail.  Coincidentally, around the same time I noticed a lot of interference on a vintage analog TV in the next room that gets used to watch the news.  There were very noticeable wide horizontal noise bars moving up the picture.  It turned out that the new bulbs were causing the interference.  The set is fed using a shielded coaxial line, so I was surprised so much noise was leaking in.  With the coax unplugged, the usual snow was completely obscured by the interference.

Out of curiosity, I used my spectrum analyzer to look at the interference.  This was with just an dangling probe as an antenna.  You can see the FM band on the left side, and a couple of ATSC TV channels on the right.  The bulbs were generating interference over the entire spectrum up to about 500 MHz.  Needless to say, they went back to the store.  The old bulbs generated no visible interference.

Feit. I could have guessed that. Those fail the quickest for me.
That is my experience as well. Nasty things. Thank you andy2000 for sharing the interference spectrum profile.

I found out that CREEs are still well done. I did a video a while ago about a 2018-era 60W equivalent bulb and it had a real constant current source.

If you are interested, check the circuit at about 10:30 of the video below (in portuguese only, sorry)

Vbe - vídeo blog eletrônico http://videos.vbeletronico.com

Oh, the "whys" of the datasheets... The information is there not to be an axiomatic truth, but instead each speck of data must be slowly inhaled while carefully performing a deep search inside oneself to find the true metaphysical sense...
 
The following users thanked this post: TimNJ

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8526
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #27 on: March 24, 2021, 12:20:09 am »
Some more discussion on the Dubai Lamp here: https://www.eevblog.com/forum/chat/the-dubai-lamp/

Capacitive droppers are cheaper and in my experience more reliable than the switching ICs. I have some lights using them and they've lasted most of a decade.

BigClive also made this video not long ago, which is mostly about a related topic, but one thing caught my attention: lights with socketed electrolytic capacitors:



I don't think I've ever seen socketed capacitors even in equipment designed to be serviced, much less a cheap throwaway lightbulb.
 
The following users thanked this post: MK14

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #28 on: March 24, 2021, 01:04:09 am »
I had a lot of Feit CFLs that didn't last very long either, I actually fixed a bunch of them at the time, it was a mylar capacitor between the two cathodes used as part of the starting circuit that would short. For LEDs I buy mostly Cree and Philips, I've generally had good luck with those. I have some of the EcoSmart (Home Depot house brand) bulbs from way back in 2010 and a few modern ones, all of those are still working surprisingly.
 

Offline BrokenYugo

  • Super Contributor
  • ***
  • Posts: 1214
  • Country: us
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #29 on: March 24, 2021, 03:27:40 am »
They socket the filter cap and mains connection to speed up/cheapen assembly, those reflow compatible sockets must be cheaper than the cost and/or time of hand soldering.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7678
  • Country: ca
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #30 on: March 24, 2021, 05:11:49 am »
LED light bulbs are shit because of intellectual property. The LED array, the heatsinking configuration etc. are all so heavily patented you can no longer make a decent LED bulb without patent infringement litigation. So many lawsuits, with the norm being to sue the retailers as well.
EPISTAR Files Second U.S. Patent Infringement Lawsuit Against Lowe’s 06/2020 over GE bulbs. Epistar (has over 4,000 patents) sued Lowe's Home Centers over the Kichler Lighting LED bulbs.
Lynk Labs suing Home Depot for patent infringement.
LED filament bulbs are really bad for litigation. Nichia Files Patent Infringement Lawsuit Against Feit Electric In California 01/2020 filament bulbs.
Seoul Semi, Cree, Epistar, Nichia, Feit, GE, Philips etc. they all are constantly suing each other and the retailers over their LED bulb IP.

It's sad that LED bulbs are still struggling to outlast incandescent bulbs.
Cree LED bulbs were terrible, not that Cree has ever been a profitable company but their bulbs barely lasted a year and LEDs failing, just pushed too hard.
Next I tried Philips bulbs- which are total shit, the PCB and electrolytic capacitor inside badly cooked due to high heat, open-circuit LEDs. But the LED's are the best I've seen for CRI. The change to COB also makes the spot lights narrow. I can't get them to last more than about 3,600hrs. You would puke if I showed pics of the failed boards.

Next, I tried Ikea bulbs and they are working great, many at the 7,000hrs point and still working fine. But only warm white is offered.
 
The following users thanked this post: MK14

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #31 on: March 25, 2021, 12:20:27 am »
That doesn't match my experience at all. I still have Cree and Philips LED bulbs I bought more than 10 years ago that still work, even some that are very heavily used. If there are patent issues that doesn't seem to be causing any problems, loads of different companies are making LED filament bulbs in all different styles. Most of the market for general illumination bulbs seems to have settled on the aluminum core PCB with surface mount LEDs on it firing into a molded plastic globe.
 

Offline MK14

  • Super Contributor
  • ***
  • Posts: 4987
  • Country: gb
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #32 on: March 25, 2021, 12:46:05 am »
Warning: Very long article, but fairly interesting explanation of the economics and stuff around Led lights.

https://www.manufacturer.lighting/info/248/
 

Offline MK14

  • Super Contributor
  • ***
  • Posts: 4987
  • Country: gb
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #33 on: March 25, 2021, 12:51:30 am »
If there are patent issues that doesn't seem to be causing any problems

https://www.forbes.com/sites/laurendebter/2019/07/31/walmart-amazon-target-ikea-bed-bath-beyond-sued-vintage-light-bulbs/

Quote
Walmart, Amazon, Target, Ikea and Bed Bath & Beyond were sued by the University of California on Tuesday for patent infringement relating to "filament" LED light bulbs.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7678
  • Country: ca
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #34 on: March 25, 2021, 01:15:40 am »
Philips LED bulbs 800lm 2700K 8.5W I put a dozen into service indoors and outdoors for night lighting, so ~8-10hrs/day in various fixtures and outdoors -35°C to +30°C, indoors +22°C.
They lasted about 1-1/2 year. On some the LED's have the "black dot of death" as BigClive calls it, the string went open circuit. But really, the power supply is a joke, it's baked phenolic with roast electrolytic. I had greater expectations for USD $4 each, so I'm calling them junk after all are dead now.
I bought a massive PAR38 Philips flood light bulb and laughed it back to the store, all it has is a small COB spot like their MR16 bulbs, it's all empty space and glass lol.

(did break the bridge rectifier taking it apart)
edit: compare the thermal conductivity of many plastics and they are basically an insulator. Patents prevent plastic heatsink fins from being used.
« Last Edit: March 25, 2021, 01:17:59 am by floobydust »
 
The following users thanked this post: MK14

Offline BrokenYugo

  • Super Contributor
  • ***
  • Posts: 1214
  • Country: us
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #35 on: March 25, 2021, 02:11:02 am »
That's incredibly roasted, I thought those driver chips all had thermal throttling. Were the bulbs in enclosed fixtures? If so were they rated for it? I know that's a gotcha that a lot of people overlook with retrofit LED bulbs, though at $4 I'd expect them to be good for it.
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8526
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #36 on: March 25, 2021, 03:28:22 am »
8.5W for 800lm, that's less than 100lm/W... but from the look of those pictures it seems a lot of the power dissipation was in the driver circuit itself.
That's incredibly roasted, I thought those driver chips all had thermal throttling.
They do, they just throttle at a very high temperature.
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #37 on: March 25, 2021, 04:47:34 am »
Hm those are newer than the Philips bulbs I have. Mine are mostly the older remote phosphor types, I have some of the ones that look like a regular A19 lamp, some of the 800 lumen tri-lobe style and some of the mushroom shaped lamps with the oddly shaped white glass globe over the phosphor dome. I also have a handful of Hue lamps that look almost identical to those.
 

Offline floobydust

  • Super Contributor
  • ***
  • Posts: 7678
  • Country: ca
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #38 on: March 25, 2021, 05:00:22 am »
The LED's are largely cooled by conduction, with the M-core board and glue to the aluminum tube.
But the driver board only gets cooled by convection inside the small chamber, which is pretty much useless obviously. Even if it (driver) dissipated 1W the temperature must be pushing well over 100°C.
IC is BP9916C with CL set to 235mA (2.55R) and ILED 118mA, IC rated max is ILED 200mA.
"The thermal regulation temperature is set to 140°C internally."  :scared:

The light fixtures are a variety, some large glass globes and metal top, like any typical fixture.
 

Offline paulca

  • Super Contributor
  • ***
  • Posts: 4361
  • Country: gb
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #39 on: March 25, 2021, 09:33:35 am »
If you think cramming all that circuitry into the base of a lamp bulb is squeezing things...

Remember a lot of LED bulbs now contain a bi-directional radio and MCU.  Radio either RF433, Zigbee/Zwave or 2,4Ghz Wifi!

The last B27 (UK Bayonet) Zigbee smart bulbs all had no visible circuitry in the bulb, just in the fitting.  I dismantled a broken one (yes I had 3 out of 4 dead within 24 hours of use).  Inside it looked like 240V LED string filaments with just a variable current limiter and a Zigbee radio board mounted 90* on to it.  All within a 2/3" x 2/3".  (Do not buy Innr bulbs!)

The Philips Hue B27 bulbs and the E14 mini-candle bulbs (working fine) show no bulk beyond the tiny little cap, so all over above must fit into the E14 screw housing..  How in hell!
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Online Siwastaja

  • Super Contributor
  • ***
  • Posts: 9328
  • Country: fi
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #40 on: March 25, 2021, 10:45:50 am »
But the driver board only gets cooled by convection inside the small chamber, which is pretty much useless obviously. Even if it (driver) dissipated 1W the temperature must be pushing well over 100°C.

Quite frankly, the only Right Thing To Do is just to make the driver part efficient and not dissipate such power.

Modern white LEDs are some 40-50% efficient, we get that, you can't do much to improve it on the bulb level. You can go a bit higher by buying the most expensive state-of-the-art LED tech; or very crappy color rendition.

But there is no real reason why the driver circuit has to be some 70-80% efficient only. IMHO, seeing a charred hotspot on the driver board is inexcusable. 1W driver dissipation in a 5W bulb is a disaster, IMO, because there is no reason to be that crappy. Make it 90% efficient to see some real gain in power savings, and additionally the cooling problem is solved.
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 8274
  • Country: de
  • A qualified hobbyist ;)
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #41 on: March 25, 2021, 11:29:41 am »
I think the big problem is that we are trying to replace incandescent bulbs with matching LED bulbs to keep the old lamp fixtures, The old bulb design is great for incandescent light, but not for LEDs. We should design new standard LED lights and fixtures to get the most out of LEDs. For example, I have several meters of 12V LED strips on aluminum U profiles. They stay cool and run great for many years.
 
The following users thanked this post: NiHaoMike, rsjsouza, Cubdriver, Jacon

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8526
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #42 on: March 25, 2021, 01:22:13 pm »
IC is BP9916C with CL set to 235mA (2.55R) and ILED 118mA, IC rated max is ILED 200mA.
"The thermal regulation temperature is set to 140°C internally."  :scared:
That's a switching buck converter, not a linear regulator... so it should be more efficient? :-\
 

Offline TimNJTopic starter

  • Super Contributor
  • ***
  • Posts: 1720
  • Country: us
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #43 on: March 25, 2021, 02:37:34 pm »
Philips LED bulbs 800lm 2700K 8.5W I put a dozen into service indoors and outdoors for night lighting, so ~8-10hrs/day in various fixtures and outdoors -35°C to +30°C, indoors +22°C.
They lasted about 1-1/2 year. On some the LED's have the "black dot of death" as BigClive calls it, the string went open circuit. But really, the power supply is a joke, it's baked phenolic with roast electrolytic. I had greater expectations for USD $4 each, so I'm calling them junk after all are dead now.
I bought a massive PAR38 Philips flood light bulb and laughed it back to the store, all it has is a small COB spot like their MR16 bulbs, it's all empty space and glass lol.

(did break the bridge rectifier taking it apart)
edit: compare the thermal conductivity of many plastics and they are basically an insulator. Patents prevent plastic heatsink fins from being used.

Lmao come on, paper/phenolic board running at ~140C.  :-DD
 

Offline james_s

  • Super Contributor
  • ***
  • Posts: 21611
  • Country: us
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #44 on: March 25, 2021, 08:01:23 pm »
I think the big problem is that we are trying to replace incandescent bulbs with matching LED bulbs to keep the old lamp fixtures, The old bulb design is great for incandescent light, but not for LEDs. We should design new standard LED lights and fixtures to get the most out of LEDs. For example, I have several meters of 12V LED strips on aluminum U profiles. They stay cool and run great for many years.


They are doing exactly that, but there are billions of existing lamps and light fixtures out there, surely you can't expect everyone to replace all of their light fixtures? There is a lot of vintage/historical/period correct fixtures out there too. LED lighting would take decades longer to catch on if not for readily available retrofit bulbs.
 

Offline MK14

  • Super Contributor
  • ***
  • Posts: 4987
  • Country: gb
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #45 on: March 25, 2021, 08:12:40 pm »
I suppose if we redesigned lighting circuits from scratch today. We might make them 36V DC, then zero flicker (apart from any voltage regulator and current regulator effects), and minimal electric shock risk, only one household mains to 36V DV converter box needed (standard plug in (into a proper unit, not a mains socket), commodity unit, user replaceable), for all lighting circuits. The LED bulbs would then just be simple (low cost) LED chains (internally), and possibly a resistor or current regulator, inside the bulbs.
Given the low wattage needed by modern LED bulbs, the power and hence currents should be low enough, that the 36V cables wouldn't need to be thick high current ones.

LED bulb life expectancy should improve, because they would have less heat generated inside  the bulb, and no need for much extra electronics, on top of the LEDs, except maybe a resistor and/or current regulator.

Summary:
A bit like the florescent tubes, holder. With its internal electronics (ballast), and user replaceable starter. You only need to change the raw tube, not the rest of the electronics.
« Last Edit: March 25, 2021, 08:18:12 pm by MK14 »
 

Offline madires

  • Super Contributor
  • ***
  • Posts: 8274
  • Country: de
  • A qualified hobbyist ;)
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #46 on: March 25, 2021, 08:24:08 pm »
They are doing exactly that, but there are billions of existing lamps and light fixtures out there, surely you can't expect everyone to replace all of their light fixtures? There is a lot of vintage/historical/period correct fixtures out there too. LED lighting would take decades longer to catch on if not for readily available retrofit bulbs.

Then we have to accept that we'll create more e-junk by using LED bulbs which fail early. Is this better or worse than using incandescent bulbs with a higher power consumption in old lamps?
 

Offline Kleinstein

  • Super Contributor
  • ***
  • Posts: 15149
  • Country: de
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #47 on: March 25, 2021, 08:46:22 pm »
The early LED lamps with still relatively low efficiency had sometime a cooling problem when used at same intensity.  It's usually the higher power versions that fail early. Some fixtures just get rather hot and also the CFL lamps could failed early in those.  With modern, more efficient LEDs the heat is usually not such a problem any more as the power is soemthing like half of what it was in the early days.
If really in doubt, reduce the intensity and add a second lamp if needed.

One sometimes see new lamps with fix installed LEDs - to make use of the smaller form factor. So it would make sense to get some new smaller standard fixture - still keeping the old E27 / E14 for old lamps.
 

Offline TimNJTopic starter

  • Super Contributor
  • ***
  • Posts: 1720
  • Country: us
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #48 on: March 25, 2021, 08:55:49 pm »
They are doing exactly that, but there are billions of existing lamps and light fixtures out there, surely you can't expect everyone to replace all of their light fixtures? There is a lot of vintage/historical/period correct fixtures out there too. LED lighting would take decades longer to catch on if not for readily available retrofit bulbs.

Then we have to accept that we'll create more e-junk by using LED bulbs which fail early. Is this better or worse than using incandescent bulbs with a higher power consumption in old lamps?

I'm sure this has been studied and thought about many times, probably even on this forum alone. Still, here's my own thought experiment

Supposing an 800lm LED lightbulb lasts 6 years running at 5 hours a day. Over six years the energy usage is about 55KWh. Let's say the life span is about 10,000 hours, after which you have to throw it out and replace it.

If you used a 60W incandescent lamp for the same duration and hours per day, the energy usage is about 660KWh. Let's generously say it's lifespan is about 2,000 hours, so you'd need to replace it 5 times over this time.

The missing pieces are: The energy input required to make each type of bulb and the environmental impact of each bulb when it's chucked in a landfill. I don't think you can put the latter in terms of energy...although maybe you can think of it as the energy required by some poor future human society to un-****-up the world the previous generations left them.

--------

Total Energy (LED) = 55KWh + (Energy to Make/Dispose LED * 1)
Total Energy (Incandescent) = 660KWh + (Energy to Make/Dispose Incandescent * 5)

--------

Energy to Make/Dispose LED < 605KHh + (Energy to Make/Dispose Incandescent * 5)


So, naively, it makes environmental sense to use an LED as long as the energy it takes to make one LED bulb is less than 605KWh + the energy it takes to make and dispose 5 incandescent bulbs. And I have no idea how to gauge how much energy is required to make an LED bulb. But, it seems like LED bulbs should make environmental sense.
 

Offline TimNJTopic starter

  • Super Contributor
  • ***
  • Posts: 1720
  • Country: us
Re: Explain to me the (de?)-evolution of LED lightbulb technology
« Reply #49 on: March 25, 2021, 09:02:20 pm »
Is the fact that you can get an LED lightbulb for $5USD a reflection of the amount of energy it takes to make it? At that price, is it reasonable to think that it really doesn't take that much energy to make one?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf