Where can I get 200 lumen/watt LEDs?

[106Lake]

I recently learned about the Philips 4w 800 lumen lightbulbs available now and bought some for my house, absolutely amazing how bright they are for 4w!! I want to build a few things that use a similar amount of LEDs and are powered from an 18v lithium battery.

Are there LED filaments available for me to buy that will match the same 200 lumens/watt? Or some other type of LED emitters or COBs? I know similar LED filaments are available, but they are more like 100 lumens/watt from random chinese sellers. I've also seen some TCP LED retrofit metal halide bulbs that use COBs and claim 200 lumens/watt.

In theory if Philips is telling the truth their actual LEDs must be more efficient than that because they have to make up for losses in the power supply.

Would it be easier to just smash open a Philips bulb? There are 8 of them that appear to be wired in parallel. I have no idea what voltage/current they are driven at...

 

[jtr1962]

Some of the Cree J-series 2835 midpower LEDs exceed 200 lm/W

The Philips bulb actually uses a capacitive current limit. Read here. That sounds similar to what I did in this project. Efficiency of the driver circuit is actually quite high, around 96%, so the LEDs only have to be a few percent more efficient than 200 lm/W to give a bulb efficiency of 200 lm/W.

 

[LEDphile]

Efficacy goes up as the current density in the LED chip goes down. So if you want high efficacy, you need both low drive current and a large chip area (either in a single die, or split over multiple dice). This isn't hard to accomplish in a device that doesn't need high lumen density, the only downside is the cost of the LEDs.

 

[106Lake]

Efficacy goes up as the current density in the LED chip goes down. So if you want high efficacy, you need both low drive current and a large chip area (either in a single die, or split over multiple dice). This isn't hard to accomplish in a device that doesn't need high lumen density, the only downside is the cost of the LEDs.

I realize this is how LEDs work, but no manufacturer I am aware of rates their LEDs at lower current levels. I really don't understand why they don't rate lumens and lifetime on a curve rather than a solid number...

With underdriven LEDs I cannot measure lumens, and even if I ran some 100 lm/w filaments at 25%, I could not confidently say what they are actually putting out in terms of efficiency. I wish I could have a lumen meter but of course it isn't as simple as just pointing a meter at it.

I understand I'm nit picking but I really love that 200lm/w number and want to know that is what I am getting. Does anybody actually sell LEDs that claim to be capable of this? If I could get something like 100 lumens at 0.5 watts I would absolutely love it...and I really want to use these with DC power for various reasons.

 

[divine]

I thought there were several Nichia chips that were at or above 200 lumens per watt. Usually the lumen per watt number is done at a sweet spot on the output range. I think Cree chips used to use 350 mA for their lumen per watt number.

I wonder if you'll only be getting 200 lumens per watt on low mode or maybe medium mode.

It's possible that the efficiency would change from a constant current driver to a PWM driver, too.

It would be nice to have a 200 lumen light that only used 1 watt of power or even better, 400 lumens with 2 watts.


I wonder if we could gather that information from the charts from one of the flashlight reviewers.

 

[LEDphile]

I realize this is how LEDs work, but no manufacturer I am aware of rates their LEDs at lower current levels. I really don't understand why they don't rate lumens and lifetime on a curve rather than a solid number...

With underdriven LEDs I cannot measure lumens, and even if I ran some 100 lm/w filaments at 25%, I could not confidently say what they are actually putting out in terms of efficiency. I wish I could have a lumen meter but of course it isn't as simple as just pointing a meter at it.

I understand I'm nit picking but I really love that 200lm/w number and want to know that is what I am getting. Does anybody actually sell LEDs that claim to be capable of this? If I could get something like 100 lumens at 0.5 watts I would absolutely love it...and I really want to use these with DC power for various reasons.

Every LED datasheet I've looked at has the LED nominal values given for a single point (the "binning conditions", usually chosen to be an expected common set of drive conditions), and performance curves for the LED that give output and Vf over current and temperature. From these curves, and the values for the binning conditions, you can determine the expected performance (output, power, and efficacy) for pretty much whatever driving conditions you'd like.

Lifetime calculation is a bit trickier, as generating the data used for the calculations is quite time consuming, but generally is done based on operating temperature, combined with picking the next highest drive current from the available dataset.

 

[106Lake]

I wonder if you'll only be getting 200 lumens per watt on low mode or maybe medium mode.

It's possible that the efficiency would change from a constant current driver to a PWM driver, too.

It would be nice to have a 200 lumen light that only used 1 watt of power or even better, 400 lumens with 2 watts.

As far as I am aware Philips only makes these bulbs in 4 flavors - cool white, warm white, 60w replacement, 100w replacement. No dimming or multiple modes.

I don't understand why if they have access to this tech, and they can sell the bulbs for about $7.50 each, why they can't make every other bulb this way. For example, the 60w bulbs are too bright for some places, but if I bought a lower lumen bulb, it would still use the same 4 watts. That drives me insane!! These would be amazing lights in smaller sizes, bigger sizes, 4 foot tubes, DC voltages, and LED tape!!

I'll probably break open one of the bulbs and see what I can do to light just one filament. I think they are relatively high voltage, making this a bit harder to do from a battery. I could just use an 18v > 120v power inverter, but there goes my efficiency!

 

[LEDphile]

As far as I am aware Philips only makes these bulbs in 4 flavors - cool white, warm white, 60w replacement, 100w replacement. No dimming or multiple modes.

I don't understand why if they have access to this tech, and they can sell the bulbs for about $7.50 each, why they can't make every other bulb this way. For example, the 60w bulbs are too bright for some places, but if I bought a lower lumen bulb, it would still use the same 4 watts. That drives me insane!! These would be amazing lights in smaller sizes, bigger sizes, 4 foot tubes, DC voltages, and LED tape!!

The drawback to the higher efficacy is cost - the standard LED A19 bulbs sell for about $1.75, so you're paying 4x as much up front for 2x the efficacy. And when you're talking low-wattage bulbs in the first place, going from a 15W bulb to an 8W bulb (100W equivalent) only saves you a fifth of a cent per hour of operation (at $0.30/kWh, which is on the high side if what is typical - much of the US is half that). So you're looking at roughly 3000 hours of operation to pay for that difference at $0.30/kWh, or double that at the more typical $0.15/kWh. Payback times are even longer for the lower output bulbs.

 

[106Lake]

going from a 15W bulb to an 8W bulb (100W equivalent) only saves you a fifth of a cent per hour of operation (at $0.30/kWh, which is on the high side if what is typical - much of the US is half that). So you're looking at roughly 3000 hours of operation to pay for that difference

Not everything has to pay for itself. I am mostly interested in a 200lm/w LED for battery use, which of course would double the life over a 100lm/w LED. I don't understand why more people here are not interested in such a thing? Sure it "cost more" but we're only talking a couple bucks in LEDs.

A 2w lantern could provide me 400 lumens for 75+ hours with one of my 18v batteries, for example. That is amazing.

Also regarding Philips products, you'd think there would be a big market for this in 4' tube replacements. Halving the cost of lighting for a commercial environment while having really long lasting bulbs could be worth thousands of dollars per month in savings...but nobody sells that.

One of the big advantages of the 4w bulb is it stays basically room temperature - so all the components will surely last a very long time. Good for places bulbs are hard to reach and commercial environments. The reason I replaced my cheap old 10w LEDs in the first place is because they cooked themselves to death, and I burned my hand on one when I was replacing it. Some of the LEDs baked so hard they fell off the board and the bulb would slowly dim as more die.

 

[Dave_H]

I recently learned about the Philips 4w 800 lumen lightbulbs available now and bought some for my house, absolutely amazing how bright they are for 4w!! I want to build a few things that use a similar amount of LEDs and are powered from an 18v lithium battery.

Are there LED filaments available for me to buy that will match the same 200 lumens/watt? Or some other type of LED emitters or COBs? I know similar LED filaments are available, but they are more like 100 lumens/watt from random chinese sellers. I've also seen some TCP LED retrofit metal halide bulbs that use COBs and claim 200 lumens/watt.

In theory if Philips is telling the truth their actual LEDs must be more efficient than that because they have to make up for losses in the power supply.

Would it be easier to just smash open a Philips bulb? There are 8 of them that appear to be wired in parallel. I have no idea what voltage/current they are driven at...

An earlier discussion on CPF identified Samsung SMT 3030 LED LM301B series reaching about 230 lumens/watt in the 0.2W-0.3W range (65mA test). However this appears a high limit as reaching this depends on a number of factors such as drive current and CRI (70 appears highest but is it suitable for particular use?).

Brightness "binning" has an impact also. Some small white LEDs have binning range up to 2x or more. Samsung in question is around 25%. Reaching the limit means highest bin, which may be cost premium and/or supply issue.

As noted, overall efficacy of the bulb depends not only on LEDs but also power conversion. Anything much less than 90% probably wouldn't fly these days. I notice some low-end bulbs using direct linear regulation, stacking LED vf to approach dc value of the rectified ac line (115vac nominal). For example 16 triple-die LEDs would get to around 144v. This avoids switching circuit complexity and noise.

I've broken glass bulbs on only a few bulbs so far. On the small filament bulb, internals were undamaged, bulb operated without it and was able to take some measurements (with great care, potentially lethal voltages). Across one filament measured around 70-75v dc so two in series can be used like this for 115vac, with parallel combination for more filaments.

On some low-power bulbs with plastic bulbs I've been able to pop the diffuser off with minimal/no damage although it is a bit of technique. For this case (glass), there must be a technique for clean break-off at the base, otherwise risk damage/injury. I recall from chemistry lab, technique for clean breaking glass tubing: small file notch, pushing out with thumbs etc, but not sure about a bulb shape.


Dave

 

[106Lake]

I've broken glass bulbs on only a few bulbs so far. On the small filament bulb, internals were undamaged, bulb operated without it and was able to take some measurements (with great care, potentially lethal voltages). Across one filament measured around 70-75v dc so two in series can be used like this for 115vac, with parallel combination for more filaments.

There is a good video on youtube from BigClive taking one of these apart. I think he measured around 100v per filament. Interestingly it does appear the European version uses 2 groups of 4 in series for 230v, and the North American version only uses one filament in series, 8 parallel.

I think I could pop the bulb in a vise without breaking the filaments...just not sure how to drive them. I see a few high voltage DC-DC converters available, but they all have low efficiency.

I would LOVE to have some LED tape with those Samsung 3030s...does anybody actually make anything out of them yet? Unfortunately micro soldering is not my thing...

 

[wus]

I can't find the Philips bulb with 200W/lm. Do you have a link for me, please?

 

[wus]

I found it now, and read up about it a bit. For me - and my application, general (home) lighting - it's inacceptable: CRI is only 80, and - probably more important - Philips allows these bulbs an SDCM of 6. My experience is that lamps should have a min. CRI of 90 in uncritical places. Where truely good light is wanted, it should be 95 or better. And where several bulbs are used, SDCM should be no worse than 3.

Already in some of my earlier lighting projects I noticed that there is a tradeoff to be made between good color rendition and efficacy. When I started reading this thread my first thought was "how miserable must be the CRI of LEDs with 200lm/W" ... well with 80 it's better than I thought, but definitely not good enough for me.

Of course if you just want emergency lighting to last as long as possible running off a given battery, such LEDs are a good choice.

 

[alpg88]

the reason you get 70 to 100v per filament is cuz you measuring the wrong thing, anytime you measure a voltage of an individual component in a series string under power, you will see voltage of entire string, not the voltage the individual component consumes.

 

[Dave_H]

the reason you get 70 to 100v per filament is cuz you measuring the wrong thing, anytime you measure a voltage of an individual component in a series string under power, you will see voltage of entire string, not the voltage the individual component consumes.

How is that? I confirmed in my 2S2P case two filament pairs wired in series with total supply (across both) voltage around 154vdc, with 77v across each filament. There was almost no ac component.

Full-wave rectified 115vac would be around 162vdc. Allowing for drop in the regulating circuit (linear), this makes sense. Efficiency in this case is not bad, around 95%.

Dimming down the filaments I can see about two dozen small dots, so at around 3v each wired in series, agrees with above.


Dave

 

[jtr1962]

This might be of interest:



The filaments in this bulb are apparently running at 100VDC.

CRI is only 80 but I suppose that's better than 70s as previously suspected. Bulb efficiency is 210 lm/W. While I prefer CRI 90 or 95 for most of my lighting this would be perfect for utility areas or outdoor lighting where the bulbs might be running a long time. The 50,000 hour lifetime is another thing this bulb has in its favor.

Philips should make a CRI 90 or 95 version. It might not exceed 200 lm/W, but my guess is it should be able to manage 160 to 180, depending upon the phosphors and CCT.

By my calculations the bulb should put out ~1.5 watts of waste heat, which means it should be more or less at room temperature. The driver might be a little warmer, but not alarmingly so.

 

[106Lake]

By my calculations the bulb should put out ~1.5 watts of waste heat, which means it should be more or less at room temperature. The driver might be a little warmer, but not alarmingly so.

This is something I love, I tried this on purpose, left one running for a few hours and put it in my hand. Every part felt like room temperature. Outside I found a big hornet just sitting on top of the bulb trying to get a tan.

I literally burnt myself on the LED bulbs I replaced, I think they were TCP brand, not dollar store, even though they don't waste a huge amount the heat is put into a concentrated area and they all cook the components.

 

[JustAnOldFashionedLEDGuy]

Bridgelux VERO 8 hit over 200 lpw at low drive current. A lot of LEDs do. It's in the datasheet but need to learn how to read the graphs.

 

[106Lake]

Bridgelux VERO 8 hit over 200 lpw at low drive current. A lot of LEDs do. It's in the datasheet but need to learn how to read the graphs.

"A lot of LEDs do"
Then why is everybody having so much trouble coming up with examples? Why are philips 60/100w replacement bulbs the only things that take advantage of this?

A quick google search shows this, which is interesting:

https://www.digikey.com/en/products/detail/bridgelux/BXRC-40E2000-C-83/15520040?utm_adgroup=General&utm_source=google&utm_medium=cpc&utm_campaign=PMax%20Shopping_Product_Zombie%20SKUs&utm_term=&utm_content=General&gclid=Cj0KCQjw3a2iBhCFARIsAD4jQB0vlytI_hK8QJyCYG2gocUkXkq4e2GxIAqMpkcNeRDQJIZN1mFUWxcaAlUwEALw_wcB

It does show 184lm/w at test current.

 

[alpg88]

I found it now, and read up about it a bit. For me - and my application, general (home) lighting - it's inacceptable: CRI is only 80, and - probably more important - Philips allows these bulbs an SDCM of 6. My experience is that lamps should have a min. CRI of 90 in uncritical places. Where truely good light is wanted, it should be 95 or better. And where several bulbs are used, SDCM should be no worse than 3.

Already in some of my earlier lighting projects I noticed that there is a tradeoff to be made between good color rendition and efficacy. When I started reading this thread my first thought was "how miserable must be the CRI of LEDs with 200lm/W" ... well with 80 it's better than I thought, but definitely not good enough for me.

Of course if you just want emergency lighting to last as long as possible running off a given battery, such LEDs are a good choice.

CRI is great and all, but there is another aspect of led bulbs that not many are aware of. some bulbs do not have properly designed circuit and it "strobes" , you do not notice it with a naked eye, but some feel it, after a while some complain of a headache. you can spot that "strobes" with a camera. i had a bulb like that in my kitchen, whenever it was on, my camera had lines moving up the screen. once i changed it, lines disappeared. i had no headaches from it, but plenty of people complain they do have headaches after replacing bulbs to led ones. very common issue, called high frequency flicker.
if i had to pick a 200lm/w bulb with a HFF over 100lm/w with no flicker, that 200lm/w bulb would go starlight into a garbage can