Why haven't filament LED bulbs taken over the market?

[TimNJ]

Do the Philips bulbs show a significantly higher quality than similar non-branded ones?

I'm not entirely convinced they're much different than any of the no name bulbs. Aishi cap and Chinese driver IC. Doubt that any of the no name bulbs are much different. Does Philips even manufacture the filaments? You'd think they would produce their own to reduce unit cost, but maybe not.

Only thing you might get with buying Philips (or other big names) is more thorough life testing of the product before pushing it out on the market. Maybe increased quality control? (Ironically, one of the six bulbs I bought was defective.)

[coppice]

Has anyone tried measuring the temperature of the LEDs in these filament style lamps? These things used to be low consumption decorative devices. Now they are making bulbs that consume 10W or more there's quite a lot of heat to remove, and apparently only poor heat conduction paths to remove it.

[IanB]

On the flicker question and design of the bulbs, there must be some differences between brands and sources because my samples of filament bulbs do not have any observable flicker. I have waved my hands and fingers across them and the light is continuous.

On the heat generation and heat removal, I do not think there is much of a thermal problem. There are thin filaments suspended in free air, allowing heat to be removed by convection and radiation over quite a large surface area. If you think how hard it is to heat up a wire suspended in free air you can get a feel for the cooling efficiency of the arrangement. If I put my hand on top of the glass bulb after the lamp has been on for a while it does not get especially hot.

[mikeselectricstuff]

Has anyone tried measuring the temperature of the LEDs in these filament style lamps? These things used to be low consumption decorative devices. Now they are making bulbs that consume 10W or more there's quite a lot of heat to remove, and apparently only poor heat conduction paths to remove it.

I just got some loose filaments. rated 0.9W 60-65W, 10mA. Obviously these figures are not self-consistent.
At 10mA, Vf =56V, 560mW, surface temp reaches 60 deg.C when vertical in free air
at 16mA ~900mw, Vf=57V, 78 deg.
This is measured with a small thermocouple in contact with the side.
Surface area is approx 85mm^2
 

[mikeselectricstuff]

Here's a datsheet for some osram filaments
http://docs-europe.electrocomponents.com/webdocs/1603/0900766b81603b8c.pdf

[mikeselectricstuff]

BTW the one I had started flashing on & off at 40mA / 140 deg. C !

[coppice]

Has anyone tried measuring the temperature of the LEDs in these filament style lamps? These things used to be low consumption decorative devices. Now they are making bulbs that consume 10W or more there's quite a lot of heat to remove, and apparently only poor heat conduction paths to remove it.

I just got some loose filaments. rated 0.9W 60-65W, 10mA. Obviously these figures are not self-consistent.
At 10mA, Vf =56V, 560mW, surface temp reaches 60 deg.C when vertical in free air
at 16mA ~900mw, Vf=57V, 78 deg.
This is measured with a small thermocouple in contact with the side.
Surface area is approx 85mm^2

That's kinda what I was expecting. So, you put a bunch of those close together, massively reduce airflow with a small sealed enclosure, and make that enclosure of a material with poor thermal conduction. Those LEDs are going to get rather toasty.

[Richard Crowley]

That's a great price. Are they 12-15W? Different manufacturers have different opinions on what 100W equiv is.

FEIT brand, 15W

[TimNJ]

On the flicker question and design of the bulbs, there must be some differences between brands and sources because my samples of filament bulbs do not have any observable flicker. I have waved my hands and fingers across them and the light is continuous.

On the heat generation and heat removal, I do not think there is much of a thermal problem. There are thin filaments suspended in free air, allowing heat to be removed by convection and radiation over quite a large surface area. If you think how hard it is to heat up a wire suspended in free air you can get a feel for the cooling efficiency of the arrangement. If I put my hand on top of the glass bulb after the lamp has been on for a while it does not get especially hot.

Not sure if it's been mentioned, but some filament bulbs are filled with an inert gas like helium which facilitates better heat transfer from the filament to the glass globe.

[james_s]

At one point I read about a prototype LED filament bulb that was filled with helium due to the high thermal conductivity of that gas. That was one of the first filament style LED bulbs I ever heard of so I don't know if that technique is used in the bulbs available now.

[TimNJ]

At one point I read about a prototype LED filament bulb that was filled with helium due to the high thermal conductivity of that gas. That was one of the first filament style LED bulbs I ever heard of so I don't know if that technique is used in the bulbs available now.

I see. Plus, that sounds like it would require a somewhat more advanced/expensive manufacturing process. Don't think the Philips bulb I took apart yesterday had an air tight seal though it did have this weird crumbly, yellow, glassy adhesive around the base.

[james_s]

Apparently it used the same manufacturing process as traditional incandescent lamps, replacing the argon/nitrogen gas fill with helium and the tungsten filament with LED filaments, the reuse of existing production lines was touted as one of the advantages. Of course if one could dispense with the airtight bulb and gas fill entirely that would be even simpler.

[CatalinaWOW]

So now we know what limits the life of these things.  How long for the helium to escape.  It is slippery stuff.

[james_s]

Well we don't even know if the helium filled tech went anywhere, but slippery as it is, I don't think helium can diffuse through hermetically sealed glass. I have decades old hard-seal HeNe laser tubes that still work just fine. The old soft seal type would fail though as helium diffused through the epoxy attaching the mirrors.

[TimNJ]

Apparently it used the same manufacturing process as traditional incandescent lamps, replacing the argon/nitrogen gas fill with helium and the tungsten filament with LED filaments, the reuse of existing production lines was touted as one of the advantages. Of course if one could dispense with the airtight bulb and gas fill entirely that would be even simpler.

Thought the same thing initially but then thought the electronics in the base might make it tricky, but probably not.

What do you mean by gas fill it entirely?

[james_s]

I meant dispense with the airtight bulb, which also means eliminate the gas fill because the bulb is not sealed anyway. Bulb is just filled with atmospheric air, should be cheaper and simpler than any special fill.

One thing that helps with these is they use a large number of LED chips run at a much lower current than the power LEDs in most LED bulbs. That should in theory result in higher efficiency and less heat.

[hayatepilot]

One thing that helps with these is they use a large number of LED chips run at a much lower current than the power LEDs in most LED bulbs. That should in theory result in higher efficiency and less heat.

Apparently it doesn't help just in theory. Just look at that datasheet that mikeselectricstuff posted above:
172lm/w efficiency at 2700k! 
That is insane 

[jmelson]

Not sure if it's been mentioned, but some filament bulbs are filled with an inert gas like helium which facilitates better heat transfer from the filament to the glass globe.

Hmm, I kind of doubt it.  First, Helium has gotten quite expensive, lately.  Second, thin plastic shells glued together will not hold Helium in for more than 15 minutes.  It passes through glass pretty easily, too.

Jon

[james_s]

Hmm, I kind of doubt it.  First, Helium has gotten quite expensive, lately.  Second, thin plastic shells glued together will not hold Helium in for more than 15 minutes.  It passes through glass pretty easily, too.

Jon

No it doesn't pass through glass easily. As I already mentioned I have helium-neon laser tubes that are made of glass and decades old that still contain their helium. It certainly would not stay inside a plastic bulb though.

This is the bulb I was thinking of http://lamptech.co.uk/Spec%20Sheets/LEDi%20Philips%20470A60IF827-E27%20Midpower-in-Glass.htm

My memory was wrong of it being a filament type, it appears to be more conventional.

[tooki]

Not sure if it's been mentioned, but some filament bulbs are filled with an inert gas like helium which facilitates better heat transfer from the filament to the glass globe.

Hmm, I kind of doubt it.  First, Helium has gotten quite expensive, lately.  Second, thin plastic shells glued together will not hold Helium in for more than 15 minutes.  It passes through glass pretty easily, too.

Jon

They're not plastic shells, never mind glued. They're traditional glass bulbs. As james_s just explained a few posts above yours, part of the impetus of designing the LED filaments was that it allows all the existing incandescent bulb manufacturing equipment and processes to be repurposed with only minimal alteration.

Apparently it used the same manufacturing process as traditional incandescent lamps, replacing the argon/nitrogen gas fill with helium and the tungsten filament with LED filaments, the reuse of existing production lines was touted as one of the advantages. Of course if one could dispense with the airtight bulb and gas fill entirely that would be even simpler.

[coppice]

This is the bulb I was thinking of http://lamptech.co.uk/Spec%20Sheets/LEDi%20Philips%20470A60IF827-E27%20Midpower-in-Glass.htm

They don't mention why there is 20% oxygen with the helium. Does anyone know?

[IanB]

They don't mention why there is 20% oxygen with the helium. Does anyone know?

As far as I know, commercial helium is supplied that way because people love to breathe it in and play with squeaky voices. If the helium was 100% pure then people would tend to die of asphyxiation, which is bad for business. So they mix oxygen with it to make it equivalent to air so it is safe to breathe.

[HighVoltage]

Filament LED bulbs have been available in Germany for at least two years.
They are called "LED-Leuchtfaden" here.

And I have been using them about that long and so far no failure.

[james_s]

I was out today so I bought a couple new types of LED filament bulbs to play with, both are Ecosmart brand from Home Depot and both claim 800 lumens. One is a 6.5W clear bulb in a compact envelope the size of a standard appliance bulb, if the ratings are believed that's a very impressive 123 lm/W. The other is a more conventional A19 size, inside frosted so it looks exactly like an ordinary incandescent bulb, it's 7.5W so slightly less efficient, likely owing to the diffuse coating. Both have a reasonably nice 2700k light, CRI is nothing spectacular, probably around 80. They run cool and put out a lot of light, and to a casual observer they look incandescent, they even dim quite nicely on a dimmer.

There is one fatal flat though for me that will restrict these to specific uses, they flicker. My other half can't see the flicker at all but it's very visible to me, especially if I scan my eyes across the room looking from one thing to another or move my hand. I suspect it's unavoidable unless we have some kind of breakthrough in compact capacitors in high enough ratings that will fit down inside the base. That right there is probably why they remain somewhat of a niche product.

[IanB]

See if you can find the Ecosmart brand. They don't seem to have any flicker. (6.5 W, 810 lumens)