Omnidirectional LED light bulb is OK?

[treez]

Hello,
The below (attached) is an omnidirectional LED lightbulb. It comprises seven PCBs, each one "aiming" in a different direction so that the light is spread out evenly, just like in one of the old incandescent bulbs.
Do you think that this solves the problem of the individual LEDs narrow beam-ness?
Surely a 15W version of this would be brighter than an old 100W incandescent bulb?

The LED PCBs would all be big enough, and comprise enough cooling copper "land" area , such that no aluminium heatsink was required. Making aluminium from bauxite takes prodigious amounts of energy, so aluminium should not be used in heatsinks for lightbulbs....do you agree?

 

[treez]

Hello again,
Is there anything intrinsically wrong with this thread?.....I posted a similar thread on a different forum, and very politely explained the whole purpose, and yet it was hotly rebuked, and after 49 threads, that forum deleted all 49 threads.

The purpose of this product is to do domestic mains lighting with repairable LED lightbulbs...so that no glass gets thrown away like it does with CFLs (since not all of them get returned to the CFL collection points).
We also want to avoid aluminium heatsinks, as these just end up getting smelted back down when the bulb dies...this is a waste of energy.

 

[yassin.kraouch]

Hello,

Surely a 15W version of this would be brighter than an old 100W incandescent bulb?

it depends on the lumen output of this LED bulb, but surely if it will have a lumen output higher than 1200 lm it will be brigher,

The LED PCBs would all be big enough, and comprise enough cooling copper "land" area , such that no aluminium heatsink was required. Making aluminium from bauxite takes prodigious amounts of energy, so aluminium should not be used in heatsinks for lightbulbs....do you agree?

not agree because many of LED light bulb use aluminium as heat sink, and the problem with LEDs is always how to cool the LED die, and how to transfer light from die to outside the PCB

 

[treez]

**broken link removed**

**broken link removed**


...but the above two are single sided LED PCBs, which are specified for use without a heatsink, and do up to 30W.

 

[yassin.kraouch]

you are not obliged every time to use heatsink,
if you do not exceed LED junction temperature, even single side PCb is sufficient,
you have to take care in the layout of the PCB, and use copper with higher thickness,
perhaps that the product you proposed have a limited lifetime,

 

[betwixt]

There are two problems I see in your reasoning:

1. Smelting Bauxite to produce Al is very energy intensive but it is one of the most recyclable materials around and with it's very low melting point, the recycling melt is probably quite economical. I have made casts at home using a plumbers gas torch to melt old overhead power cables, it was quite easy to do.

2. the multi-faceted PCB may be quite good at lighting a small area. The light cone will be the sum of the cones of each LED though and that will give quite a confined circle of illumination. A conventional lamp is unfocussed and much of the light aims upward and reflects off the surrounding ceiling where it can spead over a wider area.

An idea for you: given that the purpose of the Al is to spread the heat so it can be radiated or convected from a larger area, why not make a similar sized enclosure but filled with non electrical conducting liquid as the transfer agent. I'm thinking of a silicone gel that can be poured and has good heat transfer characteristics and also protects all the internal wiring/components from water ingress.

Brian.

 

[treez]

perhaps that the product you proposed have a limited lifetime,

..the HarvardEng LED PCB in post #4 above is specified as having 50,000 hours of lifetime.

why not make a similar sized enclosure but filled with non electrical conducting liquid as the transfer agent. I'm thinking of a silicone gel that can be poured and has good heat transfer characteristics and also protects all the internal wiring/components from water ingress.

....this sounds like a good concept..(hopefully decently thermally conductive silicone) -kind of using the silicone gel like a potting compound around the LED PCB.(?)
Do you think that this idea would be effective when the LED PCB is contained inside a completely sealed, transparent plastic enclosure, with the silicone around the LED PCB?
With the enclosure being absolutely sealed, would it not mean that the silicone would receive heat from the LEDs, but then have nowhere to conduct the heat off to....and thus the LED PCB would just end up being surrounded by a lot of heated silicone? (maybe this is still a good idea as the leds woul dbe cooler than without the silicone?)
If the enclosure was perfectly sealed, then would surrounding the LED PCB by silicone be more effective at keeping the LEDs cool than just having the LED PCB inside a sealed, transparent plastic enclosure containing just air around the LED PCB?

the multi-faceted PCB may be quite good at lighting a small area. The light cone will be the sum of the cones of each LED though and that will give quite a confined circle of illumination

...I wonder if it would be best to just do a single large , round PCB like the PCBs shown in post #4 above, and just direct it downwards. It would certainly be cheaper, and maybe such a PCB would end up giving sufficient light at wider angle anyway. Page 11 of the following Cree xpe led datasheet shows how badly the led light drops off as one goes outwards from the zenith.....

Cree XPE LED datasheet:
https://www.google.co.uk/url?sa=t&r...NJJtkxlVDa8ImxuMg&sig2=yMRUgH9mEhbhbEM1euWoBg


One other reason to use many small leds (& no alu heatsink), instead of a few bigger power leds (but with an alu heatsink), is that bigger power leds comprise a central thermal pad which would be difficult to hand solder when the LED PCB is being repaired after its lifetime use. (you can see the thermal pad on page 14 of the golden dragon datasheet)....

Osram golden dragon 3W power led:
**broken link removed**

...you see, in the other forum where this whole idea was rebuked, they seemed to think that it would be best to simply use an alu heatsink, of a standard size, and literally re-use that heatsink. -That all sounds very well, however, firstly, how are we going to make multiple private companies come together and agree on one (or a few) standardised heatsink shapes? Also, as just mentioned, if an alu heatsink is used, then we'll obviously be using a few small, higher power leds, (since the heatsink is there), and thus the problem with desoldering and re-soldering leds to the PCB when the LED PCB eventually fails and needs repair (due to the difficulty of hand soldering the thermal pad)

Anyway, this whole idea of repairable domestic mains lightbulbs was quite viciously rebuked in a different forum....there were 49 posts in the thread, where the idea was passed off as almost lunacy, then they deleted the entire thread, and banned me from the forum. Is this idea that bad?

In that (now extinct) thread, I mentioned that the repairable bulb was LED based, and at least LED bulbs don't contain mercury, but this was rebuked, and their forum members told me that 20 cans of tuna contains as much mercury as one CFL.....so , they said, "just don't *eat* CFLs or tuna". I knew that the mercury_in_CFLs problem was not that bad, but I didn't know it was that insignificant?

 

[betwixt]

Mercury IS a problem. The argument that the amount is small is flawed, it is based on the CFL using less mercury than would be emitted to the atmosphere if coal was used to create the equivalent light output using an incadescent lamp. Coal contains traces of mercury, albeit very small amounts. It assumes that less mercury is released in the production and destruction of a CFL than in the fuel that otherwise would be used for lighting. The flaw is that not all electricity is generated by burning coal and the mercury levels in coal vary considerably from one source to another. Of course, eating LEDs isn't a good idea either but at least the chemicals in them are more stable and less likely to leach into food/water sources. Did they mention the toxic paint in CFLs I wonder?

Brian.

 

[treez]

Did they mention the toxic paint in CFLs I wonder?

they didn't mention this, is it worse than the mercury?

Anyway, may I enquire if any readers are for or against the idea of a re-usable domestic mains lightbulb?
i.e., a lightbulb where the LED PCB, when failed with age, gets new leds soldered to it, and then sent back out into service.
The SMPS would also hopefully be repairable....or at least we would do it with film caps, so itlasted a long time.
Do you agree that the bulb can't have glass in it, so CFLs are out.
That leaves us with LED.
So do you think it should be a small led pcb + alu heatsink or
Large LED PCB with no alu heatsink.
If you want the heatsink, then how do you propose to arrange things such that the alu heatsinks are re-used and not smelted back down when the bulb is recycled. (I know alu smelting is relatively low energy, but its still a fair bit of energy with 680c melt point)

 

[betwixt]

I think the absence of replies answers your question!

Lighting is so widespread and mass manufactured that the cost has to be rock bottom and I think that makes repair costs unreasonably high in comparison. A more reliable PSU for existing LED lights is probably what is needed. If a company produced an efficient lamp with a life guarantee of several years it would be a winner, at the moment they just quote they last longer than an incandescent lamp but never specify the conditions of testing. One of the problems with LEDs is they are too fast to respond to the current changing through them and therefore need more than just a rectified AC supply or they flicker. If the capacitors needed to filter the DC could be eliminated it would go a long way to improving reliability.

And yes, the paint is highly toxic, as is the Thorium used to coat the heater filaments.

Brian.

 

[treez]

Recyclable led lights hardly grasps the imagination, I would admit that...its a rather dull aspect of the 'glitzy' world of hi-tech, sure.

But LED light bulbs are retailing at £25 and I think it would be cost effective to repair led pcbs........
£25 12W LED lightbulb
https://www.leds4less.co.uk/led-100...-lamp-12w-instant-light-bayonet-b22-194-p.asp

Also, the UK government now has a zero waste policy, where everything must be recycled, and certainly in Gloucestershire, we have scores of workers literally hand sifting through literally every scrap of waste to pull out the aluminium cans, newspapers etc etc.
...they will be tasked with also pulling out alu heatsinks from led bulbs....as well as the pcbs, because inevitably, many people will just lob them in the trash when theyre dead and not take them to proper recycling points.......if the government is willing to pay for this kind of total rubbish sifting, then its really just part of that process to repair stuff like led lightbulbs......you could even pinch a few of the rubbish sorting staff to come and do the solder repairs on the led pcbs..

Your point about caps is a good one...and I think its the electrolytic cap that is the Achilles heel of the led lightbulb......I reckon replacing them with polystyrene film caps is the answer?...

 

[Jibby Benjamen]

No I don’t agree. Aluminum is the best solution, As because copper equivalent in lieu of aluminum is much more costly

 

[treez]

yes but you re-use the copper

 

[chuckey]

I personally think that LED bulb at £10 are a bargain compared with the downlighter that they are fixed in a £5. I reckon with the long life of LEDs they should be made integral to the lamp holder, so you buy the lot at one go. The design of our light fittings is still constrained by the filament bulb which is about 100 years old. The same argument should be made for CFLs, trying to wrap them up to make them into a globe-like thing, then hanging them up with an external shade, seems irrational to me. With a larger housing the CFLs and LEDs will run cooler and more reliable components could be used.
Frank

 

[treez]

Incidentally, if you look at this link, somebody has now done this idea...

https://www.kickstarter.com/projects/nanoleaf/nanoleaf-bloom-the-first-bulb-that-dims-without-a

 

[betwixt]

Curiously, nowhere does it mention how the user controls the light level, only a brief mention that it may be compatible with existing dimmer switches.

Brian.

 

[treez]

I reckon they'll dim it via the lightswitch being ON.............OFF/ON'd.
Or there will be a bit of circuitry in the switch plate which lops out a mains half cycle to the bulb when a button on the switchplate is pressed, and this results in it getting dimmed down to the next level.

Dimming via an "OFF/ON" of the switch is so obviously a great idea, that im amazed nobody is doing it.......it requires no extra signal wire, and no dimmer fitting/installation.

 

[D.A.(Tony)Stewart]

There are too many false assumptions in this thread.

If repairability is key with high recycle value, then standards are essential. ZAGRA is one such standard.

Single LED repairs compromise quality or cost due to Blue-shift and LM70 aging mismatches.

Cost of cooling is not always feasible by just spreading out the SMD's. This is not scaleable as ambient rises in centre.

I prefer the ideal bulb replacement using a BLUE LED with remote phosphor on bulb with very thin tubes for heat removal with chimney flow.

Amalgam is toxic as 9 tins of Tuna due to significant reduction of mass in new CFL's and reduction of soluble Mercury in the compound itself.

LED bulbs that are dimmable are currently $0.5/W in PAR20,35 and will migrate towards the cost of SMDs which are 0.1$/W in volume.

Whereas retail started at $2/W then $1/W and luminaires much higher.

- - - Updated - - -

The ON/OFF problem in future must satisfy active PFC solutions to reduce burden on distribution power transformers with harmonic and neutral currents. This adds some cost but will become mandatory in future.

- - - Updated - - -

They can make FL tubes now with same qualities as LED bulbs with CRI index of >=80 using tri-phosphor and efficacy of >=88 LPW and last 50kh

but not as good as those with >120LPW.

 

[betwixt]

In days of old they used to send Morse code by flashing light bulbs.
If Treez is right, it means we now have to send Morse code TO the light bulbs using the power switch. !

I doubt that is the method used, it gives no control over ramp up / ramp down and if user timing is required, I can imagine a lot of frustrated people who can't set the level they want.

Brian.

 

[Audioguru]

The street lights and traffic lights in my city are all LED type. Of course they have pretty big aluminum heatsinks.
Each light uses at least 30 LEDs and after one year some LEDs can be seen not working. Some entire lights were dead when they were brand new.
Nobody is going to repair them.