Just reading of LED needing to be kept cool and I was wondering why. Is it
because standard components are used which will fail, or is there some
electrical principle that will always cause failure at a particular
temperature? Could some re-design make them happier at higher temperatures?

DerbyBorn wrote:
Just reading of LED needing to be kept cool and I was wondering why. Is it
because standard components are used which will fail, or is there some
electrical principle that will always cause failure at a particular
temperature? Could some re-design make them happier at higher temperatures?

All semiconductors need to be kept relatively cool. A max of about 200C
comes to mind from years ago.

I think its a problem of removing the heat fast enough. Most semiconductors
become problematic if they get too hot.
Brian

--
----- -
This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...

Blind user, so no pictures please!
"DerbyBorn" wrote in message
2.222...
Just reading of LED needing to be kept cool and I was wondering why. Is it
because standard components are used which will fail, or is there some
electrical principle that will always cause failure at a particular
temperature? Could some re-design make them happier at higher
temperatures?

Germanium was very bad indeed, you could not get that semiconductor very hot
before thermal runaway happened.
I don't know the figures for Galium arsanide and the other materials used
in LEDs.
Brian

--
----- -
This newsgroup posting comes to you directly from...
The Sofa of Brian Gaff...

Blind user, so no pictures please!
"Bob Minchin" wrote in message
...
DerbyBorn wrote:
Just reading of LED needing to be kept cool and I was wondering why. Is
it
because standard components are used which will fail, or is there some
electrical principle that will always cause failure at a particular
temperature? Could some re-design make them happier at higher
temperatures?

All semiconductors need to be kept relatively cool. A max of about 200C
comes to mind from years ago.

On 09/11/2016 15:48, DerbyBorn wrote:

Just reading of LED needing to be kept cool and I was wondering why. Is it

Diffusion kills semiconductors when they get too warm and it also cooks
the phosphor used in white LEDs. Some show obvious signs of scorching
after a length of time in service. There are images online somewhere.

Efficiency also suffers when they get warm as more leakage current flows
through a hot semiconductor junction without making light.

http://www.lrc.rpi.edu/programs/nlpi...s/led/heat.asp

Although I think their curves are a bit out of date (ie pessimistic).

because standard components are used which will fail, or is there some
electrical principle that will always cause failure at a particular
temperature? Could some re-design make them happier at higher temperatures?

You can design semiconductor devices to survive at higher temperatures
but in domestic lighting applications the weakest link is usually the
capacitors in their power supply rather than the actual LED itself.

For white LEDs the phosphor tight up against the emitting blue LED die
is a weak point in that it is operated close to its limit and suffers
photon damage losing output in the longer term even if it doesn't char.

--
Regards,
Martin Brown

On 09/11/16 16:21, Bob Minchin wrote:
DerbyBorn wrote:
Just reading of LED needing to be kept cool and I was wondering why.
Is it
because standard components are used which will fail, or is there some
electrical principle that will always cause failure at a particular
temperature? Could some re-design make them happier at higher
temperatures?

All semiconductors need to be kept relatively cool. A max of about 200C
comes to mind from years ago.

200C storage. Genereally 180C operating junction temps.

But lifetime is a lot less at that sort of temp.

Dopant actually migrates through the semiconducting barriers as they
age, and heat speeds it up.

--
€œBut what a weak barrier is truth when it stands in the way of an
hypothesis!€�

Mary Wollstonecraft

On 09/11/2016 17:12, Martin Brown wrote:
On 09/11/2016 15:48, DerbyBorn wrote:

Just reading of LED needing to be kept cool and I was wondering why.
Is it

Diffusion kills semiconductors when they get too warm and it also cooks
the phosphor used in white LEDs. Some show obvious signs of scorching
after a length of time in service. There are images online somewhere.

Efficiency also suffers when they get warm as more leakage current flows
through a hot semiconductor junction without making light.

http://www.lrc.rpi.edu/programs/nlpi...s/led/heat.asp

Although I think their curves are a bit out of date (ie pessimistic).

because standard components are used which will fail, or is there some
electrical principle that will always cause failure at a particular
temperature? Could some re-design make them happier at higher
temperatures?

You can design semiconductor devices to survive at higher temperatures
but in domestic lighting applications the weakest link is usually the
capacitors in their power supply rather than the actual LED itself.

For white LEDs the phosphor tight up against the emitting blue LED die
is a weak point in that it is operated close to its limit and suffers
photon damage losing output in the longer term even if it doesn't char.

Another problem with many LED lamps is the electronic drivers don't like
heat much either. So if your lamp base has an electrolytic cap in it (as
many do) its not going to last long at over 100 degrees.


--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

In article ,
Bob Minchin wrote:
DerbyBorn wrote:
Just reading of LED needing to be kept cool and I was wondering why.
Is it because standard components are used which will fail, or is
there some electrical principle that will always cause failure at a
particular temperature? Could some re-design make them happier at
higher temperatures?

All semiconductors need to be kept relatively cool. A max of about 200C
comes to mind from years ago.

Capacitors too have a shorter life with heat. And SWPS rely on fairly high
stressed caps.

--
*You're never too old to learn something stupid.

Dave Plowman London SW
To e-mail, change noise into sound.

All semiconductors need to be kept relatively cool. A max of about
200C comes to mind from years ago.

Capacitors too have a shorter life with heat. And SWPS rely on fairly
high stressed caps.


I have some kitchen downlights (Halers) where the driver assembly is
separate from the lamp unit. They have a 7 year warranty - I guess this is
the way to go. Imagine the electronics in a "ceiling rose) and the lamp on
a pendant. I guess that would work. However, makers have felt pressure to
make everything fit a conventional lamp envelope.

On 10/11/2016 10:07, DerbyBorn wrote:

All semiconductors need to be kept relatively cool. A max of about
200C comes to mind from years ago.

Capacitors too have a shorter life with heat. And SWPS rely on fairly
high stressed caps.


I have some kitchen downlights (Halers) where the driver assembly is
separate from the lamp unit. They have a 7 year warranty - I guess this is
the way to go. Imagine the electronics in a "ceiling rose) and the lamp on
a pendant. I guess that would work. However, makers have felt pressure to
make everything fit a conventional lamp envelope.

Yes, we're in a transition phase at the moment. The problems are really
with LED lamps which have to retrofit an existing, often unsuitable,
luminaire. The purpose made LED lights should be much better in this
respect, and with the greatly improved lifetime, replacing a whole unit
rather than just a lamp shouldn't cost more in the long run.

Cheers
--
Syd

On Wednesday, 9 November 2016 18:02:43 UTC, John Rumm wrote:
On 09/11/2016 17:12, Martin Brown wrote:
On 09/11/2016 15:48, DerbyBorn wrote:

Just reading of LED needing to be kept cool and I was wondering why.
Is it

Diffusion kills semiconductors when they get too warm and it also cooks
the phosphor used in white LEDs. Some show obvious signs of scorching
after a length of time in service. There are images online somewhere.

Efficiency also suffers when they get warm as more leakage current flows
through a hot semiconductor junction without making light.

http://www.lrc.rpi.edu/programs/nlpi...s/led/heat.asp

Although I think their curves are a bit out of date (ie pessimistic).

because standard components are used which will fail, or is there some
electrical principle that will always cause failure at a particular
temperature? Could some re-design make them happier at higher
temperatures?

You can design semiconductor devices to survive at higher temperatures
but in domestic lighting applications the weakest link is usually the
capacitors in their power supply rather than the actual LED itself.

For white LEDs the phosphor tight up against the emitting blue LED die
is a weak point in that it is operated close to its limit and suffers
photon damage losing output in the longer term even if it doesn't char.

Another problem with many LED lamps is the electronic drivers don't like
heat much either. So if your lamp base has an electrolytic cap in it (as
many do) its not going to last long at over 100 degrees.

same thing happened with PC PSUs a while back someone had one the other day and I brought some capacitors rated at 130C 2000hrs I think, you can get 105C and the normal range are mostly 85C.
This is why a lot of LED lamps donl;t last the 50K hours stated, if you read the full specs some are only rates at 50K if you use them for less than x hours at a time.

Of course there are some 'ankers on her that will say I'm wrong.
So before tehy embarress themselfs lijke a high wranking academci telling me there no such thing as the factories and workplaces act.....

https://www.ledhut.co.uk/spot-lights...placement.html
Go to the specs tab on the 3.6 Watt GU10 LED Spotlight - 35W Replacement
hover the mouse over the average life hrs the green.

For dave plowman who has trouble reading it says.
"When the bulb is on for 4 hours a day".

and this is another reason I dim LEDs it's not brightnes or to save money on electricity costs.

In article 2,
DerbyBorn wrote:


All semiconductors need to be kept relatively cool. A max of about
200C comes to mind from years ago.

Capacitors too have a shorter life with heat. And SWPS rely on fairly
high stressed caps.


I have some kitchen downlights (Halers) where the driver assembly is
separate from the lamp unit. They have a 7 year warranty - I guess this
is the way to go. Imagine the electronics in a "ceiling rose) and the
lamp on a pendant. I guess that would work. However, makers have felt
pressure to make everything fit a conventional lamp envelope.

True. And the same applied to CFL. Making any energy saving type which
needs a power supply of some sort with the PS built in is invariably a
compromise. I'm rather surprised new standard lighting wiring for LEDs -
perhaps with a central power supply or two - hasn't been devised.

--
*I dropped out of communism class because of lousy Marx.*

Dave Plowman London SW
To e-mail, change noise into sound.

"whisky-dave" wrote in message
...
snip
For dave plowman who has trouble reading it says.
"When the bulb is on for 4 hours a day".

Cheeky monkey.

On 10/11/2016 10:43, Dave Plowman (News) wrote:

True. And the same applied to CFL. Making any energy saving type which
needs a power supply of some sort with the PS built in is invariably a
compromise. I'm rather surprised new standard lighting wiring for LEDs -
perhaps with a central power supply or two - hasn't been devised.


It has, they are 12V usually.

For some reason people fit 240V ones instead.

In article ,
whisky-dave wrote:
Of course there are some 'ankers on her that will say I'm wrong. So
before tehy embarress themselfs lijke a high wranking academci telling
me there no such thing as the factories and workplaces act.....

I'd have thought those academics able to afford teeth?

--
*Income tax service - We‘ve got what it takes to take what you've got.

Dave Plowman London SW
To e-mail, change noise into sound.

In article . com,
dennis@home wrote:
On 10/11/2016 10:43, Dave Plowman (News) wrote:

True. And the same applied to CFL. Making any energy saving type which
needs a power supply of some sort with the PS built in is invariably a
compromise. I'm rather surprised new standard lighting wiring for LEDs -
perhaps with a central power supply or two - hasn't been devised.


It has, they are 12V usually.

I've not seen much of a range of 12v LEDs - apart from MR16 and those made
for cars. Can you give a link to a site that sells them?
But a 12v LED would also include a power supply of some sort.

For some reason people fit 240V ones instead.

They're likely to want something looking like they're accustomed to see.

--
*Bigamy is having one wife too many - monogamy is the same

Dave Plowman London SW
To e-mail, change noise into sound.

On Thursday, 10 November 2016 14:39:35 UTC, Dave Plowman (News) wrote:
In article ,
whisky-dave wrote:
Of course there are some 'ankers on her that will say I'm wrong. So
before tehy embarress themselfs lijke a high wranking academci telling
me there no such thing as the factories and workplaces act.....

I'd have thought those academics able to afford teeth?

Maybe they just like sucking ;-P

"Dave Plowman (News)" wrote in message
...
In article 2,
DerbyBorn wrote:


All semiconductors need to be kept relatively cool. A max of about
200C comes to mind from years ago.

Capacitors too have a shorter life with heat. And SWPS rely on fairly
high stressed caps.


I have some kitchen downlights (Halers) where the driver assembly is
separate from the lamp unit. They have a 7 year warranty - I guess this
is the way to go. Imagine the electronics in a "ceiling rose) and the
lamp on a pendant. I guess that would work. However, makers have felt
pressure to make everything fit a conventional lamp envelope.

True. And the same applied to CFL. Making any energy saving type which
needs a power supply of some sort with the PS built in is invariably a
compromise. I'm rather surprised new standard lighting wiring for LEDs -
perhaps with a central power supply or two - hasn't been devised.

Already there, often with 12V for the LEDs themselves.

Same with low voltage halogens.

"Dave Plowman (News)" wrote in message
...
In article . com,
dennis@home wrote:
On 10/11/2016 10:43, Dave Plowman (News) wrote:

True. And the same applied to CFL. Making any energy saving type which
needs a power supply of some sort with the PS built in is invariably a
compromise. I'm rather surprised new standard lighting wiring for
LEDs -
perhaps with a central power supply or two - hasn't been devised.


It has, they are 12V usually.

I've not seen much of a range of 12v LEDs - apart from MR16 and those made
for cars. Can you give a link to a site that sells them?

http://www.ebay.co.uk/sch/i.html?_fr...v+led&_sacat=0

On 10/11/2016 10:35, whisky-dave wrote:
On Wednesday, 9 November 2016 18:02:43 UTC, John Rumm wrote:

Another problem with many LED lamps is the electronic drivers don't
like heat much either. So if your lamp base has an electrolytic cap
in it (as many do) its not going to last long at over 100 degrees.

same thing happened with PC PSUs a while back someone had one the
other day and I brought some capacitors rated at 130C 2000hrs I
think, you can get 105C and the normal range are mostly 85C. This is
why a lot of LED lamps donl;t last the 50K hours stated, if you read
the full specs some are only rates at 50K if you use them for less
than x hours at a time.

Yup indeed. I retired a GU10 spot the other day that was still working
LED wise, but was flickering because the cap in its base had lost so
much capacitance.


Of course there are some 'ankers on her that will say I'm wrong. So
before tehy embarress themselfs lijke a high wranking academci
telling me there no such thing as the factories and workplaces
act.....

https://www.ledhut.co.uk/spot-lights...placement.html

Go to the specs tab on the 3.6 Watt GU10 LED Spotlight - 35W
Replacement
hover the mouse over the average life hrs the green.

For dave plowman who has trouble reading it says. "When the bulb is
on for 4 hours a day".

It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - thats between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",

which (to me at least) does not seem to imply any maximum duty cycle or
period of on time to achieve that lifetime on that particular product.
(others may differ obviously)



--
Cheers,

John.

/================================================== ===============\
| Internode Ltd - http://www.internode.co.uk |
|-----------------------------------------------------------------|
| John Rumm - john(at)internode(dot)co(dot)uk |
\================================================= ================/

In article ,
whisky-dave wrote:
For dave plowman who has trouble reading it says.
"When the bulb is on for 4 hours a day".

and this is another reason I dim LEDs it's not brightnes or to save
money on electricity costs.

So that's why you keep on having to buy new dimmers?

BTW, have you any data to prove altering the waveform into the LED PS will
actually improve the life? My guess is not.

--
*We have enough youth, how about a fountain of Smart?

Dave Plowman London SW
To e-mail, change noise into sound.

In article ,
John Rumm wrote:
It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that‘s between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


Heh heh. If the average life of an LED was actually 15 years, why don't
they give them a lifetime warrenty? The number of people who could produce
a receipt after 15 years would be very very small.

which (to me at least) does not seem to imply any maximum duty cycle or
period of on time to achieve that lifetime on that particular product.
(others may differ obviously)

Any claim made to the life of a product should be taken as Trump speak
unless backed up with a decent warrenty.

--
*Why is it that rain drops but snow falls?

Dave Plowman London SW
To e-mail, change noise into sound.

"Dave Plowman (News)" wrote in message
...
In article ,
John Rumm wrote:
It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that's between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


Heh heh. If the average life of an LED was actually 15 years, why don't
they give them a lifetime warrenty? The number of people who could produce
a receipt after 15 years would be very very small.

which (to me at least) does not seem to imply any maximum duty cycle or
period of on time to achieve that lifetime on that particular product.
(others may differ obviously)

Any claim made to the life of a product should be taken as Trump speak
unless backed up with a decent warrenty.

Reminds me of our local barber.
OAPs cut free when accompanied by a parent.

On 11/11/2016 00:45, Dave Plowman (News) wrote:
In article ,
John Rumm wrote:
It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that‘s between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",

Heh heh. If the average life of an LED was actually 15 years, why don't
they give them a lifetime warrenty? The number of people who could produce
a receipt after 15 years would be very very small.

The LEDs themselves are incredibly reliable provided that you don't
abuse them. I find dipping the old dim ones in LN2 to demonstrate the
improvement in efficiency with a cooled more rigid crystal lattice tends
to do for them after a while, but indicator LEDs I installed when they
were new and novel in the mid 70's are still going strong. Unlike the
old low voltage indicator bulbs they are almost indestructible.

which (to me at least) does not seem to imply any maximum duty cycle or
period of on time to achieve that lifetime on that particular product.
(others may differ obviously)

Any claim made to the life of a product should be taken as Trump speak
unless backed up with a decent warrenty.

MTBF of bare LEDs driven properly is very long.

The electronics in consumer LED lights is by far the weakest component.
I have seen just one lamp failure due to a single defective LED in a
long chain going down the others were due to cooked PSU capacitors. I
look at all the failed ones (three so far) to see how they failed.

--
Regards,
Martin Brown

On 10/11/16 17:57, John Rumm wrote:
On 10/11/2016 10:35, whisky-dave wrote:
On Wednesday, 9 November 2016 18:02:43 UTC, John Rumm wrote:

Another problem with many LED lamps is the electronic drivers don't
like heat much either. So if your lamp base has an electrolytic cap
in it (as many do) its not going to last long at over 100 degrees.

same thing happened with PC PSUs a while back someone had one the
other day and I brought some capacitors rated at 130C 2000hrs I
think, you can get 105C and the normal range are mostly 85C. This is
why a lot of LED lamps donl;t last the 50K hours stated, if you read
the full specs some are only rates at 50K if you use them for less
than x hours at a time.

Yup indeed. I retired a GU10 spot the other day that was still working
LED wise, but was flickering because the cap in its base had lost so
much capacitance.


Of course there are some 'ankers on her that will say I'm wrong. So
before tehy embarress themselfs lijke a high wranking academci
telling me there no such thing as the factories and workplaces
act.....

https://www.ledhut.co.uk/spot-lights...placement.html


Go to the specs tab on the 3.6 Watt GU10 LED Spotlight - 35W
Replacement
hover the mouse over the average life hrs the green.

For dave plowman who has trouble reading it says. "When the bulb is
on for 4 hours a day".

It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - thats between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


So fas so good...

which (to me at least) does not seem to imply any maximum duty cycle or
period of on time to achieve that lifetime on that particular product.

???WTF does '4 hours a day' mean if it doesn't mean 'duty cycle'

30,000 hours is about 4 years if you leave it on all the time.

Around 1960 my late data rigged up a fluorescent tube in the garage, he
died a couple of years later, but that tube was still there working,
though it flickered badly, when the house was sold in 2004.

It probably was on for less than an hour a month.




(others may differ obviously)





--
"What do you think about Gay Marriage?"
"I don't."
"Don't what?"
"Think about Gay Marriage."

On 11/11/2016 00:45, Dave Plowman (News) wrote:
In article ,
John Rumm wrote:
It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that‘s between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


Heh heh. If the average life of an LED was actually 15 years, why don't
they give them a lifetime warrenty? The number of people who could produce
a receipt after 15 years would be very very small.


If you sold 10 million of them and 0.01% claimed each year you would
need quite a few people to keep up with the claims.
It just wouldn't be economical on the margins they make.

One car manufacturer offered lifetime warranties for a while but they
stopped doing it.

On 11/11/2016 08:54, The Natural Philosopher wrote:

Around 1960 my late data rigged up a fluorescent tube in the garage, he
died a couple of years later, but that tube was still there working,
though it flickered badly, when the house was sold in 2004.

It probably was on for less than an hour a month.

You would never use a fluorescent tube if the life was that poor.

In article ,
Martin Brown wrote:
Heh heh. If the average life of an LED was actually 15 years, why don't
they give them a lifetime warrenty? The number of people who could produce
a receipt after 15 years would be very very small.

The LEDs themselves are incredibly reliable provided that you don't
abuse them. I find dipping the old dim ones in LN2 to demonstrate the
improvement in efficiency with a cooled more rigid crystal lattice tends
to do for them after a while, but indicator LEDs I installed when they
were new and novel in the mid 70's are still going strong. Unlike the
old low voltage indicator bulbs they are almost indestructible.

Yes - same here. But once you start driving a LED hard to use as a light
source, the life comes down.
I've had a couple of early ones explode. And they claimed a very long life
too. Given their high cost put me off them - especially since the light
they produced was terrible.

--
*I'm pretty sure that sex is better than logic, but I can't prove it.

Dave Plowman London SW
To e-mail, change noise into sound.

In article . com,
dennis@home wrote:
On 11/11/2016 00:45, Dave Plowman (News) wrote:
In article ,
John Rumm wrote:
It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that‘s between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


Heh heh. If the average life of an LED was actually 15 years, why don't
they give them a lifetime warrenty? The number of people who could produce
a receipt after 15 years would be very very small.


If you sold 10 million of them and 0.01% claimed each year you would
need quite a few people to keep up with the claims.
It just wouldn't be economical on the margins they make.

Are you then saying none ever fail within the existing warranty period?
Most things are likely (percentage wise) to fail early on. If not, last
the design life before failing. The bathtub effect.

One car manufacturer offered lifetime warranties for a while but they
stopped doing it.

How long a car lasts depends on lots of things. Like type of use and
servicing, etc. That doesn't apply to a lamp.

--
*Isn't it a bit unnerving that doctors call what they do "practice?"

Dave Plowman London SW
To e-mail, change noise into sound.

Rod Speed wrote:

"Dave Plowman (News)" wrote in message
...
In article 2,
DerbyBorn wrote:


All semiconductors need to be kept relatively cool. A max of about
200C comes to mind from years ago.

Capacitors too have a shorter life with heat. And SWPS rely on fairly
high stressed caps.


I have some kitchen downlights (Halers) where the driver assembly is
separate from the lamp unit. They have a 7 year warranty - I guess this
is the way to go. Imagine the electronics in a "ceiling rose) and the
lamp on a pendant. I guess that would work. However, makers have felt
pressure to make everything fit a conventional lamp envelope.

True. And the same applied to CFL. Making any energy saving type which
needs a power supply of some sort with the PS built in is invariably a
compromise. I'm rather surprised new standard lighting wiring for LEDs -
perhaps with a central power supply or two - hasn't been devised.

Already there, often with 12V for the LEDs themselves.

Same with low voltage halogens.

But this does not solve the problem. Bare LEDs (the diodes themselves)
are not happy at a fixed voltage, they need circuitry to control the
current through them. So a 12V 'LED bulb" is little different from a
240v one in the complexity of circuit required.

And even if it was possible to run LEDs from a remote power supply,
there would still be the problem of a keeping the actual diode, itself a
semiconductor, adequately cooled.

The only good engineering solution is a light fitting containing the
circuitry to work from a fixed voltags supply (which might just as well
be 240v AC as anything else) and designed to allow appropriate cooling
both for the LED and the associated electronics.


--

Roger Hayter

The Natural Philosopher wrote:

On 10/11/16 17:57, John Rumm wrote:
On 10/11/2016 10:35, whisky-dave wrote:
On Wednesday, 9 November 2016 18:02:43 UTC, John Rumm wrote:

Another problem with many LED lamps is the electronic drivers don't
like heat much either. So if your lamp base has an electrolytic cap
in it (as many do) its not going to last long at over 100 degrees.

same thing happened with PC PSUs a while back someone had one the
other day and I brought some capacitors rated at 130C 2000hrs I
think, you can get 105C and the normal range are mostly 85C. This is
why a lot of LED lamps donl;t last the 50K hours stated, if you read
the full specs some are only rates at 50K if you use them for less
than x hours at a time.

Yup indeed. I retired a GU10 spot the other day that was still working
LED wise, but was flickering because the cap in its base had lost so
much capacitance.


Of course there are some 'ankers on her that will say I'm wrong. So
before tehy embarress themselfs lijke a high wranking academci telling
me there no such thing as the factories and workplaces act.....


https://www.ledhut.co.uk/spot-lights...tt-gu10-led-sp
otlight-35w-replacement.html


Go to the specs tab on the 3.6 Watt GU10 LED Spotlight - 35W
Replacement
hover the mouse over the average life hrs the green.

For dave plowman who has trouble reading it says. "When the bulb is
on for 4 hours a day".

It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that's between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


So fas so good...

which (to me at least) does not seem to imply any maximum duty cycle or
period of on time to achieve that lifetime on that particular product.

???WTF does '4 hours a day' mean if it doesn't mean 'duty cycle'

30,000 hours is about 4 years if you leave it on all the time.

Around 1960 my late data rigged up a fluorescent tube in the garage, he
died a couple of years later, but that tube was still there working,
though it flickered badly, when the house was sold in 2004.

It probably was on for less than an hour a month.

That the duty cycle affects the actual life in years for a given
duration of switched on life is triivially true. The original claim was
that a reduced duty cycle, or rather on time, increases the actual
switched on life. This suggestion is not made in the quoted material.
Given the likely heating rate of a light fitting it is likely that the
on time would have to be reduced to less than a few minutes to afffect
lifespan. It is very unlikely that 4 hours would be better than 24
hours in terms of heat damage.





--

Roger Hayter

On 11/11/2016 11:41, Roger Hayter wrote:
Rod Speed wrote:

"Dave Plowman (News)" wrote in message
...
In article 2,
DerbyBorn wrote:


All semiconductors need to be kept relatively cool. A max of about
200C comes to mind from years ago.

Capacitors too have a shorter life with heat. And SWPS rely on fairly
high stressed caps.


I have some kitchen downlights (Halers) where the driver assembly is
separate from the lamp unit. They have a 7 year warranty - I guess this
is the way to go. Imagine the electronics in a "ceiling rose) and the
lamp on a pendant. I guess that would work. However, makers have felt
pressure to make everything fit a conventional lamp envelope.

True. And the same applied to CFL. Making any energy saving type which
needs a power supply of some sort with the PS built in is invariably a
compromise. I'm rather surprised new standard lighting wiring for LEDs -
perhaps with a central power supply or two - hasn't been devised.

Already there, often with 12V for the LEDs themselves.

Same with low voltage halogens.

But this does not solve the problem. Bare LEDs (the diodes themselves)
are not happy at a fixed voltage, they need circuitry to control the
current through them. So a 12V 'LED bulb" is little different from a
240v one in the complexity of circuit required.

And even if it was possible to run LEDs from a remote power supply,
there would still be the problem of a keeping the actual diode, itself a
semiconductor, adequately cooled.

The only good engineering solution is a light fitting containing the
circuitry to work from a fixed voltags supply (which might just as well
be 240v AC as anything else) and designed to allow appropriate cooling
both for the LED and the associated electronics.

A DC supply would be better as you wouldn't need relatively large
electrolytics in the internal LED switch mode PSU which have a limited
life or a high cost. If you're using DC, 48V would be a good choice as
it's high enough not to need too thick wires (given that LEDs use less
current than incandescents) and low enough not to be a shock hazard.

No point replacing existing wiring, but maybe one day for new buildings.

Cheers
--
Syd

On 11/11/16 11:41, Roger Hayter wrote:
The Natural Philosopher wrote:

On 10/11/16 17:57, John Rumm wrote:
On 10/11/2016 10:35, whisky-dave wrote:
On Wednesday, 9 November 2016 18:02:43 UTC, John Rumm wrote:

Another problem with many LED lamps is the electronic drivers don't
like heat much either. So if your lamp base has an electrolytic cap
in it (as many do) its not going to last long at over 100 degrees.

same thing happened with PC PSUs a while back someone had one the
other day and I brought some capacitors rated at 130C 2000hrs I
think, you can get 105C and the normal range are mostly 85C. This is
why a lot of LED lamps donl;t last the 50K hours stated, if you read
the full specs some are only rates at 50K if you use them for less
than x hours at a time.

Yup indeed. I retired a GU10 spot the other day that was still working
LED wise, but was flickering because the cap in its base had lost so
much capacitance.


Of course there are some 'ankers on her that will say I'm wrong. So
before tehy embarress themselfs lijke a high wranking academci telling
me there no such thing as the factories and workplaces act.....


https://www.ledhut.co.uk/spot-lights...tt-gu10-led-sp
otlight-35w-replacement.html


Go to the specs tab on the 3.6 Watt GU10 LED Spotlight - 35W
Replacement
hover the mouse over the average life hrs the green.

For dave plowman who has trouble reading it says. "When the bulb is
on for 4 hours a day".

It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that's between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


So fas so good...

which (to me at least) does not seem to imply any maximum duty cycle or
period of on time to achieve that lifetime on that particular product.

???WTF does '4 hours a day' mean if it doesn't mean 'duty cycle'

30,000 hours is about 4 years if you leave it on all the time.

Around 1960 my late data rigged up a fluorescent tube in the garage, he
died a couple of years later, but that tube was still there working,
though it flickered badly, when the house was sold in 2004.

It probably was on for less than an hour a month.

That the duty cycle affects the actual life in years for a given
duration of switched on life is triivially true. The original claim was
that a reduced duty cycle, or rather on time, increases the actual
switched on life. This suggestion is not made in the quoted material.
Given the likely heating rate of a light fitting it is likely that the
on time would have to be reduced to less than a few minutes to afffect
lifespan. It is very unlikely that 4 hours would be better than 24
hours in terms of heat damage.



THAT I agree with.

except that of course many heat/cool on/off cycles will affect the kit
through differential thermal expansion, and that in itself causes problems.






--
"I am inclined to tell the truth and dislike people who lie consistently.
This makes me unfit for the company of people of a Left persuasion, and
all women"

Dave Plowman (News) wrote:
In article ,
John Rumm wrote:
It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that‘s between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


Heh heh. If the average life of an LED was actually 15 years, why don't
they give them a lifetime warrenty? The number of people who could produce
a receipt after 15 years would be very very small.

which (to me at least) does not seem to imply any maximum duty cycle or
period of on time to achieve that lifetime on that particular product.
(others may differ obviously)

Any claim made to the life of a product should be taken as Trump speak
unless backed up with a decent warrenty.


Not LEDs, but have Just worked out that some of my CFLs are 35
yeares old. I shudder to think what their light output is.

The Natural Philosopher wrote:

On 11/11/16 11:41, Roger Hayter wrote:
The Natural Philosopher wrote:

On 10/11/16 17:57, John Rumm wrote:
On 10/11/2016 10:35, whisky-dave wrote:
On Wednesday, 9 November 2016 18:02:43 UTC, John Rumm wrote:

Another problem with many LED lamps is the electronic drivers don't
like heat much either. So if your lamp base has an electrolytic cap
in it (as many do) its not going to last long at over 100 degrees.

same thing happened with PC PSUs a while back someone had one the
other day and I brought some capacitors rated at 130C 2000hrs I
think, you can get 105C and the normal range are mostly 85C. This is
why a lot of LED lamps donl;t last the 50K hours stated, if you read
the full specs some are only rates at 50K if you use them for less
than x hours at a time.

Yup indeed. I retired a GU10 spot the other day that was still working
LED wise, but was flickering because the cap in its base had lost so
much capacitance.


Of course there are some 'ankers on her that will say I'm wrong. So
before tehy embarress themselfs lijke a high wranking academci telling
me there no such thing as the factories and workplaces act.....


https://www.ledhut.co.uk/spot-lights...tt-gu10-led-sp
otlight-35w-replacement.html


Go to the specs tab on the 3.6 Watt GU10 LED Spotlight - 35W
Replacement
hover the mouse over the average life hrs the green.

For dave plowman who has trouble reading it says. "When the bulb is
on for 4 hours a day".

It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that's between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


So fas so good...

which (to me at least) does not seem to imply any maximum duty cycle or
period of on time to achieve that lifetime on that particular product.

???WTF does '4 hours a day' mean if it doesn't mean 'duty cycle'

30,000 hours is about 4 years if you leave it on all the time.

Around 1960 my late data rigged up a fluorescent tube in the garage, he
died a couple of years later, but that tube was still there working,
though it flickered badly, when the house was sold in 2004.

It probably was on for less than an hour a month.

That the duty cycle affects the actual life in years for a given
duration of switched on life is triivially true. The original claim was
that a reduced duty cycle, or rather on time, increases the actual
switched on life. This suggestion is not made in the quoted material.
Given the likely heating rate of a light fitting it is likely that the
on time would have to be reduced to less than a few minutes to afffect
lifespan. It is very unlikely that 4 hours would be better than 24
hours in terms of heat damage.



THAT I agree with.

except that of course many heat/cool on/off cycles will affect the kit
through differential thermal expansion, and that in itself causes problems.

I agree that would be a factor, and already is for incandescent bulbs.
However, an on time measured in minutes would also render LED lights
pretty useless for any practical lighting use.



--

Roger Hayter

Syd Rumpo wrote:

On 11/11/2016 11:41, Roger Hayter wrote:
Rod Speed wrote:

"Dave Plowman (News)" wrote in message
...
In article 2,
DerbyBorn wrote:


All semiconductors need to be kept relatively cool. A max of about
200C comes to mind from years ago.

Capacitors too have a shorter life with heat. And SWPS rely on fairly
high stressed caps.


I have some kitchen downlights (Halers) where the driver assembly is
separate from the lamp unit. They have a 7 year warranty - I guess this
is the way to go. Imagine the electronics in a "ceiling rose) and the
lamp on a pendant. I guess that would work. However, makers have felt
pressure to make everything fit a conventional lamp envelope.

True. And the same applied to CFL. Making any energy saving type which
needs a power supply of some sort with the PS built in is invariably a
compromise. I'm rather surprised new standard lighting wiring for LEDs -
perhaps with a central power supply or two - hasn't been devised.

Already there, often with 12V for the LEDs themselves.

Same with low voltage halogens.

But this does not solve the problem. Bare LEDs (the diodes themselves)
are not happy at a fixed voltage, they need circuitry to control the
current through them. So a 12V 'LED bulb" is little different from a
240v one in the complexity of circuit required.

And even if it was possible to run LEDs from a remote power supply,
there would still be the problem of a keeping the actual diode, itself a
semiconductor, adequately cooled.

The only good engineering solution is a light fitting containing the
circuitry to work from a fixed voltags supply (which might just as well
be 240v AC as anything else) and designed to allow appropriate cooling
both for the LED and the associated electronics.

A DC supply would be better as you wouldn't need relatively large
electrolytics in the internal LED switch mode PSU which have a limited
life or a high cost. If you're using DC, 48V would be a good choice as
it's high enough not to need too thick wires (given that LEDs use less
current than incandescents) and low enough not to be a shock hazard.

No point replacing existing wiring, but maybe one day for new buildings.

Cheers

Is it reallly possible to design a DC to DC converter that is both
stable and efficient without using large value electrolytics?

--

Roger Hayter

On Friday, 11 November 2016 00:46:59 UTC, Dave Plowman (News) wrote:
In article ,
whisky-dave wrote:
For dave plowman who has trouble reading it says.
"When the bulb is on for 4 hours a day".

and this is another reason I dim LEDs it's not brightnes or to save
money on electricity costs.

So that's why you keep on having to buy new dimmers?

No thicko. I've not brought a new dimmer since I replaced my faulty one with a working one.


BTW, have you any data to prove altering the waveform into the LED PS will
actually improve the life? My guess is not.

No why would I, as it has nothing to do with it.
You just have no idea have you, it's the temperature that that's the key.
We're running a lab on that today. Where students 'roast' an LED.
The ide is they hold a solderign iron up to the LED to make it hot and to destroy it by just heat no voltage applied.
Make : Electronics Second edition by charles platt
Experiment 13 Roasting an LED Page 121.

Get this book and learn something just for once.

we brought about 60
https://www.amazon.co.uk/Make-Electr.../dp/1680450263

On Friday, 11 November 2016 00:46:59 UTC, Dave Plowman (News) wrote:
In article ,
John Rumm wrote:
It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that€˜s between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


Heh heh. If the average life of an LED was actually 15 years, why don't
they give them a lifetime warrenty? The number of people who could produce
a receipt after 15 years would be very very small.

Why ask for a receipt ?


which (to me at least) does not seem to imply any maximum duty cycle or
period of on time to achieve that lifetime on that particular product.
(others may differ obviously)

Any claim made to the life of a product should be taken as Trump speak
unless backed up with a decent warrenty.

if yuo donlt understand the specs you wonlt be able to claim anything on any warrenty.

On Friday, 11 November 2016 09:47:08 UTC, dennis@home wrote:
On 11/11/2016 00:45, Dave Plowman (News) wrote:
In article ,
John Rumm wrote:
It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that€˜s between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


Heh heh. If the average life of an LED was actually 15 years, why don't
they give them a lifetime warrenty? The number of people who could produce
a receipt after 15 years would be very very small.


If you sold 10 million of them and 0.01% claimed each year you would
need quite a few people to keep up with the claims.
It just wouldn't be economical on the margins they make.

But it doesn;t stop them doing it because they know few will attempt to claim and those that do you cabn ask have you ever had them on for more than 4 hours at a time. If they asnwer yes (which most peolpe will have had) then the warrently is invalid just as it would be if the do not remove sticker was removed.


One car manufacturer offered lifetime warranties for a while but they
stopped doing it.

On Friday, 11 November 2016 11:21:13 UTC, Dave Plowman (News) wrote:
In article . com,
dennis@home wrote:
On 11/11/2016 00:45, Dave Plowman (News) wrote:
In article ,
John Rumm wrote:
It actually says: "Manufacturers will give lifetime claims in multiples
of 1000 hours. One thousand hours is equivalent to about 6-months use
when the bulb is on for 4 hours per day. A traditional halogen spotlight
bulb is expected to last about one year (2000 hours) where as an LED
will last between 30,000 and 50,000 hours - that€˜s between 15 and 25
years! Switching to LEDs means you may never have to change your light
bulbs again.",


Heh heh. If the average life of an LED was actually 15 years, why don't
they give them a lifetime warrenty? The number of people who could produce
a receipt after 15 years would be very very small.


If you sold 10 million of them and 0.01% claimed each year you would
need quite a few people to keep up with the claims.
It just wouldn't be economical on the margins they make.

Are you then saying none ever fail within the existing warranty period?
Most things are likely (percentage wise) to fail early on. If not, last
the design life before failing. The bathtub effect.

One car manufacturer offered lifetime warranties for a while but they
stopped doing it.

How long a car lasts depends on lots of things. Like type of use and
servicing, etc. That doesn't apply to a lamp.

types of use does.
Hence the 4 hours per day, or in hot confined envioments.
Thsi is also why capactitors have temerture ratings, and resistors have rated wattages. No suprise yuo know notheing about this though.

Even computers Apple stae theres as I've ready them they have temerature and altidue limits (really a pressure limit)

Even trams have limits on what the are meant to do under certain situations..
Sure trams can go 50+ MPH but that doesnlt; mean you can take sharp corners doing those sppeds that is OUTSIDE the specs.