None of the responders have addressed the concern I raised, life span. Larry
is correct, in order to dim the light output of LEDs, you limit the duty
cycle, not the current. Yes Bruce, resisters limit the LED current, but to
what level? The greater the current, the brighter the output, but the cost
of this is durability. The brighter the burn, the shorter the life. What is
the correct device current? Normally it is whatever the LED draws .2V more
than on voltage, which differs depending on the LED color. Anything more
than that affects device life. It is that voltage you should regulate to for
maximum life. It is not to the manufacturer's advantage to tell you that.
They are perfectly happy to sell you replacements. This rule is also valid
for Halogen lamps. Voltage regulators are prudent, but not required, just
understand the trade off in longevity.
Steve

wrote in message
...
Electricity has always been a scarce commodity for us. Lights really
are the only essential component, but I have grown accustomed to
refrigeration and the computer ( a 12VDC mini-itx unit built in with a
19" LCD monitor, also 12VDC)...and then of course there are the lights
and the various pumps.

To make the long story short I've been converting the lights over to
LEDs so I can still have cold drinks and bang on the keyboard without
worrying if I'll have enough juice to see and be seen. The Anchor/Tri-
color was the first. It was great; 12 amp hours down to about 1...it
was expensive up front but made me less of a electricity nazi.

I experimented with cheaper alternatives for the interior lighting,
but because of component cost I was having difficulty finding workable
solutions that agreed with my miserly ways...that is until I was
poking around eBay. I stumbled on:

30 Red Surface Mount LED 19 in Strip Super Bright Light

It was 12VDC. It was 24 bucks to my door. I'll give it a shot.

Four days later I make 3 little stainless clips and mount the strip
over the Nav Station, *in the packing tube it came in*; I tossed in a
sleek rocker SPST rocker switch and it is bright; it is red; it is
perfect.

Cool. What else can I do?

30 Warm White Surface Mount SMD SMT LED 19 in Strip 12V

Same thing different color. I pick the warm white over plain white
because I like the color temp...but it's the same deal. They can be
ganged sequentially so I get 4 of these buggers. Two by two, port and
starboard in the salon/galley. They are bright, warm, and use a
fraction of the electricity the fluorescent light used...why stop
now? I go crazy.

Lot of 10 x Under Cabinet G4 Light Warm White LED 12V for about 10
bucks a piece, plus a wad of cheap prewired MR16/MR11 sockets to plug
them into....

I'm sticking these things everywhere the sun don't shine...and liking
it.

In the aft cabin the fluoro fixture was crapping out. I gutted it and
screwed in 3 sockets and popped in 3 of those LEDs. On the other side
I dropped in 2- 4 Red LED Surface Mount 12 Volt Modules and re-used
the SPDT switch that was already there. The white side might be too
bright and I might take out one of the LEDs

I also tried one of the larger 120 lumen G4 LEDs. It is so bright I
went to Lowe's and picked up a 16 dollar gooseneck light fixture; took
out the halogen bulb; cut off the AC plug and wired it up for the
workbench.

The upshot is the solar panels now keep up and I don't have to fire up
the engine or start the generator just to charge batteries.

....and most importantly my wife says to me, "Thanks for the LEDs; I
can see much better."

On Sun, 28 Dec 2008 11:24:43 +0100, "Steve Lusardi"
wrote:

None of the responders have addressed the concern I raised, life span. Larry
is correct, in order to dim the light output of LEDs, you limit the duty
cycle, not the current. Yes Bruce, resisters limit the LED current, but to
what level? The greater the current, the brighter the output, but the cost
of this is durability. The brighter the burn, the shorter the life. What is
the correct device current? Normally it is whatever the LED draws .2V more
than on voltage, which differs depending on the LED color. Anything more
than that affects device life. It is that voltage you should regulate to for
maximum life. It is not to the manufacturer's advantage to tell you that.
They are perfectly happy to sell you replacements. This rule is also valid
for Halogen lamps. Voltage regulators are prudent, but not required, just
understand the trade off in longevity.
Steve

Every document and specification I have read related to LED output or
life states that current is the limiting factor and data is freely
available concerning output, current, life, etc. Plus there are
innumerable circuits available for LED circuits.

Using a pulse rate or width supply is just a simple, cheap, and easy
method of controlling the total current through the LED.

The point with the LED "tubes" is that they are designed to be used in
auto circuits (I believe) and they are built to survive in that
service. Initially I did install regulators for each light but found
that the tubes were already current limited to the point that adding
any resistance at all to the circuit caused the lights to dim
noticeably and certainly having them on with 14.1 VDC system voltage
hasn't bothered them yet.

I really don't know how long these lights will last and to a great
extent I don't care. The individual LED's can easily be changed so
other then minor inconvenience a burned out LED isn't really a
problem.
Cheers,

Bruce
(bruceinbangkokatgmaildotcom)

Bruce In Bangkok wrote in
:

I really don't know how long these lights will last

In America we have a new kind of red, yellow and green LED sold to
government bureaucrats as traffic signal lights. It differs from the 5
baht LEDs you buy in that it has a mean time between failures, once hung
over the street and turned on, of about 3 weeks. They must have put some
kind of magic timer in each of the LEDs in the big round panel set to a
different failure time. You pass a light and everything is lit the first 3
weeks. Then, one at a time, they start to fail as the magic timers wink
them out. After 2 months, half the LEDs in the array no longer light up.
After 3 months, they've already replaced it with the SAME STUPID CRAP the
first one was so it will run another 2 months......EXACTLY like the light
bulbs the LEDs replaced!

American engineering.....Planned Obsolesence....

On Sun, 28 Dec 2008 15:11:13 +0000, Larry wrote:

Bruce In Bangkok wrote in
:

I really don't know how long these lights will last

In America we have a new kind of red, yellow and green LED sold to
government bureaucrats as traffic signal lights. It differs from the 5
baht LEDs you buy in that it has a mean time between failures, once hung
over the street and turned on, of about 3 weeks. They must have put some
kind of magic timer in each of the LEDs in the big round panel set to a
different failure time. You pass a light and everything is lit the first 3
weeks. Then, one at a time, they start to fail as the magic timers wink
them out. After 2 months, half the LEDs in the array no longer light up.
After 3 months, they've already replaced it with the SAME STUPID CRAP the
first one was so it will run another 2 months......EXACTLY like the light
bulbs the LEDs replaced!

American engineering.....Planned Obsolesence....

Called "planned obsolesce" and in financially difficukt times useful
to keep Chinese electronic firms in business.

Cheers,

Bruce
(bruceinbangkokatgmaildotcom)

Bruce In Bangkok wrote in
:

Called "planned obsolesce" and in financially difficukt times useful
to keep Chinese electronic firms in business.


No, I don't think the Chinese can make an LED last only 3 weeks. The state
redneck bureaucrats would have a "Buy American" sticker on the bulletin
board at the highway department office. This crap must be Made in USA.

On Mon, 29 Dec 2008 03:15:52 +0000, Larry wrote:

Bruce In Bangkok wrote in
:

Called "planned obsolesce" and in financially difficukt times useful
to keep Chinese electronic firms in business.


No, I don't think the Chinese can make an LED last only 3 weeks. The state
redneck bureaucrats would have a "Buy American" sticker on the bulletin
board at the highway department office. This crap must be Made in USA.


Engineered in the USA for a one time engineering fee, manufactured in
China with a profit of 0.1 cent/LED. The Yellow Peril in action.

Cheers,

Bruce
(bruceinbangkokatgmaildotcom)

Bruce In Bangkok wrote in
:

Engineered in the USA for a one time engineering fee, manufactured in
China with a profit of 0.1 cent/LED. The Yellow Peril in action.



You got an Apple iPhone?......(c;]

Sorry, cellular newsgroups. I just can't help myself but to point out
iPhone's glaring deficiencies to their adoring fans.

"Steve Lusardi" wrote in
:

None of the responders have addressed the concern I raised, life span.
Larry is correct, in order to dim the light output of LEDs, you limit
the duty cycle, not the current. Yes Bruce, resisters limit the LED
current, but to what level? The greater the current, the brighter the
output, but the cost of this is durability. The brighter the burn, the
shorter the life. What is the correct device current? Normally it is
whatever the LED draws .2V more than on voltage, which differs
depending on the LED color. Anything more than that affects device
life. It is that voltage you should regulate to for maximum life. It
is not to the manufacturer's advantage to tell you that. They are
perfectly happy to sell you replacements. This rule is also valid for
Halogen lamps. Voltage regulators are prudent, but not required, just
understand the trade off in longevity.
Steve


The LED is first a DIODE. It's a very non-linear device that conducts
in one direction. You MUST, repeat MUST use a current limiting resistor
which is very easily calculated by the simple series circuit it creates
as a minimal circuit.

Now, these lighting LEDs don't need any resistor. They aren't just
diodes. It is just SO easy to add solid state devices to the same chip
the LED is made of what they do is add a constant current regulator to
the chip. To keep the regulator from making the chip hot, they use a
switching regulator, not an old analog transistor acting like a
resistor.

To find out if your LED is this kind of device, light it up on a
variable voltage power supply. Turn the power supply up until the LED
lights, much lower than the rated applied voltage. Bring the voltage up
while watching the light. At some point, the self-regulated LED will
get just so bright and then no brighter. If you move it back and forth
above the point where it stops getting brighter, a switcher regulator
will start pulsing the light way faster than you can see on and off.
Moving it rapidly sideways you can see it strobing on and off. If it
keeps getting brighter with voltage, it's a simple LED with series
resistor. If you advance the voltage from zero to 0.2V and it suddenly
gets very bright, indeed, and the current jumps up really fast with
voltage applied, it's a raw LED you must provide an external resistor to
protect.

There are all kinds of LEDs produced now, not just plain LED diodes.
The clear ones you can use a loupe and see the regulator circuitry
inside of next to the diode that lights up. Some of the better
regulator ones will run on any AC or DC voltage from 3V to 50V
continuously with life spans of several hundred thousand hours. The
ONLY thing that kills any LED is HEAT. Like any IC, heat causes
migration of the various doping in the silicon, destroying the chip.
Kept at a reasonable temperature, much higher than you expect, it'll run
nearly forever.

It all runs on magic smoke. Any solid state device will work fine
unless the magic smoke escapes. Once the magic smoke escapes, there's
no more magic to keep it performing miracles like they do.