Practical LED illumination issues

[gonzo]

"D Yuniskis" <not.going.to.be@seen.com> wrote in message
news:h9v2m0$jpo$1@aioe.org...

Hi,

Motivated by the CFL thread, I'm looking for info on
fabricating "light panels" for use under kitchen
cabinets. Typically, low profile fluorescent fixtures
are used in these places. But, they often require
the addition of a molding to completely conceal their
presence. This reduces the working distance between
the countertop (which is fixed by the height of the
"base cabinets") and the underside of the "wall
cabinets" (which is fixed by the height of the ceiling!).

One amusing approach considered was to use the backlight
assemblies out of LCD monitors (i.e., remove the active
display and just use the CCFL's with the mechanical
diffuser). While this would be an amusing approach, I
don't think it practical. :< (I'm not keen on having
all that high tension wiring around *water*!)

The next approach was a panel made of (white) LEDs
possibly embedded in some white/opaque plastic to
produce a luminous surface. This would probably lose a
fair bit of light due to the diffuser.

Another approach was to just tightly pack LEDs in a
linear array (similar to a fluorescent tube in mechanical
design) and attach *that* to the cabinet underside.

But, I am unsure of how realistic such an approach would
be. The total light available would be an issue as would
the power requirements and the heat dissipation. Cost,
of course, is also an issue.

Aside from empirical testing, I'm at a loss as to how to
evaluate this approach without investing lots of time
and/or money.

Thanks!

I had the same idea, but used LEDs. I used lengths of
flat plastic conduit, with holes drilled at regular intervals
for the LEDs. Two lengths of wire run parallel inside the
conduit, each LED and resistor then soldered to these.
The conduit is hidden completely by the moulding under
the units, and it looks really nice when on. On reflection,
"straw hat" LEDs would give a better spread of light, but
it's quite nice as it is. Runs from a spare 12v PSU I had
lying around, and cost buttons.

 

[Gnack Nol]

On Thu, 01 Oct 2009 03:32:05 +0900, Michael Kennedy wrote:

"D Yuniskis" <not.going.to.be@seen.com> wrote in message
news:ha03vp$5fc$2@aioe.org...
William Sommerwerck wrote:
GE used to make "stick-up" fluorescents, which could be attached
with screws or double-sided tape.

It's called a Bright Stick.
http://www.thehardwarecity.com/getProductDetail.html?sku=5617816

Do they make an LED Bright Stick?

Great idea, but apparently not.

(sigh) That was the point of my post...


Why not use some cold cathode lamps that people use for PC modding?
They're the same as the CCFL in a scanner / LCD monitor / etc.. except
they come in a plastic tube with a small inverter that runs on 12VDC.


"http://cgi.ebay.com/12-DUAL-WHITE-COLD-CATHODE-LIGHT-KIT-MOD-CASE-BRIGHT_W0QQitemZ350255372273QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item518cd8d7f1&_trksid=p3286.c0.m14"

I found this one on ebay for about $10 including shipping. Just search
for CCFL or Cold Cathode and you'll get a lot of hits. Different colors
too.

- Mike

All Electronics which is one of my faviorite sources for odd parts and
assemblies has been selling the CCFL lamps along with the inverters for a
long time.

Here is the page for them.

http://www.allelectronics.com/make-a-store/category/320100/Lamps/Fluorescent/1.html

They also carry lots of Led stuff Too.

http://www.allelectronics.com/make-a-store/category/340/LEDs/1.html

You can find light strips in this listing that operate from 12 and 24
volts as well as a lot of other types.

they specalize in surplus and new parts of interest.

Gnack

 

[rush14]

On Sep 30, 2:22 pm, D Yuniskis <not.going.to...@seen.com> wrote:

rush14 wrote:
On Sep 30, 3:55 am, D Yuniskis <not.going.to...@seen.com> wrote:

Another approach was to just tightly pack LEDs in a
linear array (similar to a fluorescent tube in mechanical
design) and attach *that* to the cabinet underside.

But, I am unsure of how realistic such an approach would
be.  The total light available would be an issue as would
the power requirements and the heat dissipation.  Cost,
of course, is also an issue.

Aside from empirical testing, I'm at a loss as to how to
evaluate this approach without investing lots of time
and/or money.

I may have found just the thing you're looking for. Inexpensive LED
modules, 3 LEDs per module, low current, made by OPTEK and  available
from Allied Electronics.  They operate on 12VDC and come in various
colors including cool white. I use them for landscape lighting.  Here
is a web page you can get all the info about them:

http://www.alliedelec.com/Search/SearchResults.aspx?N=0&Ntk=Primary&N...

These look pretty cool!  But, I am a bit confused as to how they
are sold...

Pricing suggests that a "three lamp module" is the basic item
that you are buying (~$2.50).  Yet, the datasheet indicates they
are sold in strings of 30 modules -- quite obviously *wired*
together.  So, you're looking at a minimum purchase of ~100 lamps
(30x3).

I guess you are expected to "trim" the string to length
(but *buy* it in a 30 module configuration)!

Nope, you can buy them individually from Allied (I did that for some)
or as many in a string as you want. And yes when I ordered 5 blue
ones they were wired together but I can't say that's going to be the
case all the time. The wires are fairly short when they cut a string
but can easily be wired together. You might want to just order a few
of various colors to see if they'll work as desired for your
application. BTW, if you do order some pay attention to the shipping
method for cost purposes. Some of the shipping options are kind of
costly. Double backed tape for mounting under a cabinet would be
ideal.

 

[Clint Sharp]

In message <h9vtn6$s9f$2@aioe.org>, D Yuniskis
<not.going.to.be@seen.com> writes

Arfa Daily wrote:
"D Yuniskis" <not.going.to.be@seen.com> wrote in message
news:h9v2m0$jpo$1@aioe.org...

The next approach was a panel made of (white) LEDs
possibly embedded in some white/opaque plastic to
produce a luminous surface. This would probably lose a
fair bit of light due to the diffuser.

Another approach was to just tightly pack LEDs in a
linear array (similar to a fluorescent tube in mechanical
design) and attach *that* to the cabinet underside.

Aside from empirical testing, I'm at a loss as to how to
evaluate this approach without investing lots of time
and/or money.
LED 'tubes' for under-cabinet illumination, already exist, I
believe. Seem to think that I've seen them in Ikea. Go take a look at
their website

I'll have a look. I was aware there were options like this
available commercially. But, assume they will be way too expensive
(based on normal rates of markup)
Perhaps;

http://www.dealextreme.com/details.dx/sku.5256

There are plenty of LED strips on that website and they also have
flexible self adhesive LED 'tape' which you can cut to length (there is
a minimum length multiple though). I've uses them quite a few times now,
not the fastest if you choose the free shipping but they've never let me
down yet.
--
Clint Sharp

 

[Clint Sharp]

In message <ha04eq$64d$1@aioe.org>, D Yuniskis
<not.going.to.be@seen.com> writes

So, it is not that the power source to the XMAS lights may
have aged. But, have you verified that your current (now dim)
lamps are still seeing the same current? I assume you just used
a fixed resistor in series with each?
Unfortunately 'white' LEDs do dim over time, the phosphors used emit

less and light output goes down even when driven consistently at lower
than maximum ratings. The cheaper the LED, the faster they dim and also
the harder you drive them...
--
Clint Sharp

 

[GregS]

In article <NsjglvIemGxKFw4u@clintsmc.demon.co.uk>, Clint Sharp <clint@clintsmc.demon.co.uk> wrote:

In message <h9vtn6$s9f$2@aioe.org>, D Yuniskis
not.going.to.be@seen.com> writes
Arfa Daily wrote:
"D Yuniskis" <not.going.to.be@seen.com> wrote in message
news:h9v2m0$jpo$1@aioe.org...

The next approach was a panel made of (white) LEDs
possibly embedded in some white/opaque plastic to
produce a luminous surface. This would probably lose a
fair bit of light due to the diffuser.

Another approach was to just tightly pack LEDs in a
linear array (similar to a fluorescent tube in mechanical
design) and attach *that* to the cabinet underside.

Aside from empirical testing, I'm at a loss as to how to
evaluate this approach without investing lots of time
and/or money.
LED 'tubes' for under-cabinet illumination, already exist, I
believe. Seem to think that I've seen them in Ikea. Go take a look at
their website

I'll have a look. I was aware there were options like this
available commercially. But, assume they will be way too expensive
(based on normal rates of markup)
Perhaps;

http://www.dealextreme.com/details.dx/sku.5256

There are plenty of LED strips on that website and they also have
flexible self adhesive LED 'tape' which you can cut to length (there is
a minimum length multiple though). I've uses them quite a few times now,
not the fastest if you choose the free shipping but they've never let me
down yet.

I was just looking at

http://www.dealextreme.com/details.dx/sku.5254

The main problem I had when they shipped requirng a signature.
Next time post man left it, but I did complain to the company.
Sometimes you will get back orders, and its confusing
on the documents.

I use their 350 ma. drivers that work off AC or DC.
Getting the right LED color is the main problems with
lighting. The Cree I mentioned fills the bill for me.

greg

 

[nesesu]

On Sep 30, 10:32 am, D Yuniskis <not.going.to...@seen.com> wrote:

nesesu wrote:
On Sep 30, 12:55 am, D Yuniskis <not.going.to...@seen.com> wrote:

The next approach was a panel made of (white) LEDs
possibly embedded in some white/opaque plastic to
produce a luminous surface.  This would probably lose a
fair bit of light due to the diffuser.

Another approach was to just tightly pack LEDs in a
linear array (similar to a fluorescent tube in mechanical
design) and attach *that* to the cabinet underside.

But, I am unsure of how realistic such an approach would
be.  The total light available would be an issue as would
the power requirements and the heat dissipation.  Cost,
of course, is also an issue.

FWIW, I needed to replace a bunch of very small 'wedge base' bulbs in
a digital readout with white LEDs due to unavailability of the bulbs
and high heat dissipation with the bulbs.

This seems to be an increasingly common problem.

I found a source of 35 'delux white' T1 LEDs with diffusers and line
operated power module for $5. They should be in your stores soon,
since they are called LED Christmas light srtings.

Yes.  But, how much light do 35 lamps give off when clustered
together?  I've been playing with one-off lamps trying to gauge their
effectiveness and, without a focused reflector, they seem pretty
dismal.  Note that you use CMAS lights differently than those
for general illumination.  I.e., you look *at* XMAS lights so
even very tiny lights that give off very little light can
"look good".  OTOH, you look *at* the surface that lights used
for general illumination shine upon.

Sort of like architectural landscape lighting:  the lights
look bright but if you look at the *ground* around the lamp,
you can't really *see* anything!  :

The one curious thing that I found with them is that they have
noticably dimmed in the year that they have been operating. The
display is 'on' 100% of the time, but 90% of the time it is set very
dim and only is set bright when the room lights are on. Those LEDs
that are on most of the time in the display appear dimmer than those
that are infrequently on. I set the 'full bright' current to be less
than the LED current in the 'string'.

Just to clarify...

You cannabilized the store-bought strings for the individual
lamps, right?  And, prior to doing so, measured the current
flowing through the string (presumably the lamps were wired in
series?) to come up with the value that you then implemented
in your "incandescent replacements"?

So, it is not that the power source to the XMAS lights may
have aged.  But, have you verified that your current (now dim)
lamps are still seeing the same current?  I assume you just used
a fixed resistor in series with each?- Hide quoted text -

- Show quoted text -

I unplugged one LED from the string and used a 'scope to measure the
voltage across the LED and a clip on 'scope current probe to measure
the current.The string connection was somewhat obscure, and I did not
trace it out, but the LEDs 'appear' to be in series. However, when I
removed one it did not turn any others off. When I had removed the 24
that I needed for the digital display and then plugged the string in,
none of the remaining 11 LEDs lit.

In my application in the digital display, there is a DC power bus of
about 8V [bright] that feeds each LED through a resistor, and then an
'open collector' to common to switch the LEDs on or off. The light
sensor varies the voltage on the power bus.

Clint Sharp's comment " Unfortunately 'white' LEDs do dim over time,
the phosphors used emit
less and light output goes down even when driven consistently at
lower
than maximum ratings. The cheaper the LED, the faster they dim and
also
the harder you drive them..."
probably explains what is happening since these are likely very cheap
LEDs.

Without the diffusers, these LEDs are semi-spot and at a range of
about 40cm would give a softly defined light circle of about 20cm
diameter. One LED does not throw a huge amount of light, but 35 of
them do.

Neil S.