Existing set of Christmas tree lights with multiple push in bulbs has
stopped working and replacing the fuse bulb doesn't fix it. Rather than
pull out, test and refix a hundred odd bulbs I am considering replacing
the entire set with some of the newer LED based ones.

But, does anyone know whether these are wired in series or parallel - or
more importantly if one LED fails will the whole set stop working?

Andrew

On 26/11/09 09:23, Andrew May wrote:

if one LED fails will the whole set stop working?

I suspect the LEDs failing need not be on your radar.

"Andy Burns" wrote in message
o.uk...
On 26/11/09 09:23, Andrew May wrote:

if one LED fails will the whole set stop working?

I suspect the LEDs failing need not be on your radar.


Not guaranteed. I have some of those garden lights that set into decking.
Each has half a dozen LEDs in it and, in two years, I have already had to
replace three of the lights because LEDs were failing. Admitted they were
only Homebase ones but I would have expected a longer life than that.
--
Tinkerer

In article ,
Andrew May wrote:
Existing set of Christmas tree lights with multiple push in bulbs has
stopped working and replacing the fuse bulb doesn't fix it. Rather than
pull out, test and refix a hundred odd bulbs I am considering replacing
the entire set with some of the newer LED based ones.

But, does anyone know whether these are wired in series or parallel - or
more importantly if one LED fails will the whole set stop working?

No - you might lose more than one but not all of them. They're usually
wired in a combination of series and parallel circuits.

--
*Why do the two "sanction"s (noun and verb) mean opposites?*

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

"Tinkerer" wrote in message
...

"Andy Burns" wrote in message
o.uk...
On 26/11/09 09:23, Andrew May wrote:

if one LED fails will the whole set stop working?

I suspect the LEDs failing need not be on your radar.


Not guaranteed. I have some of those garden lights that set into decking.
Each has half a dozen LEDs in it and, in two years, I have already had to
replace three of the lights because LEDs were failing. Admitted they
were only Homebase ones but I would have expected a longer life than that.
--
Tinkerer

............or could it be the power supply thingy?

In article ,
Andy Burns wrote:
On 26/11/09 09:23, Andrew May wrote:

if one LED fails will the whole set stop working?

I suspect the LEDs failing need not be on your radar.

LEDs only have a near infinite life when under-run. A bit like tungsten.
;-)

--
*One of us is thinking about sex... OK, it's me.

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

Andrew May
wibbled on Thursday 26 November 2009 09:23

Existing set of Christmas tree lights with multiple push in bulbs has
stopped working and replacing the fuse bulb doesn't fix it. Rather than
pull out, test and refix a hundred odd bulbs I am considering replacing
the entire set with some of the newer LED based ones.

But, does anyone know whether these are wired in series or parallel - or
more importantly if one LED fails will the whole set stop working?

Andrew

Do a binary chop - at most, assuming it's one failed bulb, you'd need to
remove about 7 bulbs and test.

1) Pick a bulb half way along the test section (initially teh test section
is the whole string)

2) Remove bulb

3) Test continuity between each bulb socket contact and L and N at the plug

4) If one contact is open circuit, determine which section that is (pull
another bulb out if it's not obvious), and defined that as the new test
section, marking your current position as the end of section (wrap of
sticky tape). Goto 1)

4a) If both contacts are open circuit, you'll need to defined two test
sections and goto 1.

5) Both contacts have continuity - throw the bulb away and put a good one
in.

OTOH if you really (I know I would) arelooking for an excuse to buy an new
set, delete this post and I won't tell :-)

--
Tim Watts

This space intentionally left blank...

Dave Plowman (News) wrote:
In article ,
Andrew May wrote:
Existing set of Christmas tree lights with multiple push in bulbs has
stopped working and replacing the fuse bulb doesn't fix it. Rather than
pull out, test and refix a hundred odd bulbs I am considering replacing
the entire set with some of the newer LED based ones.

But, does anyone know whether these are wired in series or parallel - or
more importantly if one LED fails will the whole set stop working?

No - you might lose more than one but not all of them. They're usually
wired in a combination of series and parallel circuits.

They seem to come in multiples of 40 so it seems likely that loosing one
will take out a string of 40 which is a substantial proportion.

What worries me is that although the mean time to failure for a single
LED may be 10,000 hours I have no idea what the expected standard
deviation from that figure is - and that is what will dictate the
expected life of a whole string.

Andrew

On 26 Nov, 10:23, Andrew May wrote:
Existing set of Christmas tree lights with multiple push in bulbs has
stopped working and replacing the fuse bulb doesn't fix it. Rather than
pull out, test and refix a hundred odd bulbs I am considering replacing
the entire set with some of the newer LED based ones.

But, does anyone know whether these are wired in series or parallel - or
more importantly if one LED fails will the whole set stop working?

Andrew

I have a Homebase 40-LED set in which 3 near neighbours are out. It
may be because I crunched one between wall and ladder last year, but
there's no visible damaged. The rest of the chain works OK, so I guess
it is a series/parallel arrangement. My plan is to buy a reduced-to-
clear set in the sales for spares, unless someone knows another
source? I assume they're all the same voltage (no internal resistors)
regardless of chain length? Of course, the package says "non-
replaceable bulbs", but, hey, this is uk.d-i-y.

Tip: When you take them down, wind them onto a cable drum to avoid
tangling.

Chris

Andrew May wrote:

What worries me is that although the mean time to failure for a single
LED may be 10,000 hours I have no idea what the expected standard
deviation from that figure is - and that is what will dictate the
expected life of a whole string.

I don't know where you got a MTBF of 10,000 hours from. High efficieny
LEDS operating at 25C have a typical MTBF(90% CI) of 5,000,000 hours.
That gives a typical failure rate of 0.037%/1000 hours.

If you run the lights for six hours per day, you can expect to lose
0.08% of the bulbs per year. That's one bulb every thirty years (ish).

Given that they are usually wired in parallel, that's going to be over a
thosand years before they've all gone or if you're only happy to have
50% of them working, about 500 years.

"Andrew May" wrote in message
...

They seem to come in multiples of 40 so it seems likely that loosing one
will take out a string of 40 which is a substantial proportion.

I doubt they are in strings of 40.
That would require about 60V to operate them.
The PSUs tend to be about 12V-24V so strings of about 8-16 is more likely.

dennis@home wrote:


"Andrew May" wrote in message
...

They seem to come in multiples of 40 so it seems likely that loosing
one will take out a string of 40 which is a substantial proportion.

I doubt they are in strings of 40.
That would require about 60V to operate them.
The PSUs tend to be about 12V-24V so strings of about 8-16 is more likely.




But that would imply a lot more cables than are actually present. In a
string of 160 anything between 10 and 20 pairs by that reckoning.
Unless, the current is so low that they are using very thin multicore cable.

I'm tempted to buy a set just to find out.

Andrew

"Steve Firth" wrote in message
.. .
Andrew May wrote:

What worries me is that although the mean time to failure for a single
LED may be 10,000 hours I have no idea what the expected standard
deviation from that figure is - and that is what will dictate the
expected life of a whole string.

I don't know where you got a MTBF of 10,000 hours from. High efficieny
LEDS operating at 25C have a typical MTBF(90% CI) of 5,000,000 hours.
That gives a typical failure rate of 0.037%/1000 hours.

I'm not sure how they work the MTBF out, some I've seen quote like that
are tested on a 50% duty cycle and as the temperature rises the life is
significantly reduced.
A friend of mine brought some of those strings of LEDs from china 24 or 36
per string
and he put them around the house for lighting, but he's had about 20% of
them failing and
the power supply 12V burnt out too. If one LED is only slightly out of spec
then if that goes
open circuit (in parallel) then this will mean that all others will get a
slight increase in voltage
due to the drop in current, but obviously this depends on the quality of the
PSU and it's regulation.


If you run the lights for six hours per day, you can expect to lose
0.08% of the bulbs per year. That's one bulb every thirty years (ish).

Given that they are usually wired in parallel, that's going to be over a
thosand years before they've all gone or if you're only happy to have
50% of them working, about 500 years.

Sounds good in theory and on paper.
I'd like to know the details of the ones used as indicators on London buses
as they should, be reliable due to H&S

Steve Firth wrote:
Andrew May wrote:

I don't know where you got a MTBF of 10,000 hours from. High efficieny
LEDS operating at 25C have a typical MTBF(90% CI) of 5,000,000 hours.
That gives a typical failure rate of 0.037%/1000 hours.

It's a figure that I was given by the gentleman in John Lewis when I
enquired if he knew how they were wired.

If you run the lights for six hours per day, you can expect to lose
0.08% of the bulbs per year. That's one bulb every thirty years (ish).

Given that they are usually wired in parallel, that's going to be over a
thosand years before they've all gone or if you're only happy to have
50% of them working, about 500 years.

If they are wired in parallel doesn't that mean that each lamp has to
have a current limiting resistor whereas if they are in series it is
only one resistor per string?

Andrew

"Andrew May" wrote in message
...


I doubt they are in strings of 40.
That would require about 60V to operate them.
The PSUs tend to be about 12V-24V so strings of about 8-16 is more
likely.




But that would imply a lot more cables than are actually present. In a
string of 160 anything between 10 and 20 pairs by that reckoning. Unless,
the current is so low that they are using very thin multicore cable.

No, three will do it if they are all on at the same time.
Two for +/- one for each string in series.

On 26 Nov, 14:49, "dennis@home" wrote:
"Andrew May" wrote in message

...

I doubt they are in strings of 40.
That would require about 60V to operate them.
The PSUs tend to be about 12V-24V so strings of about 8-16 is more
likely.

But that would imply a lot more cables than are actually present. In a
string of 160 anything between 10 and 20 pairs by that reckoning. Unless,
the current is so low that they are using very thin multicore cable.

No, three will do it if they are all on at the same time.
Two for +/- one for each string in series.

Yup, tend to be series /paralell with 3 wires on a 24V wall wart.

MTBF is calculated with a roulette wheel, big brand white is now
typically rated to 50K hours to 70% lumens output, off brand white
might be 10% of the price and may or may not be worth the gamble
depending on application.

ForeverBright, U.S., came up with first LED Christmas lights, 35 is
series on 120V mains, have couple of sets , circa 2000/1 they set the
now standard LED shapes of berry, pine cone and stave.

Aldi had some nice cone shapes in colours and ice white LED last week.

John Lewis have an excellent set of colour morphing LED lights , if
you can find stock.

B&M Homestores, if you have one local, have a dazzling array , almost
exclusively LED.

B&Q have nice set of colour changing acrylic jewels, have had a set
from last year on 24/7 here with zero failures.

Commercial dispalys are almost exclusively LED nowadays,Glasgow George
Sq tree , 2 500m, 2.5Km, of white LED, power draw 400W.

Life span for big brand really shouldn`t be issue nowadays.

Cheers
Adam

Adam wrote:

John Lewis have an excellent set of colour morphing LED lights , if
you can find stock.

On the web site, if not in stores
http://www.johnlewis.com/230487349/Product.aspx

Apparently they are suitable for use both indoors and ... errm ... only
indoors

whisky-dave wrote:

I'm not sure how they work the MTBF out,

This might help:

http://www.itl.nist.gov/div898/handb...ion2/apr25.htm

I've got 'flu at the moment and all I can see is dancing equations.

In article ,
"Dave Plowman (News)" writes:
In article ,
Andy Burns wrote:
On 26/11/09 09:23, Andrew May wrote:

if one LED fails will the whole set stop working?

I suspect the LEDs failing need not be on your radar.

LEDs only have a near infinite life when under-run. A bit like tungsten.

Also depends on how well they're made.
There are lots of poor quality ones going very cheaply,
and guess the sort of thing they end up in...

--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]

In article ,
(Steve Firth) writes:
Andrew May wrote:

What worries me is that although the mean time to failure for a single
LED may be 10,000 hours I have no idea what the expected standard
deviation from that figure is - and that is what will dictate the
expected life of a whole string.

I don't know where you got a MTBF of 10,000 hours from. High efficieny
LEDS operating at 25C have a typical MTBF(90% CI) of 5,000,000 hours.
That gives a typical failure rate of 0.037%/1000 hours.

The LEDs used in christmas decorations are cheap chinese ones
manufactured for the purpose, or rejects from batch tests.
Life expentancy can be incredibly short, because they only
need to work for about 45 days. (They only have to last one
season - no one complains if they die the following year,
they won't even remember which shop they were bought from,
and if they do, it likely won't still be there anyway.)

These LEDs are a problem for real equipment manufacturers.
There's an enormous glut of them, at less than 100th of
the price of proper spec LEDS, and there's much opportunity
for distributors to pass them off as such and make big bucks.
Many respectable companies have been caught out in this way,
and ended up manufacturing commercial products which don't
last even 6 months before the LEDs are dead.

--
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]

Andrew Gabriel
wibbled on Friday 27 November 2009 12:14

The LEDs used in christmas decorations are cheap chinese ones
manufactured for the purpose, or rejects from batch tests.
Life expentancy can be incredibly short, because they only
need to work for about 45 days. (They only have to last one
season - no one complains if they die the following year,
they won't even remember which shop they were bought from,
and if they do, it likely won't still be there anyway.)

I bought some LED clip lights to hang on my daughter's bed, for fun, being
12V and not hot running. All from B&Q. Several 5mm white LEDS dies withing
2 months. Got a replacement set under guarantee. Same thing. Gave up and
vowed to make my own next time...


Are there any Xmas lights with *decent* LEDs in? I fancy a set, but I'll buy
filament ones unless I can be sure I'm going to get 10 years service out
of the LEDS...

--
Tim Watts

This space intentionally left blank...

In message , Andy
Burns writes
On 26/11/09 09:23, Andrew May wrote:

if one LED fails will the whole set stop working?

I suspect the LEDs failing need not be on your radar.

I wouldn't put money on that, I have a set of mains LED lights that have
several dead bits. The lights carry on working though.
--
Clint Sharp