Your string of christmas lights goes out...

[NivagSwerdna]

Your string of christmas lights goes out...

The string is the type where many (say 80) incandescent bulbs are wired in series.  The plug on the end has a fuse which is intact.

How do you approach finding the broken bulb(s)?

(On the string is a bulb that has a white top, all the others are identical)

[Twoflower]

That's a very easy task: Grab your TDR equipment and just measure the distance to your broken bulb and you're done. In the end we're engineers ain't we? 

But if you lend the TDR to a friend with the same problem: Maybe you can remove a bulb in the middle and check the connectivity with a multimeter in both directions of the string to see in which direction the broken one is.

By the way usually there are usually not all 80 in series as it would exceed the low voltage security limits. So there are segments which are in parallel.

[madires]

But if you lend the TDR to a friend with the same problem: Maybe you can remove a bulb in the middle and check the connectivity with a multimeter in both directions of the string to see in which direction the broken one is.

... and do the same for the broken half, again, and again until you've found it. Search optimization

[sleemanj]

I solve the problem by throwing away the 0ld string and buying a new LED one

[Stupid Beard]

I don't. Bah humbug! 

[mikeselectricstuff]

Those volt-stick devices can be useful. You may need to swap Live/Neutral to get sensible results

[hopski]

Check the one with the white top first, that's the fuse bulb so if there's a prob that one should go first.

[CJay]

I solve the problem by throwing away the 0ld string and buying a new LED one

And when the same occurs on the LED string as happened to me last week?

[NivagSwerdna]

Checking the fuse in the plug and the fuse bulb (the white one) is an obvious first step...

I don't have a TDR

... but then... do You leave the power on?

Assuming you don't... then if you put one side of your meter on a plug terminal you need a very long fly lead to get to somewhere on the string before attempting the binary search.

I was wondering if I could inject a signal into the string?

[Gyro]

Finally switched to LED after 32 years with the same set from Woolworths (2 strings in parallel), coiled up each year around the same rolled up newspaper from 1984! I think the last time we had to change a bulb was about 10 years ago. I still think they looked nicer - we'll probably go back to them when the LED ones die. 

Edit: If the white topped fuse bulb is intact then it's sometimes as simple as going round and tapping each of the other bulbs with your fingernail, sometimes the internal 'shorting mechanism' (a bit of aluminium wire wrapped round the filament posts) doesn't break down when the filament blows. Then replace the culprit before it increases the stress on the other bulbs. The fuse bulb doesn't have this shorting machanism for obvious reasons!

[Twoflower]

Surely you might be able to use a TDR. But then you might read the manual first. Some allow to keep power running, some not. Some have a signal source build in, others not. Also you should (if you insist to go this way) chose one with the right resolution and maximum distance and such.

My first paragraph was meant ironic, even if it should be possible. And I don't know how well it will actually work as the cable and connections will cause lots of reflections even at good points.

The second paragraph is meant to be a fully separate approach: Unplug your string. Remove bulb #40 (the one in the middle). Now measure with a ohmmeter if there is a reasonable resistance between one contact of the bulb socket and the end of the string. If yes: This side should be OK, then test the other contact of the bulb socket. The part of the string which has 'infinite' resistance is the defective half. So now you can half this part of the string again and so on.

[NivagSwerdna]

My first paragraph was meant ironic, even if it should be possible. And I don't know how well it will actually work as the cable and connections will cause lots of reflections even at good points.

Really?

[Twoflower]

Yep.

[NivagSwerdna]

Yep.

It's OK.  I got it first time. 

[Twoflower]

  Well played, sir!

The spread of knowledge here is just to big to be sure if such thing is understood the right way.

[Mark]

Use a  high voltage, low current supply like something out of a plasma ball...  apply the HV in a darkened room and the blown bulb will light up. 

[Heszu]

Use a  high voltage, low current supply like something out of a plasma ball...  apply the HV in a darkened room and the blown bulb will light up.

10 + points.. This one is best in a looong time !

[CatalinaWOW]

I can't compete with the humor, but maybe some advice.  As previously noted most of these aren't a pure series string.  The ones I have dealt with have a series-parallel combo with usually 3 substrings.  If the whole chain has gone out either the fuze has gone, or you have more than one dead bulb.  Seems like that would be unusual, but I have hit it many times.  Maybe a line spike kills all the weak ones at once.  And have found only three solutions.  The obvious, easy and correct one which is to trash the whole string and buy a new one (preferably a day or two after Christmas when they are really, really cheap.)  The even more obvious and economical one, which may not play for family reasons is to just dump the whole Christmas light concept.  The least preferred answer is to go through the string checking and replacing bulbs.  There are actually worse things to do on a cold and blustery day.  At least for those of us in the northern hemisphere.

[NivagSwerdna]

Thanks for getting back on track CatalinaWOW...

My approach to date has been two fold.... visual inspection... faster than swapping bulbs and with a decent amount of magnification works reasonably well, or to take a spare (tested bulb) and swap one at a time, pretty time consuming.

I had hoped that the great minds here could have some up with something, I was thinking that some kind of signal injection would work?

We still have a few of these strings... the one that stopped working yesterday is a single strand (prob less than 80 bulbs), the one that failed the day before had two halves in parallel as you describe.

[NivagSwerdna]

Maybe... I need to set my meter to continuity/bleep and then have a fly-lead with two old bulb connectors and to short segments of the path in turn...

[N2IXK]

Those volt-stick devices can be useful. You may need to swap Live/Neutral to get sensible results

I'll second this!  Only method I've found that worked consistently. Hold the probe against each bulb in the string in order. The last one that shows the presence of voltage is the open one.

[Zero999]

Those volt-stick devices can be useful. You may need to swap Live/Neutral to get sensible results

I'll second this!  Only method I've found that worked consistently. Hold the probe against each bulb in the string in order. The last one that shows the presence of voltage is the open one.

I third this. It works well.

[PA4TIM]

TDR is not a joke, it does work.  did it a few years ago, and tried some other ways too. See: http://www.pa4tim.nl/?p=345

[Twoflower]

Nice feedback about the TDR. I have expected a more reflections of the individual sockets/bulbs making the defective bulb drown in the noise floor.

[ZeTeX]

Use a  high voltage, low current supply like something out of a plasma ball...  apply the HV in a darkened room and the blown bulb will light up.

I was surprised nobody mentioned it earlier, it is a pretty known trick.
the idea is to get the HV high enough so it will spark in the blown bulb.