LED Christmas light problem / how are they wired?

I've noticed that warning, too. Sad to say, until you pointed it out, I didn't even think about how puzzling the warning was.

Reminds me of the product that offered you a chance to win a prize, details inside the bag, no purchase necessary. I guess you could just rip the bag open in the store to get the information.

Do they visibly flicker or leave a pattern of dots on your retina if you move your head round quickly? If not then there is a rectifier and capacitor somewhere. Some cheap skates just put one string on each half phase and they flicker like hell. The circuit typically used for domestic LED lights was discussed in an adjacent thread only recently.

It is much more likely to be a bad joint or a weak point breaking in the cable that has failed in the wind and weather. Look for sharp bends or even broken wires. LEDs are intrinsically very reliable provided that you don't abuse them with over voltage or yanking their leads.

If they are series wired then remember that you are potentially dealing with live mains and lethal voltages. Visual inspection of the entire length of the failed string would be worthwhile.

A 9v battery and a 10k resistor will do as an LED tester. One way round the LED will light - no need for a multimeter.

OFF PLANET USE. (not space qualified)

Regards, Martin Brown

Multiple fuses means the string is broken down into parallel strings, one for each fuse.

Most likely all the LEDs are in series, and driven with a C, R, BR psu.

I'd use a cmos invertor IC plus 2 square inch copper plate to work as a voltage sniffer that can pick up mains through insulation. Run it along the set to get a good idea of where somethings failing.

Do not use a 9v battery and resistor to test LEDs. LEDs usually have a reverse V rating of max 5v or less, 9v reverse would slaughter them.

NT

AC---+--->|--->|--- ... -->|---+ | | +---|

I have not noticed any flickering, but I never really stared at it up close. The display looks best from a distance.

What is the name (subject line) of the recent thread you're referring to. This is a pretty busy newsgroup so it might take awhile to find.

Thanks for the help.

You don't need to be up close - just with it in your peripheral vision and turn your head quickly to look at it. Multiplexed LED displays look very funny when you do this as does anything that flickers.

About half a dozen threads further down the list. I'd check fuse(s) for the strings first before doing anything else - and remember there are potentially mains voltages involved. Series strings across mains voltage are uncommon in the UK now - most lights for outside use is now on 24 or less with a wall wart transformer.

Regards, Martin Brown

The strings my folks had (from the '50s) had a fuse in each prong (non-polarized plug; parallel lamps). Plug it in (either direction) then have the cheezy wire get pinched over a earth-potential metallic edge, and there's no drama.

Thanks for the link and the help. I'll check those fuses. I'm not too worried about the mains, because this is on a GFCI outlet. Of course I still practice safety, which is the reason for the GFCI. All outdoor stuff should be on GFCIs, except for the occasional use stuff such as a power tool, which these days are all plastic anyhow.

I guess that means that all the LEDs I've tested for function and polarity over the last 25 years - using a 12V source and typically a

and there needs to be some current limiting resistor or capacitor...

if one string opens, not only will those not be lit, the string that still works will be exposed to excessive reverse voltage since the other string is no longer conducting.

Mark

Not familiar with these lights myself, my understanding (via a recent conversation with another engineer who claims to understand these things) is that: 100-bulb strings are wired in series-parallel. Each bulb socket has some sort of internal collapsible coil the makes the circuit (keeps the circuit) if the bulb burns out. The coil somehow manages to shunt the bulb. On some other lights, this functionality is built into the bulb, not the socket... or so I am told.

Thus, a single bulb burnout would disable the whole string.

He showed me some tool you had to squeeze the trigger on up to 30 times to repair a string of lights. I confess I have no idea how it works (doesn't interest me anyway), but I could post a link to it if enough want to know, or someone else here doesn't beat me to the punch.

-mpm

Yes, please post the link.....

I have to admit that I never thought these LED strings could be as complicated as they are. In fact I thought they were each wired individually, and those old days of fighting with series wired bulbs was gone. Boy was I wrong!

Wiring one LED in a circuit is very simple, but these strings are hard to understand because the resistors and any other parts apparently hidden in the molded plastic globs at certain points along the string. (and of course no schematic).

The good news is that I found the problem on my display. It turns out there are THREE sections that light individually on this display. One of the three was dead. I began to substitute bulbs from a working section, and it ended up there were 3 dead LED bulbs and all of them were next to each other (I wonder why that is?). The funny thing is that after I replaced the first bad one, the other bulbs got a very dim glow. All I had to do was look for others that had no glow and I found the other two dead ones. (I dont understand that dim glow).

Oddly enough, I have had a few LEDs burn out on this string, and the rest of the string still worked. (Apparently this must be that "shunt" which I guess dont always work). Anyhow, the whole display has white LEDs, but I did not have any spare white ones so I replaced those initial two with yellow ones. This is the section that now burned out. I now wonder if the reason for the recent burnout was because of incorrect voltages in the string due to using those 2 yellow bulbs ???????

Although I got it working, I unplugged it. I only have colored replacement bulbs. I'll have to see if I can track down some white ones tomorrow.

Actually, I wish I could leave the colored ones. I put them in specific places on the display and like the colored ones in those places. But I guess this will likely cause more failures so I better stick to the white ones.

Actually, the thought of rewiring the whole display with each bulb having a separate resistor, and using a 6vdc wall wart seems appealing, but this seems like a huge job for little gain. Anyhow, now that I worked on this thing, I want to learn more about these light strings.

is

ew

ng

le

Link:

Not necessarily. With the right number and combination of LEDs, no LED gets enough voltage to drive excessive current. It's like those cheap flashlights with white LEDs that run off 2 AA's in series. They don't use a current limiting component because the voltage isn't high enough to drive excessive current. Of course, the better flashlights use a driver circuit - and the better xmas lights use a driver (smps) too.

Excessive reverse voltage would depend on what's in the string. LEDs typically have a reverse voltage rating of 5 volts, so if there are enough of them in the string that shouldn't be a problem. That said, it seems to me that the manufacturers should include a regular diode in each string anyway, even in junk xmas lights. Not sure how they overcome flicker in minimal component junk - maybe they don't?

Ed

is

ew

ng

le

it shoudl be fine with a few yellows in

NT

Some of them use the internal resistance of the battery, too. Substitute an alkaline cell in for the supplied zinc and *PHUT*. This is a really bad plan for line connected LEDs, too. Line variation is just too much.

Some don't. One set (ten strings) I bought a few years ago flickered so bad I threw the junk away when I took it all down - went back to incandescents.

Not sure how all do it, but anti-parallel LEDs (should be cheap on the same die) would solve both problems. A Zener (or two) across each LED will "solve" the "one light out" problem, too.