LED Christmas Lights and Timers - On Faintly Even When 'Off'

[imagoon]

Hmm never thought of that, they ARE diodes after all, don't really need a rectifier, they just have to be rated to handle 120v (ex: C7/C9's which use 120). Also explains why they flash since they turn off when the AC is phased off or opposite polarity. So... I wonder what would happen if you hooked these up to a 120vdc power supply, bet they would work, and they would not flash, and probably appear brighter. I need to try that experiment one day. First I should probably open up one of those bulbs to see how they're setup.

A common "upgrade" for these lights is to use a capacitor and a bridge rectifier to eliminate that blinking, crammed in to a little box at the start of the strings.

 

[Red Squirrel]

A common "upgrade" for these lights is to use a capacitor and a bridge rectifier to eliminate that blinking, crammed in to a little box at the start of the strings.

Cool that's great to know it would work.

I want to build a light control switch in the future so I will probably incorporate a DC rectifier into it that can be turned on/off on a per channel basis. Basically it will just be a box with lot of plugs and relays that hooks to a computer so it can be programmed. Nothing insane like 1000's of channels like those music synced displays, maybe like 20 channels of on/off alternating lighting.

 

[imagoon]

Well here is what I know about them so far:

Typically they are bunch of LEDs in series so that the voltage drop across all the junctions is some what close to 120v with a resistor in there for current control. Some strings have a "reverse" some where. Basically if you rectify them to try and remove the blinking, only 1/2 light because they built the string expecting AC. Current per string is often really low. I have seen some 70 "light" strings only pulling just under 5 watts. It makes the really easy to control with solid state because the amperage is low. The also make RGB strings for more interesting color shows

 

[Rubycon]

LEDs can be powered by tiny induced currents.

Yes they can!

https://www.youtube.com/watch?v=OXx1qxH247k

WARNING! Keep hands and fingers AWAY from the wiener while the circuit is live. Failure to do so may result in unexpected electrical shock and touching the wiener in an unpredictable way!

😎

 

[Red Squirrel]

Yes they can!

https://www.youtube.com/watch?v=OXx1qxH247k

WARNING! Keep hands and fingers AWAY from the wiener while the circuit is live. Failure to do so may result in unexpected electrical shock and touching the wiener in an unpredictable way!

😎

The first time I saw this I totally did not expect it to work. It's not even a proper completed circuit lol. But guess enough current still makes it and there's enough potential variations throughout the wiener for there to be a voltage between both leeds of a led. Fun stuff. Guess that's why if you are on electrified ground such as a scene of a downed power line you keep your feet together.

 

[rockyct]

A common "upgrade" for these lights is to use a capacitor and a bridge rectifier to eliminate that blinking, crammed in to a little box at the start of the strings.

I wouldn't do the upgrade yourself for a couple reasons. First, if you did a bridge rectifier with a RC filter, you're essentially turning 120 V AC into 170 V DC and that at the very least will shorten the LEDs lifespan. Second, depending on how it's wired, I believe that only half of the string may be light up.

I do wish more LEDs had the rectifiers in from the factory though. The 60 Hz flicker pisses me off a lot and the colors are so brilliant when they are full wave. They've also seemed to have figured out "white" LEDs a bit better now so I hope that technology trickles down. Since a "white" LED is really just a blue LED with some yellow phosphor, it is a bit tricky to get them to a nice color. Plus, people just seem to prefer their clear Christmas lights with a bit more yellow in them.


Anyway, to answer the OP. This is normal with these timers but it's not a big issue really. The LEDs are not a resistive load so they aren't really triggering the shutoff completely. A snubber (aka a resistor) of some sorts would make the LEDs turn off. You could do this with a high kohm resistor but something like a C7 bulb would do it as well.

Like this:
http://www.amazon.com/Darice-6402-Co...s=c7+bulb+cord

Don't worry about the life span either. The sockets in the string are going to wear out far sooner than the bulbs. If the bulbs do fail, it's because they got rusted. The pins are still made of tin and those replaceable bulbs actually can defeat the seal pretty easily.

 

[edro]

A snubber (aka a resistor) of some sorts would make the LEDs turn off. You could do this with a high kohm resistor but something like a C7 bulb would do it as well.

Like this:
http://www.amazon.com/Darice-6402-Co...s=c7+bulb+cord

That's a great idea. Simple/easy/safe solution. (as long as flipping polarity at the timer doesn't work)

 

[iamwiz82]

Thank you all, I did some searching around the web and saw that night lights are an acceptable option to use in this situation but if there is no harm in leaving them dull I won't bother. It's so faint that you have to be looking for it before you notice.

 

[Rubycon]

LEDs hooked to AC have a much more pronounced flicker due to their rapid attack and decay times. Incandescent lamps also flicker but their slow attack/decay smooths it out sufficiently enough so the human eye does not perceive it. It's still there, however. A small solar cell hooked into a headphone amplifier and pointed at one will reveal a slight hum if you want proof.

Those cute little neon glow lamps (NE2, for example) still found in some pilot lamps and night lights will actually "switch poles" depending on the input frequency. This can be revealed with a video camera with variable shutter speed. Like mains driven fluorescent lighting (50/60Hz) they tend to have a more pronounced flicker (not as "sharp" as direct driven LED) but sufficient enough to be a hazard around moving parts as to give the illusion that a part is moving slow or even in some cases not at all!

A two pole relay / contactor will eliminate the partial glow. How bright is it? (pic) Generally a single pole switch turning off the neutral (which in itself is hazard!) is going to cause a plethora of issues with sensitive / high impedance loads. This is why modern plugs have one prong larger than another, for example.

 

[imagoon]

I wouldn't do the upgrade yourself for a couple reasons. First, if you did a bridge rectifier with a RC filter, you're essentially turning 120 V AC into 170 V DC and that at the very least will shorten the LEDs lifespan. Second, depending on how it's wired, I believe that only half of the string may be light up.

Yes, like I mentioned above some are wired flipped. 170V vs 120V over 35 or so LEDs in series is about a 1V increase per lamp. You can either run them hot, shed the extra voltage or use a transformer to drop to around 75VAC before you rectify. It isn't all that hard to correct.

 

[Jeff7]

LEDs hooked to AC have a much more pronounced flicker due to their rapid attack and decay times. Incandescent lamps also flicker but their slow attack/decay smooths it out sufficiently enough so the human eye does not perceive it. It's still there, however. A small solar cell hooked into a headphone amplifier and pointed at one will reveal a slight hum if you want proof.
...

That quick on/off time of LEDs plays hell with my oddly-fast eyes. I can see PWM flicker of up to around 4kHz. I didn't think that was possible, but that's what the o-scope confirms. Beyond 5kHz, then I'm finally starting to see the light as steadily-on.

LED vehicle tail lights are especially bad - some operate normally at some flickery low rate, and others only flicker when they're dimming. But it's got to be <1kHz. They "buzz" visibly, and every time I move my eyes anywhere, they look like some kind of strobing emergency light. Very distracting. :\

 

[Rubycon]

That quick on/off time of LEDs plays hell with my oddly-fast eyes. I can see PWM flicker of up to around 4kHz. I didn't think that was possible, but that's what the o-scope confirms. Beyond 5kHz, then I'm finally starting to see the light as steadily-on.

LED vehicle tail lights are especially bad - some operate normally at some flickery low rate, and others only flicker when they're dimming. But it's got to be <1kHz. They "buzz" visibly, and every time I move my eyes anywhere, they look like some kind of strobing emergency light. Very distracting. :\

It's the waveform. If you think those lights are annoying on shore power plug them into your typical cheap UPS' 'stepped approximation to sinewave' and pull the power plug!

It's pretty easy to determine the PWM frequency of a tailight. Attach a silicon photovoltaic cell to the eyepiece of a 8x monocular and point it at the offending light. Connect the output to a recording device so you can use an audio editor later to examine it. They will have more apparent flicker when dimmer (headlights on but running, when driver's foot is on brake pedal they go full bright, less flicker).