Power-line LEDs, part 54,926

[carbonzit]

OK, so we've been having a few discussions here lately about powering LEDs directly from the power line (sans transformer or proper power supply). At one point I proposed a way to do this, and was immediately shot down by people saying "That won't work!". I must say, whenever I hear that, it just makes me want to go out and try it even worse. Must be the rebel in me, I guess.

Anyhow, I just finished cobbling together the following:

**broken link removed**

It works. I witnessed it with my own two eyes. LED lights up, brightly but not too brightly. Resistor doesn't even get warm.

Notice the lack of a bridge rectifier, bleeder resistor, protection diode, etc., etc.

So tell me again why this won't work? 'Cause I'm telling you, it does ... the LED doesn't seem anywhere near being stressed.

Now, I'm not suggesting anyone copy this design for a production model. If I were going to actually use this, I think I'd throw a 1N4001 or so in series with the LED to give it a little more protection.

Oh, yeah, the capacitor had a pretty good charge on it (~60V) when I pulled the plug, but what the hey--so long as nobody's fingers come into contact with it, so what?

Minimal parts count "R" us ...

 

[Mike_2545]

You are almost correct, by using Xc=1/2PiFC you can calculate the final mA supplied to the led. I jotted down a circuit some time ago for an opto-isolator, but it will work for just a single led.

 

[carbonzit]

You are almost correct, by using Xc=1/2PiFC you can calculate the final mA supplied to the led. I jotted down a circuit some time ago for an opto-isolator, but it will work for just a single led.

"Almost correct"? In what way am I not correct? By the way, I used that formula (the formula for capacitive reactance) to determine the resistor size.

 

[Mike_2545]

Sorry, I just came up with a different set of values based on 12mA...some leds dont need to be driven too hard

 

[carbonzit]

Well, my ASS-umption was 20 mA @ 2V or so, having no idea what type of LED it was (nor caring--it came from my parts box, where most stuff is salvaged from stuff left on the curb by my neighbors).

Close enough for government work.

 

[MikeMl]

It will work until the first transient (lightning, motor, inductive load) comes down the AC line. LEDs are very sensitive to excessive reverse voltage. A reverse polarity transient with a fast dv/dt will sail through the capacitor like it wasn't there, and the 1K resistor is not enough to limit the current to keep from zapping the LED.

 

[carbonzit]

It will work until the first transient (lightning, motor, inductive load) comes down the AC line. LEDs are very sensitive to excessive reverse voltage. A reverse polarity transient with a fast dv/dt will sail through the capacitor like it wasn't there, and the 1K resistor is not enough to limit the current to keep from zapping the LED.

Which is why I'd add a 1N4001 if I cared about this circuit.

 

[MikeMl]

Which is why I'd add a 1N4001 if I cared about this circuit.

The 1n4001 would go across the LED in inverse-parallel, not in series. Most Leds have a reverse breakdown voltage of only 5V or so...