LED rope lights

[hrc]

I've been doing some searching on the net, and I can find several vendors but nothing that describes how the LED rope lights actually work. They run off of 120VAC (LED's can do that?) but are they in series, parallel or what? Im sure someone out here knows.

 

[BrianG]

I thought this question or - one very similar - had come up before, but I can't find the thread now!

Anyway, the essential thing to remember is that LEDs are d.c. devices. As diodes they produce light in the forward direction but unlike "normal" diodes they have a low breakdown voltage in the reverse direction.
Therefore to run them on a.c., unless the voltage is very low, it requires reverse voltage protection for the LED such as a high voltage diode in series to block any reverse current.

Unless the LEDs are of the type which are "constant current" all LEDs require a method of limiting the maximum forward current to safe levels (e.g. a resistor on d.c. circuits; a.c. circuits could use a resistor or a capacitor).

To make a chain which can run on a.c. the series string would be simplest. A parallel circuit would require a low voltage (e.g. from a transformer, so not practical as part of the string) plus individual current limiting resistors or capacitorsfor each LED. Series-parallel circuit combinations could use groups of LEDs each with common current limiting. All configurations would need diode(s) to block reverse voltage, as above.

For the series method only one reverse diode is required. The string is designed so that "umpteen" LEDs whose forward voltage drops (plus the reverse protection diode) in total add up to near the nominal a.c. supply voltage. One overall current limiting resistor or capacitor would suffice for the whole string, calculated to limit the peak current through the LEDs at the peak a.c voltage (1.414 x nominal a.c voltage.)

Hope that helps!

 

[jon1]

There are a few LED makers who have not only integrated a series resistor in with the LED, but also a high voltage blocking diode, and they have done this so cheaply as to make the added cost less than a 10% adder. At the moment I can't recall who they are, but they were all together in a major distributor's catalogue, and they are a neat package, since the resistor allows for a wide operating voltage range, and the blocking diode allows for AC operation (half wave), the result being that you can simply string 24 in one direction in parallel with 24 in the other direction and get an 48 lamp array that comsumes only 5 to 20 mA (depending on desired brightness) from both halves of your 120V AC line (.6 to 2.4 watts total)

 

[OperaHouse]

It may be a "stupid engineer trick", but a reverse diode is not necessarily needed. I saw this in some cheap import where the LED was connected to the 120V line without a diode. I tested about ten different LEDs with only a 27K resistor in series with the line for a week. None of them failed! Certainly a stack of them in series would have enough reverse voltage.

 

[hrc]

LED's are a current device, correct? So is it possble to dim the entire string by just varying the current (the series resistor)?

 

[buzzp]

Yes it is.

 

[BrianG]

Operahouse: Yes, you can probably get away without a reverse blocking diode - if you are lucky. But the problem is this: although the reverse voltage breakdown of most LEDs is around 5V - which I grant you is not much different to the forward voltage, especially for a whole series string - the maximum reverse current allowed is measured in uA whereas the forward current is in mA. The series resistor is calculated for limiting the forward current...... go figure, as I believe Americans say.

Leaving out the reverse blocking diode is value-engineering taken a step too far!

 

[jon1]

Brian G is absolutely right. You don't have to be designing circuits for the military to be aware of the hazards of a sucessful "sample test" like a previous poster did for one week. His LEDs probably had a parasitic reverse bleed current protecting them, but unless the manufacturer tells you that every device will have this, you can't count on it, and your results (failures) will vary.