Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

LED ligth(ning) on 240V no trafo

Status
Not open for further replies.
Hi folks,

conceived an idea for a LED light fitting. Plan is 19LEDs in series and R56 and a capacitor in parallel to one rectifier, this would be 60V. This arrangement 4x in series gets 240V and I have 76 LEDs lined up, maybe put them into a fancy pattern, mirror ball or so.

So I would not need transformers.

For the attachement: not completed yet, the LM317 would not do the trick it seems but I found LT3014/LT3009/LT3008. The 3014 would drive 20mA on up to 60 or so V. Could not find LT3014/LT3009/LT3008 in Multisim 10.1, and not much going on reg spice models for them in the www it seems

LEDs would be 5mm white 20mA:
**broken link removed**
 

Attachments

  • LED lamp 1.pdf
    25.5 KB · Views: 242
Don't you have LED Christmas tree lights in Australia??
They don't use a transformer and I haven't seen a capacitor in them either.
Mine appear to flicker at 60Hz.
 
YouTube - 240 Volts LED pilot lamp -- project 4

Above YouTube link is of a LED lamp for 240 Volts I have constructed many off with great success.
As long you keep the led current below 20 mA the leds should last very long.
There may be issues with the LED's fading faster when exposed to higher currents, especially when series Capacitors are used for voltage droppers.
 
Looking at your schematic, I think you make it overcomplicated.
For long time reliabilty forget about the LMxxx and just set your series R or C to the value required to drop the voltage difference between the mains and the led string.

Take 3 to 3.3 Volts for the white leds in series therefore 228 to 250 Volts is required for the 76 led string.

Experiment here carefully and use a higher value series resistor and mA meter to check that the current remains below 20 mA.

Personally I rather drop a little more voltage in a series R than relying on 5 to 10 Volts. in case of voltage fluctuations.
 
Don't you have LED Christmas tree lights in Australia??
They don't use a transformer and I haven't seen a capacitor in them either.
Mine appear to flicker at 60Hz.

should check in my local hardware store (ebay ,-) . Suppose switching power supply. Or just a rectifier, so its not exactly DC and you get a 50/60Hz flicker?

edit: Holy Moses, 48 AUD for a 300 LED one from HK, cheaper than the sum of the parts.

But I want to create a light fitting for indoor, as efficient as possible.
 
Last edited:
10 Watt, don't know where you get that wattage from ?
 
acc to your formula I get for 76 LEDs, 3.1V each, for Rs 240 Ohm, 18.33mA through the diodes.

The resistors anticipate 80mW.

But the simulation comes up with 0.8A, the LEDs are not going to like that?
 

Attachments

  • 25 LEDs.pdf
    23.1 KB · Views: 291
  • 76 LEDs.pdf
    28.4 KB · Views: 210
76 leds with a voltage of 3.1 V is 235 V, leaving only 5V drop if you have a 240 V supply.

That's not going to work in the real world. The main problem is that 240 V ac is a sinewave with an RMS voltage of 240 V (ish) so the peak voltage is 240 * sqrt(2) = 339 V. That gives you a peak of 100 V across resistors designed for 5 V drop.

There is also the problem of the variation of the 240 V and the 3.1 won't be exact.

I would suggest making the resistors about 2500 Ω each. Also, if you put one resistor between the rectifier and the LEDs, the capacitor will reduce the ripple current a lot more and you won't get any visible flicker.

There should not be an earth connection. I don't know if you need it for the simulation but it shouldn't be there in the actual circuit.

On your earlier circuit with three voltage regulators, that would not have worked at all because variations between the current taken by the three parts would have resulted different voltages across each part.
 
Diver300 is correct.

Design the led string for a value of about 200 Volts and on average 3.0 Volts per LED which yields 66 leds in series.

230 Volts case. The resistor required is than U=I*R -- 30/0.02 = 1500 ohms.
Power rating I²R -- 0.02²*1500= 0.6 Watts.

To alowe for a safety margin ( Australia is 240-250 Volts) I would take a 2k2 or 2k7 ohm resistor of at least 2 Watts, or use 2 resistors in series of 1200 ohms 1 Watt.

The peak value from the mains is 325 Volts although I calculate usually from the RMS value. Start with a higher value series Resistor and measure the led current.
Most leds will take an impulse current of 30 mA as long there is also a period that the led chip can cool down when the current is say 5 mA. (ON - OFF ratio)
 
I find it always wise with mains operated led lamps to keep the led current to below 20 mA. and aim for 10 - 15 mA.

On my YouTube video the leds run at 11 mA. This is more than adequate for a night light.

If you want real bright light it is cheaper to use CFL's and leds are way to expensive at this time.
They have failed at this stage to become streetlight bulb replacements because of the huge costs and minimal power savings compared to SOX or SON lamps.
 
All the traffic lights and pedestrian lights in my city have been changed to LEDs.
The city saves money because of the low power consumption and light bulbs no longer need to be replaced.
 
All the traffic lights and pedestrian lights in my city have been changed to LEDs.
The city saves money because of the low power consumption and light bulbs no longer need to be replaced.


Correct, same here in NZ.
Although we have sigle leds and led strings failing on some traffic light already, but they are a lot brighter and faster in response.

But for streetlighting a high intensity bundel of light is required which needs lots of leds to be driven at full power. hence reasonable high power usuage.
 
I am driving then see a bright red LED traffic light in front of me so I stop. But the light is many km away. It is much brighter than an old incandescent traffic light.
 
The LED traffic lights are much better and brighter, defenitely and can be seen more far away than any standard traffic light.

even with some leds out.
 
wound it down to 17mA. The voltages are a bit low now, 117V won't really drive 76 LEDs...

CFL would be cheaper I know, but want to create a funny light fitting. Maybe half a mirrorball with LEDs in the gaps so the mirrors reflect the light as well and I have half a globe with light radiating everywhere. As opposed to downlights.
 

Attachments

  • graetz leds.pdf
    30 KB · Views: 332
What are the U1, U2 and U3 rectangular boxes in series with the leds ?

You should get between 240 and 339 Volts on the 100 Hz output from the bridge rectifier, depending upon the size of the capacitor.
 
ok, just realised, properties been set as Vf 1.66V, I 5mA by default in Multisim. Have to change them to 3.1V versus 12mA, simulate again.
From the graph attached I picked 3.2 V, 12.5mA. That gives 256 Ohm. Replaced the LEDs in the simulation with 75 x 256 Ohm and a 256 ohm in series. Legal? Then modified the other Rs so I end up with 3.11V and 12mA. Just relize, but it is RMS? But I am on the right track, just have to modify it for 2.26 V RMV across the diode?
 

Attachments

  • led 2.pdf
    92 KB · Views: 162
  • graetz.pdf
    20.7 KB · Views: 172
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top