This is the test of a simple power supply for the LED filament lamp.
It uses full wave rectification and proper smoothing for flicker-free operation.
It's basically a 100 ohm resistor to limit inrush current to a full wave rectifier charging a 4.7uF 400V capacitor. The resultant DC voltage is then applied to the LED lamp via current limiting resistors. I used two quarter watt 10K resistors to get a current of 4mA, which is about a third of the normal current for this "4W" lamp, but kept the dissipation well within the limits of the resistors. (just 160mW for 250mW resistors). I could have increased the current by reducing the value of the resistors, but I prefer my electronics to run cool.
Note that I added some flexible leads to the glass LED globe and resined them into the base to avoid accidentally breaking off the existing wires coming through the glass seal.
This was really just an experiment to see how simple I could make a power supply for the lamp using standard robust components. It also suggested that the original power supply for the lamp was probably dissipating near to 1W from quite a small linear regulator chip.
Here's the data:-
Smoothed DC voltage 336v
Lamp voltage 256v
Voltage across resistors 80V
Lamp current with 2 x 10K resistors 4mA
Resistor dissipation 160mW per resistor
It uses full wave rectification and proper smoothing for flicker-free operation.
It's basically a 100 ohm resistor to limit inrush current to a full wave rectifier charging a 4.7uF 400V capacitor. The resultant DC voltage is then applied to the LED lamp via current limiting resistors. I used two quarter watt 10K resistors to get a current of 4mA, which is about a third of the normal current for this "4W" lamp, but kept the dissipation well within the limits of the resistors. (just 160mW for 250mW resistors). I could have increased the current by reducing the value of the resistors, but I prefer my electronics to run cool.
Note that I added some flexible leads to the glass LED globe and resined them into the base to avoid accidentally breaking off the existing wires coming through the glass seal.
This was really just an experiment to see how simple I could make a power supply for the lamp using standard robust components. It also suggested that the original power supply for the lamp was probably dissipating near to 1W from quite a small linear regulator chip.
Here's the data:-
Smoothed DC voltage 336v
Lamp voltage 256v
Voltage across resistors 80V
Lamp current with 2 x 10K resistors 4mA
Resistor dissipation 160mW per resistor
Just a question, I would like ro use 1 x filament per light source, ie 4 in series but in separate fittings. I would need to smash the glass carefully and disassemble the lamp. Is there an issue with heat and using the filaments in a sealed bottle with no gas. I would obviously use current limiting as you have.
Eat some garlic moron.your voice scratch my ears membranes.
on a 25 ceiling led kit i need a driver….please
Hi Clive. I've been toying with the idea of modifying existing lighting fixtures rather than the lamps. I recently picked up a handful of different LED lamps and have found that all of them seem to work just fine with a resistor (so far I've tested up to 1500 ohms using 5W wire wound ceramic bricks) in series with the lamp socket. Are there any traps I should be concerned about regarding any active circuitry like switching regulators in the lamps?
Sorry to post this on such an old video, but this one is more relevant to my question than anything more recent.
Great video.
I always find them very informative.
Keep up the good videos!
Is it possible to dim this lamp? Not with a normal dimmer, but with a led (pcm) dimmer?
Thanks for sharing 😀👍
Again very informative. Thanks Clive, I'm trying my best to get my head around all of this electronics, so thank you much for all of your advise.
I thought you were going to change out the capacitor (4.7uF X 400V) in this from the earlier video?? Thanks, tom.
I am curious about your use of the current limiting resistors, since current is constant in a series circuit, why not just a single 20k?
Would this circuit allow me to dim the bulb? Or would the capacitor negate a dimmer like a trailing edge, or triac based dimmer?
Ok Clive, I have a techtronix scope with a failed CCFL tube. Unfortunately the parts have been discontinued. I want to use cool white led filaments to replace it. I tested it with a cheap led strip and it lights up great it just has concentrated light spots. so i was hoping connecting 3 in series of the filaments would eliminate the spots. your circuit is great but what would i need to change for a 110v system?
Hello! I just came across this video and was slightly confused by something. This may be a simple question but I'm somewhat new to this and can't find the answer elsewhere. The mains voltage is 240v, but the voltage across the capacitor is significantly higher. Even generously accounting for fluctuations in mains voltage this seems high. Why does that happen? Thanks!
Dear Clive, actually I did not expect to receive an answer and am therefor very pleasantly surprised. Thank you very much for your explanation. Lots of success and I hope you will continue your very entertaining 'reverse engineering' video's. Actually infotainment is a dutch word which I think better describes you efforts to educate me and others. Thanks again,
Instead of 100 ohm at the input, why not move one of the 10K resistors there? This will further reduce inrush current, limit peak currents, and it lowers the voltage across the capacitor, extending its life.
I do not know if the total power dissipation will be higher or lower, or the same.
It may also increase the light fluctuation, as the capacitor is at a lower voltage now, the ripple will be higher if seen as a percentage.