I have some LED lights that are flickering at High frequency, and I believe it's because they're essentially running off of rectified but otherwise unconditioned AC power. I got the idea to try capacitors to smooth it out, but I feel like that's potentially dangerous. It seems to me that logically, adding the capacitor would results in overall greater power usage, unless I may be lower the voltage by using additional diodes or a resistor. What do y'all think about this? I know that in the current setup it's 1.4A according to my meter, so my first thought to test it safely would be to use some high wattage power resistors to make sure that the current is limited and see what happens. I have several 100 ohm, 100 watt resistors, and I figured it would be pretty safe to use one of those for a test run just to see how they respond. Is there a safer and/or smarter way to do this? How can I estimate what would be a good size capacitor to try this? Also I have tested the output going into the LED strip, it's 120VDC according to my multimeter, and by the time it gets to the end of the run it's dropped to 106VDC. If my theory and understanding are correct, I could add the power resistor and that wouldn't reduce the overall brightness of the lights. But then if I add the capacitor back in, since it's getting fed pumped DC, that should cause the brightness to go back up shouldn't it? It would increase the average current, just not the maximum current, correct? At least in theory? Or am I just totally off my rocker about the way electrical theory works? Would it be better to use some diodes to reduce the voltage, then add a capacitor and see what happens? Actually I just had the thought, since the DC output is just rectified, it would increase the total current consumed, but it would not cause current spikes, right? So if I started out with a super tiny capacitor, like 100nf and started testing with just a tiny capacitor like that and no resistors or diodes, that would probably be a safe way to start wouldn't it? Or would even a capacitor like that have the potential for huge effects on current?

 

Smoothing caps are an ordinary feature of decent power supplies but are often omitted to save money on cheaper ones. Putting a sufficiently rated cap in across the output is not at all dangerous but you should include a bleeder (discharge) resistor of high value to bleed the charge from the cap after power is removed, for protection against surprise shocks.

 

Adding a capacitor would smooth the voltage, increase the average voltage, and potentially improve brightness. Begin with a small capacitor (e.g., 100nF to 1μF) to observe its effect. Gradually increase capacitance and monitor voltage and current.

 

Adding a capacitor would smooth the voltage, increase the average voltage, and potentially improve brightness. Begin with a small capacitor (e.g., 100nF to 1μF) to observe its effect. Gradually increase capacitance and monitor voltage and current.

I couldn't find my box of loose capacitors, so I stole a pair of 63V 100uf from other boards and put them in series on a much smaller test strip I have. As I understand it, putting capacitors in series gives you the addition of the voltage but since farads is a measure of electrical charge, you have to halve the farad rating, giving me effectively a 126V 50uf cap (way bigger than what you said to try, but it's what I had).

Current consumption on the AC line at Max power with no filtering was approximately 0.25A. Adding the series caps made the average current consumption go up by a factor of 6 to about 1.5A. it also had a very interesting effect. In what I would say is approximately the 10 to 20% range (I have a triac dimmer switch connected), the flickering was much worse. It went from very fast flickering to much slower flickering. I tried adding a 10R 100W power resistor to the capacitor in series, this cut the total current at maximum to about 0.8A (appx 3x increase from base), and I was not readily able to tell a difference in total brightness, but I was not able to test the two side by side. Next I moved the resistor to the 120VAC line from the triac to the lights, this resulted in current usage of about 0.75A, and no noticeable differences otherwise. All of the filtered cases seem to have the same effect visually. Reduces the flickering at High power, but at low and maybe even around medium the flickering is worse. I was not able to capture how much worse on camera even with a 960 FPS slow-mo. For some reason to the camera with and without filtering caps at low power looks almost identical. I can't tell them apart, despite the fact that it's an obvious difference in person.

I also tried putting the capacitors directly on the DC output and then using a resistor to reduce the LED power. At this time I was measuring DC current. The DC current at the LEDs with no filtering was about 140mA. Adding the capacitor brought it up to almost 300mA. Adding the resistor after the capacitor brought it down to about 260mA. I added additional 10R resistors in series all the way up to a total of 40R (again, between the capacitors and the LEDs) and it caused only minor additional drops in current. I think the next thing is to try a different dimmer switch to see how it behaves. This one seems to have some unexpected behavior. But the final answer I'm guessing is going to be a different switch, huge capacitors, and either power resistors, diodes, or combination, to reduce the total output current...or an entirely different light source lol.

 

Here in the UK we have 12V LED light that essentially replace 12 mini spot halogen bulbs. Typically the non-dimmable bulbs will run on either AC or DC, but the majority of the 'transformers' designed for halogen bulbs produce AC. Dimmable LED bulbs almost only run on AC only. If you run either of these types of bulbs from the halogen style transformers they often flicker, presumably because the electronic transformers become unstable when the current drawn is below the designed minimum. In my experience the dimmable LEDs exhibit worse flicker than the non-dimmable types. Often one transformer will run several lamps in parallel and leaving one halogen in a circuit of 5 or 10 lamps will eliminate flicker whilst still giving significant power saving over halogens.

Ideally you should replace the transformer with a unit sized for the actual consumption of the circuit. However, this assumes the transformer and cables are easily accessible. Adding capacitance to an AC low voltage circuit is not going to help and will inevitably be a fire hazard.

 

Here in the UK we have 12V LED light that essentially replace 12 mini spot halogen bulbs. Typically the non-dimmable bulbs will run on either AC or DC, but the majority of the 'transformers' designed for halogen bulbs produce AC. Dimmable LED bulbs almost only run on AC only. If you run either of these types of bulbs from the halogen style transformers they often flicker, presumably because the electronic transformers become unstable when the current drawn is below the designed minimum. In my experience the dimmable LEDs exhibit worse flicker than the non-dimmable types. Often one transformer will run several lamps in parallel and leaving one halogen in a circuit of 5 or 10 lamps will eliminate flicker whilst still giving significant power saving over halogens.

Ideally you should replace the transformer with a unit sized for the actual consumption of the circuit. However, this assumes the transformer and cables are easily accessible. Adding capacitance to an AC low voltage circuit is not going to help and will inevitably be a fire hazard.

Interesting information, unfortunately none of it is applicable to my situation. But thanks anyway!