THE BEST COLOR LED BULB HAS THE WORST SERVICE RECORD!

Flicker issues ... I've noticed it.
LED stands for Light Emitting Diode. It's the diode part that's the problem here.
They are pulling technologies out of the electronics parts bin for these things. They only flow electrons one way, blocking flow the other. In an AC environment its on less than half the time.
If full wave bridge rectified, it still has some ripple
In audio, this would be equal to the amount of hum you hear from a cheap radio. In a light, it's flicker.
Audio proves they can eliminate it as you climb the quality food chain. Unlike high end audio, you won't pay for that supply and filtering in a light bulb, so they don't offer it.

This is completely incorrect. I think you are trying to apply a little knowledge of rectifying diodes to some assumptions of how LED lighting might work. They aren't the same; not at all.

Also, audio filtering is vastly different than switching converter filtering. Unless you have a switching audio amplifier, linear audio filters are physically large. They work at audio frequencies; below 20 kHz. Switching converters strive for much higher frequencies in the interest of reducing size; 100 kHz to 500 kHz. I have worked with switching converters in the 5 MHz region. There is no free lunch. Higher frequencies come with much tougher design tasks.

An LED is a diode in name only. Yes, it can rectify but has very low reverse voltage capability - around 5 - 20 V. That makes it useless for 120 Vac rectification where it would need in excess of 170 V reverse withstand capability; more for line surges and spikes. One might argue to put more diodes in series for better withstand capability, but they won't share voltages in reverse blocking. You'd have cascading failures. Again, LEDs are not rectifying diodes. Not in the least.

One needs to manage the LEDs' forward currents. Current, not voltage. Their brightness, and operational stress is a function of current. This is done with a switching converter to take 120 Vac sine wave voltage, step it down and create a controlled current source. This takes a switching converter. The idea is to create a smooth and regulated dc current to feed the LEDs.

This is where flicker comes from. The switching frequency is a design consideration. Filtering is required to smooth the dc current. Without it, you'd have pulsating dc. Not ac, but chopped dc. Filtering makes it into smoother dc and reduces flicker.

Cheap means less filtering and lower switching frequencies. Both conspire to produce more detectable flicker. Smooth, clean dc current through the LEDs means far less, perhaps undetectable flicker.
 
This is completely incorrect. I think you are trying to apply a little knowledge of rectifying diodes to some assumptions of how LED lighting might work. They aren't the same; not at all.

Also, audio filtering is vastly different than switching converter filtering. Unless you have a switching audio amplifier, linear audio filters are physically large. They work at audio frequencies; below 20 kHz. Switching converters strive for much higher frequencies in the interest of reducing size; 100 kHz to 500 kHz. I have worked with switching converters in the 5 MHz region. There is no free lunch. Higher frequencies come with much tougher design tasks.

An LED is a diode in name only. Yes, it can rectify but has very low reverse voltage capability - around 5 - 20 V. That makes it useless for 120 Vac rectification where it would need in excess of 170 V reverse withstand capability; more for line surges and spikes. One might argue to put more diodes in series for better withstand capability, but they won't share voltages in reverse blocking. You'd have cascading failures. Again, LEDs are not rectifying diodes. Not in the least.

One needs to manage the LEDs' forward currents. Current, not voltage. Their brightness, and operational stress is a function of current. This is done with a switching converter to take 120 Vac sine wave voltage, step it down and create a controlled current source. This takes a switching converter. The idea is to create a smooth and regulated dc current to feed the LEDs.

This is where flicker comes from. The switching frequency is a design consideration. Filtering is required to smooth the dc current. Without it, you'd have pulsating dc. Not ac, but chopped dc. Filtering makes it into smoother dc and reduces flicker.

Cheap means less filtering and lower switching frequencies. Both conspire to produce more detectable flicker. Smooth, clean dc current through the LEDs means far less, perhaps undetectable flicker.

Yes ... they are.
be it a 1N4148 small signal, 1N4001 rectifier, or any LED, the same rules apply. They must be used within specifications.
None of the above will withstand being plugged into an outlet directly. In the absolute simplest of circuits, a current limiting resistor will be required, and far from optimal. (Christmas light strings are an example of this)
Any electronics project, no matter how great or small, is really a study in power supply design.
old tube fired radios were often very close to being the current limiting resistor example.
Here, we need a bit of sophistication.
Ideally, it'll be a buck/boost converter able to supply no less than 200% of the wattage required, to prevent sag, then current regulated so as not to over drive the diode.
This product must then be filtered to cut the remaining ripple.

The economics of it negotiate this down to a buck converter with what little overhead it has over driving the cheapest LED's that won't instantly fail.
 
My house has a outside light next to the side garage door. When the house was built there was woods next door, we left that incandescent bulb on 24/7. Only time it was off is if the power was out. That bulb lasted 6 years, and was still going when I replaced it with a yellow bulb.

Not sure who it was made by, the builders put it in.

I have warm white LED lights all over my house. Never had a problem.

Builders used to use 130 volt incandescent bulbs. The light output was slightly lower than the regular 110-120 volt bulbs, but they lasted for years. They also tended to have issues removing them when they did go out due to corrosion. I have broken many of those over the years when trying to remove them from the socket after they expired.
 
Builders used to use 130 volt incandescent bulbs. The light output was slightly lower than the regular 110-120 volt bulbs, but they lasted for years. They also tended to have issues removing them when they did go out due to corrosion. I have broken many of those over the years when trying to remove them from the socket after they expired.

Dad taught me how to remove a broken, jagged bulb with a potato.
 
Slightly OT, but it is interesting to see how the LED lights on many cars appear when recorded or viewed on CCTV. They appear to strobe, showing quite clearly that the current is pulse width modulated, probably to reduce heat and extend the unit's life.
 
Some last, some don't...

I'm not REAL picky about the color as long as it's not too far afield from a 'daylight' bulb. I believe that power surges will wreck them quicker, like a lot of electronics.
 
I like the Eiko and Cree bulbs. When I was still working, light bulbs of various kinds in my museum was an ongoing job for years. Over 1,000, all in dimmer-controlled fixtures. Par-16's and 30's, 12V MR-16's. Some have to stay on 24/7 - that's 8,760 hours a year. Those on a day/night schedule still get 3,300 hours. The incandescents usually had to be changed every four-six months. The 24/7's were immediate candidates for LED's as soon as they were introduced. I swapped nearly all the MR-16's and Par 16's for LED, mostly with Eiko's, and by the time I retired, some had 5 years, or 16,425 hours, and the 24/7's 43,800 hours. Only five lamps had to be changed in the first 5 years. My successor is working hard to swap out everything else over the next few months, as he is very tired of changing bulbs. I never liked the color and life span of GE and Philips, and they didn't like our dimmer system.
 
Back
Top