RFI from Cree LED bulbs?

[California]

Curious what would cause RFI in a LED bulb. There is probably not much of a PSU but maybe just a voltage regulator which oscillates and could cause RFI I guess.

I don't understand the details but I have a LED bulb that scrambles the flat-screen TV so bad that I couldn't use it anywhere in the house. But the old CRT tv out at the ranch isn't affected by it.

I posted about this, far back in this thread.

 

[ultrarunner]

All the LED bulbs I've bought... about 50 so far are for conventional Edison Base sockets... some of the lamps date from the 1920's with most dating from the 1950's...

Can't comment on flat screen interference as we don't own any!

Even my Desktop is a CRT!

 

[Redneck in training]

Curious what would cause RFI in a LED bulb. There is probably not much of a PSU but maybe just a voltage regulator which oscillates and could cause RFI I guess. How many LEDs are actually in a household Eddison replacement type bulb anyway?

Brightness of the LED is proportional to current. If you would limit the current by a resistor the resistor would generate heat defeating the LED efficieny. The way around this is to "chop" the current by switching it ON and OFF. Then the brightness is proportional to the ratio of time ON versus time OFF. The problem is that switching generates whole spectrum of frequencies so there has to be some kind of filter preventing them to propagate back to the power line and also the components should have such dimensions not work such as an antenna emitting the radio signal into surrounding space or be shielded.

 

[k0ua]

The thing is, house wiring makes a fair antenna at least on some frequency's, and some switching power supply's can amazingly emit a lot of RF energy. I have one that wipes out the complete spectrum from below the AM broadcast band to up into VHF/UHF territory. I kept it as a novelty.

 

[MossRoad]

Why don't you show them that at the pacemaker's club meeting next month....

Just kidding. We've had a few at my employer's over the years. Plays heck with radios and such. Not so much anymore.... and barely anyone that knows what even to look for.

Several times I've been called to someone's desk on reports of a wavy computer screen. For the life of them they could not figure out why they had wavy lines all over their screen. I'd reach over and turn off the desk fan, or the desk fan in the cubicle behind their monitor and low and behold, the waves go away.

Other times we'd be sitting in a cubicle working on something and answer our walkie talkie and hear someone nearby say WHAT THE HECK? Their monitor would go nuts every time we'd key the mike. Funny thing was, most of the monitors were the same make and model, but only some of them were affected by the keying of the mike. Quality control issue at the monitor factory?

 

[California]

Hmm. I wonder if there is an opportunity here to devise a cell phone blocker with 'plausible deniability' to use in a movie theater or similar.

 

[MossRoad]

About the time you deploy a cell phone blocker in a theater, someone will have a medical emergency and won't be able to call 911. As much as I dislike cell phone usage in a theater, I'm more of a "Please shut that thing off!" kind of person.... :laughing:

 

[CreativeWorksLocal]

I got my off road lights you would not believe how it gives me lighting support on night roads.

 

[bumperm]

Brightness of the LED is proportional to current. If you would limit the current by a resistor the resistor would generate heat defeating the LED efficieny. The way around this is to "chop" the current by switching it ON and OFF. Then the brightness is proportional to the ratio of time ON versus time OFF. The problem is that switching generates whole spectrum of frequencies so there has to be some kind of filter preventing them to propagate back to the power line and also the components should have such dimensions not work such as an antenna emitting the radio signal into surrounding space or be shielded.

That's "almost" the way it works (grin). You are describing PWM or pulse width modulating. This is done with some LED flashlights to reduce brightness, though better ones regulate voltage and current. PWM is mostly done at lower frequencies - fast enough to fool the eyes but not nearly so high as to cause RFI issues with anything.

What is typically used in a LED driver circuit is a buck regulator (or if needing to increase the DC voltage, a buck-boost regulator. This is a AC to DC or a DC to DC converter. To keep size small, higher frequencies are used for the switching circuits to increase efficiency as well as use smaller components. The "AC" to drive the small transformer is a fast rise time square waves, rich in harmonics (read, lots of RFI). The circuitry takes the incoming voltage/current and if AC, rectifies it to DC, then that drives a multi-vibrator or switching circuit to give a high frequency square wave which is then stepped down through a toroidal transformer, rectified to DC, then current and voltage regulated for the lower DC voltage needed to drive the LED.

Too much info, I know.

Check eBay for prices. 36 watt bar lights are below $30 w/free shipping. I have two front and rear - cool!

 

[KubotainNH]

My son was at Ikea yesterday and called me about LED bulbs they were selling. 2 - 6.3w (40w comparable) for $4.50. The 60's are close to $5 each. The bulbs seem good so far, they light up quicker than the Lowe's ones I mentioned and are not dimmable. They list them as 10 year bulbs. At this price I may put them in all my ceiling fans.