LED light bulb meaurements

[The Electrician]

I was in the local warehouse store (Costco) and noticed that LED bulbs were on sale: http://www.feit.com/led-lamps/performance/Performance_LED/Performance_LED

I bought a couple and made some measurements.  First I just measured the current drawn by each; the first image shows the current drawn by the 40 watt equivalent globe:



The second image is of a 65 watt equivalent floodlight style lamp current:

[The Electrician]

Here is the result of analysis of the 40 watt equivalent bulb with a power analysis application.  The orange trace is line voltage, blue is bulb current and red is instantaneous power.  The true power, reactive power, power factor, etc., can be seen at the right of each display.:



And, here is the analysis of the 65 watt equivalent bulb:



Here is an analysis of a compact fluorescent for comparison:



Here is a 60 watt incandescent bulb:

[TerraHertz]

Typical isn't it. The watermellons push to have incandescent  bulbs banned 'because they are so evil' (somewhat inefficient at making light), and instead we're told to use overcomplicated, unreliable, poisonous (mercury plus phosphors in the case of CFLs), high energy-cost-of-manufacture CFLs and LEDs.

Which also as you've demonstrated, use primitive electronics with very poor power factors, ie they draw all their power form the AC mains voltage peaks, thus messing up the mains sine wave and generating harmonics. And as more and more such loads are added to the grid, the stability of the entire system is degraded.

[ConKbot]

At least the 65W equivalent one has a smoothed current input instead of square like the 40W, or pure shit waveform like the CFL,  at least *some* attention has been paid to the current waveform.

[SeanB]

The ay they get the current waveform smooth is that they omit the input capacitor in the circuit, using only a small low value ( sub 100n) capacitor. Thus the light has a pronounced flicker on it.

[skipjackrc4]

Very interesting, thank you for posting this.

In the 65W equivalent LED bulb, the scope lists the phase angle as 36 degrees.  The current and voltage do not visibly seem to be that far apart.  Am I missing something?  I'm not terribly familiar with power systems.

[ConKbot]

Very interesting, thank you for posting this.

In the 65W equivalent LED bulb, the scope lists the phase angle as 36 degrees.  The current and voltage do not visibly seem to be that far apart.  Am I missing something?  I'm not terribly familiar with power systems.

The scope is probably counting the start of the cycles as the rising edge from 10% to 90%,

I.e. on this picture



The rising current is at 0 divisions, where your trigger happened. The rising edge of the voltage sine wave is 1/2 of a division before that, so the difference comes out to be 36 degrees. In reality, I wouldnt even know what to call that power factor.  1 with a bunch of harmonics and other crap?

[jahonen]

That angle based definition applies only to linear loads, i.e. consisting only of pure R, L and C. Anything non-linear must be calculated via definition of the power factor, PF = real power/apparent power. Real power is average of product of instantaneous voltage and current, and apparent power is RMS(u)*RMS(i). Another way to say this is that only fundamental frequency current carries real power, but harmonics are form of reactive power.

So don't waste your time trying to figure out the "phase shift" with non-linear loads.

Regards,
Janne

[The Electrician]

Very interesting, thank you for posting this.

In the 65W equivalent LED bulb, the scope lists the phase angle as 36 degrees.  The current and voltage do not visibly seem to be that far apart.  Am I missing something?  I'm not terribly familiar with power systems.

The angle is the arccosine of the power factor, and is not necessarily the apparent angle between the current and voltage.

Have a look at the heading "Distortion power factor" here:

http://en.wikipedia.org/wiki/Distortion_power_factor

[skipjackrc4]

Duh, that makes perfect sense.  Thanks guys.