LED bulb question

Two different LED bulb models by the same manufacturer. Packaging says:

bulb1: 800 lumens, 2700K, 150 mA @ 120V, 8.5 watts

bulb2: 800 lumens, 2700K, 120 mA @ 120V, 10 watts

Can someone explain what’s going on here?

The two bulbs might operate at different power factors (PF).

Power factor is defined as the ratio of power (Watts) consumed to apparent power (Volt-Amperes) drawn. Resistors and old-fashioned incandescent light bulbs operate at unity power factor, because all the current they draw follows the applied voltage instantaneously; no lag, no distortion. Inductive loads such as transformers and AC motors operate at less than unity power factor because their current draw lags the applied voltage. Electronic loads, such as the AC to DC converter circuits inside LED light bulbs, operate at less than unity power factor because their current draw is distorted relative to the line voltage. But, not all AC to DC converter circuits are created equal; some distort more than others, so some operate at lower power factor.

The first bulb in your example operates at PF = 8.5 / (120 x 0.15) = 0.47

The second bulb operates at PF = 10 / (120 x 0.12) = 0.69

It would be worthwhile to test these two bulbs using an inexpensive meter such at the Kil-A-Watt, to see how well actual performance tracks their markings.

Interesting my first thought was different efficiency emitters for the different models could account for the same light output with different consumption. but the math on the wattage doesn’t add up.

The other question is one a dimmable and the other not?

I had an interesting experience recently when I installed a non-dimmable LED bulb in a fixture that was connected to a lighted switch. Turned on the switch and everything worked fine. Then I turned off the switch and periodically the bulb would flash on for a very short period. Scratching my head and then looking at the lighted switch the light in the switch would flicker corresponding to the bulb flashing. I later put a dimmable bulb in that fixture and the flashing stopped.

In retrospect I concluded that the slight voltage leak through the switch’s illumination allowed the capacitor in the non-dimmable bulb to charge to the point where current would flow through the LED. It would then drain the capacitor, the light would go off, the capacitor charge again and the cycle repeat.

So a dimmable bulb would have higher consumption at the same light output due to the added control circuitry’s power consumption. But that of course does not explain why the math doesn’t calculate out.

I was wondering about that.

Bulb2 is dimmable; bulb1 is not.

Interesting you should mention that. It’s on my list of useful tools I should look into. Do you have one of that model?

Yes, and you are welcome to borrow it. PM to arrange a meeting time.

I happened to have two Cree LED bulbs marked 2700K 800 Lumens 9.5W 120V 60 Hz 79 mA on hand. From my household 120V, both draw about 8 Watt, 70 mA, and both measure PF = 0.88 using my Kil-A-Watt.

Their markings indicate ~unity power factor: 9.5W / (120V x 79mA). Small print also says they are UL Listed and suitable for damp locations. Another marking 2213 suggests they were manufactured in Week 22 (late May/early June) of 2013.

What brand are your LED bulbs?

Walmart “Great Value”.

They were selling them for 82 cents a piece so I picked up a dozen.

Hey Richard,

I picked up a Kill-A-Watt yesterday.

I tested it with a standard* 60 watt incandescent bulb in an outlet whose voltage reads 123.6 V. The Kill-A_Watt says it’s drawing 33.9 watts / 33.9 VA (PF=1) and 0.27 A. I had not expected that.

Would you please test a standard* 60 watt incandescent bulb with your meter?

*** i.e. not “long life”**

Will do that. Right now I am on the road coming back from New Hampshire, so it may be late Saturday before I can post again.

*I just tested a 750/1500 watt dual heat resistive space heater with the Kill-A-Watt. The readings made sense. Power Factor 1.00 (as expected), and watts approximately 765 (on low) and 1443 (on high).

When I get a chance, I’m going to test a larger sample of incandescent light bulbs of various brands and power ratings.

I just tested three GE Reveal incandescent light bulbs with my Kil-A-Watt.

The 29W bulb measured 28 Watts, 0.23 Ampere, 1.0 power factor.

The 60W bulb measured 60 Watts, 0.5 Ampere, 1.0 power factor.

The 75W bulb measured 76 Watts, 0.63 Ampere, 1.0 power factor.

As expected.

I don’t have any GE Reveal here, but I tested a bunch of GE Soft White incandescents and they all measured roughly half their wattage rating (with PF=1.00).

Do you have any GE Soft White bulbs? If so, could I ask you to test them?

What about the LED bulbs that started this discussion, have you tested them? Inquiring minds want to know.

I do have to wonder if there is a problem with Ether’s Kill-a-watt in accurately measuring at the lower end of the range. I was under the impression that the difference between the standard, soft white and reveal bulbs was purely the coating on the inside of the glass and not due to a different filament. So in theory they should all match their wattage ratings within the expected tolerance of a high volume mass produced item at low price point.

How about plugging a power strip in with that heater on low and a lamp with the bulb and comparing the results with and without the bulb on?

Thanks for the excellent suggestion.

34.2W bulb only (60 watt soft white incandescent)

755W heater only

811W heater plus bulb

811-755 = 56W

Power Factor was 1.00 for all readings above.

So in theory they should all match their wattage ratings within the expected tolerance of a high volume mass produced item at low price point.

Spec accuracy is listed as 0.2%

I think I may have a defective unit.

I was thinking about the bulb’s tolerances, but the Kill-a-watt’s tolerance is an important part of the equation as well.

Is there anything in the documentation that would indicate the tolerance is different at different parts of the range? If it is just a stated .2% then I’d have to agree that you seem to have a defective unit based on Richard’s results with his. +/- 1 Watt seems to be as good as would be expected on a product like an incandescent bulb.

Have you tried measuring the bulb resistance with a DVM?

The max watts spec for the unit is 1800W.

So even if 0.2% means 0.2% of 1800, the unit is still way out of spec.

The resistance of the bulb changes substantially as the filament gets white hot.

I don’t have a DVM that uses 120 volts to measure resistance.


… and the inexpensive DVMs I have access to don’t support measuring AC current with 120V applied.

*I just called the tech support number and the friendly fellow who answered wasn’t able to shed any light on the situation.


*Richard, If you are still following this thread, could you please take a peek at the label on the back of your Kill-A-Watt and compare it to this and flag any differences for me? Note especially the red arrows.

I’m wondering if your unit is an improved version.

I bought mine at Menard’s, and they appeared to have been hanging on the rack for quite some time (packaging covered in fine dust). Also, the printing on the cardboard label inside the thermal-welded plastic packaging is copyright 2011.