Tuesday, July 21, 2009

Watts up with these CFL bulbs?

The other day, I was doing a check of our power outage gear and I encountered a problem with the power draw of some compact fluorescent light bulbs.  In the photo above, you might be able to read that it's a 20 watt bulb.  Even though the bulb is only a 20 watt bulb, because of some weirdness in how CFL bulbs work, the draw of the bulb can be considerably more than 20 watts (more later).

Don't get me wrong, CFL bulbs are great.  I have a zillion CFL bulbs installed all over the house, including some nifty recessed ones in our kitchen (see above).   We have 8 recessed lights in the kitchen and they used to have 85 watt incandescent bulbs.  When I switched over to CFLs, we went from 680 watts to only around 160 watts and we get roughly the same light.  Since this light is on for 6-8 hours a day, the savings are substantial.  You just can't beat the efficiency of CFL bulbs.

But back to my issue - well sort of.  First a detour.  While I was away from home in the winter of 2006, DW had to deal with a nasty 3-day power outage.  If it weren't for her ingenuity and persistence, we would have certainly had frozen pipes.  Following this near disaster, I purchased and installed a 5,500 watt generator.  It gives us water, heat, light, and a few other essentials whenever the power goes out.

I talked to my electrician and did some research on connecting the generator and found that most transfer switches (that allow you to plug a generator into your home wiring) are limited to just 6 or 8 circuits.  The way modern homes are wired, the actual draw from each circuit is pretty small and 5,500 watts can run a lot more than 6 circuits if you're careful about how you do it. 

I decided on this GenTran 50 amp transfer switch.  This is way oversized for our generator, but it gives us lots of options.  At first, I thought I'd need to be picky about which circuits I lit up when on generator power, but after some experience, I learned that as long as you pay attention to the big stuff (toasters, coffee makers, hair dryers, microwaves, etc), everything else is just noise and runs fine.  Of course, I make sure I switch on the circuits one at a time to ease the startup load on the generator.  Also, I installed voltage meters to go with the amp meters so I can watch the load levels.

This setup worked great during the December 2008 ice storm outage.  As great as it was to have, I was hesitant to run the generator non-stop.  It's loud and it seems wasteful to have it on all the time.  For us, running it in the morning for a couple of hours, and then again in the early evening for dinner seemed to be a good balance.   Also, when the power goes out for a short time, it doesn't make sense to haul out the generator and fill it with gas just for a 2-3 hour outage (which most tend to be).  

So to help with the time when the power is out and the generator is off, I picked up this Xantrex portable power pack (basically a battery and an inverter) to give us some light and a little power.  It works great and even though it was pricey (I got it on sale for around $120), it fills an important role.  It's great for camping and other remote power needs and I actually use it a lot more than the generator.

Since the power pack above takes 35 hours to recharge with the AC charger, I also wanted a smart battery charger that could do quick charging during my limited generator runs.  This 25/10/2 amp Black & Decker unit is currently on sale at the B&D outlet in Kittery for $27 and fits the bill nicely.  It's a smart charger that automatically adjusts its output to the battery, so it can really pour in the juice in a hurry.  With the 35 28 Ah Xantrex unit, it gets the job done in 2-3 hours instead of 35 (This isn't great for battery life, but it's workable in a pinch - I use the slow charger whenever I have time).

Also, while I was at the Black & Decker store, I picked up an additional power pack (this one has a 19 Ah battery, with DC only output) for $41 on sale.  I couldn't let such a great deal go, and I have a small dc/ac inverter that I can use with it.  This setup can give me some additional runtime, or power a couple laptops or a small tv.  It also can be used to jumpstart a car or inflate tires.

Anyhow, back to my CFL problem and the power outage test-run I was doing.  I wanted to be able to really light up a room with the non-generator power backup system.  Not just with a faint candle glow, but with bright light.  The Xantrex power pack, with its 28 Ah battery (~330 watts for 1 hour), should have no problem doing that, even for several hours, as long as I use efficient CFL lighting.  So I dusted off an old 3-socket lamp from the basement and inserted 20 watt CFL bulbs (see first photo) in each of the sockets.

I expected no problems with this setup, since the power pack can sustain 480 watts of draw and the 3 bulbs added up to only 60 watts.  Unfortunately, I was wrong.  When I hit the switch I was greeted with a sustained flicker from all three bulbs.  Confused and perplexed, I did what any guy would do.  I tried it again, and this time it worked.  Then I let it rest and tried it again, and got flicker.  Finally, I unscrewed one of the bulbs and tried yet again and it lit fine.  Next, I screwed the third bulb back into its socket.  It also lit up fine and now the room was as bright as day.  For fun, I hooked up a 100 watt fan (because I could), and sure enough, it ran fine along with the lights.

Hmmm, I thought.  That doesn't make sense.  I tore everything down, let everything rest a bit and repeated my experiment.  Sure enough, same result.  What was going on I wondered? Why the intermittent results with the 3-bulb setup?

I set out to track this thing down with some help from google.   Within a few minutes, I learned about something called "Power Factor" and the issues that the low power factors of many CFL bulbs can cause.  Power factor has to do with how a CFL bulbs sips  juice from its 120 volt AC input.  After reading a bunch of stuff, including this thread on an alternative energy site, I thought I had it figured out.  

But on reflection, it didn't quite make sense.  The inverter I was using to generate the AC can provide almost 10 times the needs of the three bulbs.  Even taking the power factor issue into account, along with other inefficiencies, the math said this should work.

Then I found an even more technical post about CFL bulbs and power factor. This one gave me the clue I needed.   The post is very technical, but it contains a key piece of info - How much instantaneous current does a 20 watt CFL draw?  The answer sure surprised me, and it explained my problem.  The scary looking waveform labeled "Tek Run" that's about mid-way down this page shows that a 20 watt CFL can draw as much as 200 watts of instantaneous current.  Multiply that times 3 and I'm right at the limit of my system, and my guess is that this is the issue that I had.  I knew that inductive loads like big motors can have a huge startup draw, but I never would have guessed that a CFL bulb could draw this much peak current, even if only for an instant.  

Apparently, the CFL draws a lot of juice for part of the time, but then nothing during the rest of the time.  My guess is that this peak current is even higher at startup, especially when I look at the size of the capacitor (black cylinder)  in the photo above.  It may be that the inverter in my power pack isn't very robust and although it had enough juice, it just couldn't hold it together well enough to get all three bulbs going on the first try.  

Anyhow, it seems that when using smaller inverters and multiple CFLs, things may work better if they're not all turned on at once.  Incidentally, once I got the bulbs on, I left them on for over three hours before the inverter cut out due to a low battery.  That was an hour or two less than theoretical max, but considering the draw of the inverter itself, as well as other inefficiencies, and the fact that the battery is a couple of years old, it seemed about right.


  1. Egad! Wouldn't it have been easier to light candles?

  2. Candles may be easier, less costly, and more environmentally friendly - but they'd be much less fun :)

    Well, that is unless we could harvest beeswax from a backyard beehive to make into candles - like a certain Nashua Telegraph columnist for example.


  3. Do You notice that You often get a noticeable "snap!"-ing noise from Your light switch when You turn on a CFL? I suspect the high inrush current with CFLs -perhaps a lot higher than with incandescent bulbs- will cause premature contact wear on electrical switches. But it may be a few years down the road before people start discovering their wall switches are failing and need replacement.

  4. John,

    Thanks for the comment. I'm no expert on light-switch construction, but my understanding is that most switches are rated for 15 or 20 amps. Even if a 20 watt bulb sees an inrush current of 200 watts, you'd need more than 10 of them on a single switch to get near 20 amps.

    OTOH, I'm not sure there's anything else in the typical household that would have the same load characteristics as 10 or more 20-watt CFLs on a switched circuit. I guess we could be in unchartered waters.


  5. This is an outstanding report! I sell CFL's and I had never seen the inside of the base nor did I realize that they spiked to light. I guess that makes sense or else they would take a much longer time to excite the gas and produce enough light to reach their expected brightness. Instead of running several smaller watt CFL's you could try one big 85w or 105w that could brighten a whole floor of the house in times of emergency.

    Here is a quick video we created for high watt spirals in case people haven't seen one in person before: High watt CFL spirals

    If you like the video please support our business Planet Bulb

  6. apart from this post, one thing i´ve noticed about what this lamps do to the wire connection boxes...

    if using this lamps, be sure to tighten the connections on the boxes.

    In my house, i´ve installed about 2 of this, and a few weeks later i had no light in there.
    turns out that the arc created by the capacitor creates a kind of repulsive energy in the connections... and they just got loose.


  7. I'm glad I found this post. I'm an electrician in training, and I do lots of side work while I'm going to school. I was working at a friends house installing some new light fixtures (2 fixtues with 3 bulbs total) and upon completion we screwed in some cfl's. The switch was a dimmer type but ( and knowing full well that dimmers and cfls are a no no) he was rushing to see them work. We fired up the breaker and POP!!! The dimmer fried. We Shut off the breaker and scatched our heads. I installed another dimmer and repeated. POP!!! WTF I said. I finally installed a single pole switch and it worked fine. Several hours later it hit me. In rush or high start up current must have fried them. Seaching the internet I came upon your post which just proves my theory. Thanks buddy, now I know...

  8. My son just did a science project on CFL bulbs, he wanted to compare the cost savings a home would experience by switching to CFL bulbs. We tested the bulbs on an watt meter (not an amp meter that then multiplys to figure watts) and we found that a Sylvania 19watt CFL bulb was drawing 89watts, and a BGE home CFL 14 was drawing 44watts. We left them on for four hours and no change occured, since its drawing 89 watts its going to cost more than an incandescent 60watt bulb (that actually draws 57watts). The idea that the bulb stops drawing current is incorrect, we then shut downn all of our breakers except one, we stopped our home meter and then plugged in the CFL bulb for 3o mins, then we did the same with the 60watt incandescent, the incandescent used less electricty by 30%. So when looking at billed watts the new bulbs are false advertising, they do not save money as they claim, they lose brightness over time, and they are more hazerdous to the enviroment when one breaks. Getting to the real why is the question no one can answer.

  9. Hi Richard,

    Thanks for commenting. I'm not sure what to say about the results that you saw. I can say definitely that I and many many others have tested the power consumption of CFLs vs incandescent bulbs and have not seen results like yours. I used a kill-a-watt meter in my own testing and saw power consumption in line with the manufacturer specs.

    I suspect there was an error in your test setup or a malfunction in the bulbs you were using.


Note: Only a member of this blog may post a comment.