geekorojo

Wednesday, February 7, 2007

More: power consumption of household device (CFLs and Cable Modem)

This is a follow-on from the Jan 22 - Feb 2 2007 posts: I have measured the cable modem and several more CFL bulbs and a fluorescent fixture. I'm seeing that CFLs consistently consume 150% to 200% of the power that their packaging claims.

My household "almost always on" total now runs up to:
0.12W neon night light
1W: cell phone charger
3W: motorized lamp timer
4W: weather station
4W: night light
5W: two laptop-computer charger units (while not connected)
5W: three strings of LED christmas lights (OK, it's seasonal)
8W: CFL 10-watt drawing 16W at 50% usage
10W: phone answering machine
12W: alarm system control box
15W: Linksys 802.11a/b/g cable modem/switch
25W: Motorola cable TV box (digital, no DVR, no HD)
92W total thus far.

More Measurements:

--------------------------------------------------------------------------------

Device: Linksys (Cisco) WCG200-CC Cable modem and 4-port switch, circa 2004. Has a corded transformer/supply rated 120V 25W, output rated 12VDC @ 1A.
Current: measured about 121 mA AC once powered-on initialization is complete.
Power: 14.5 W.
Usage: This typically is plugged in 24/7/365.

Having a computer online and active on the internet didn't seem to impact this device power by more than 1mA.

--------------------------------------------------------------------------------

Device: Compact Fluorescent Lamp (CFL) "floodlight" GE FLE15/2/R30SW. Rated on the package as "15 W".
Current: First second: 1000 mA inrush, almost no light for first 2 seconds.
215mA at 3 seconds. Light is dim.
178mA at 30 seconds
184mA at 2 minutes
192 mA at 3 minutes.
202mA at 5 minutes (change scale to 2000mA on meter)
205mA at 7 minutes = 24.6W
198mA at 12 minutes = 23.8W
194mA at 18 minutes. = 23.3W
Power: Power draw was very dependent on elapsed time, but seems to stabilize towards 23 W.
Usage: This is in a fixture over the kitchen sink. Can be left on overnight, or just for minutes.

Once again, worth noting how the real power draw of 23W is 150% of the claimed 15W draw, and the bold packaging mis-states both the actual printed rating and the actual measured rating.

This is also one of those old-school CFL's with an annoying property: it takes a second or two for the lamp to produce any light, and even in the first minute of use it is noticably dim compared to when it's left on a few hours. This is marketed as a replacement for the interior "R-30" 75-W flood light used in recessed ceiling fixtures.
--------------------------------------------------------------------------------

Device: CFL "twister" bulb, Lights of America 2009AN9W. Rated on the package as "9 W". Fine print on the bulb base says 140mA at 120V, which would equate to a 16.8W rating.
Current: First second: 625 mA inrush, but produces some light right away.
126mA to 133mA range over a ten-minute warm-up period.
126mA stable after warm-up. = 15.4W
Power: Power draw was very dependent on elapsed time, but seems to stabilize towards 15 W.
Usage: This is in a table lamp in the living room that is a "security light" on a timer, turned on half the day.

Once again, worth noting how the real power draw of 15W is 166% of the claimed 9W draw, and the bold packaging mis-states both the actual printed rating and the actual measured rating. This older bulb tends to go dim-to-bright but not as badly as the CFL floodlight above.

--------------------------------------------------------------------------------

Device: fluorescent single tube ballasted fixture, one tiny tube type "F14T5" measuring 22" long and about 1/2" diameter. Rating plate says 14W.
Current: First second: 500 mA inrush, produces bright light right away.
230mA after 1 minute. Second fixture was measured, draws 212 mA.
Power: Seems to be about 27 W.
Usage: Two of these are mounted under the kitchen cabinets, and I use them a lot so I can turn off the bright overhead floods.

Once again, worth noting how the real power draw of 27W is 193% of the claimed 14W draw, almost double. This was purchased at Home Depot, and I don't know what the original packaging claimed but I'd bet it was 14W. This is a fixture purpose-built to mount under cabinets and light a countertop. The bulb is an odd compact size, the smallest of the tube bulbs that I have seen.

--------------------------------------------------------------------------------

So what explains the CFL discrepency?

Is this a measurement artifact? One could wonder if the manufacturers present a lying "come on" to get consumers to switch. No doubt the bulbs still draw less current per lumen compared to incandescent bulbs. So why under-state the CFL power consumption across the board?

Like fuel economy in cars, the flaw could be in the measurement method. RMS (root mean square) is one measure that can be used to quantify AC current draw. My DMM measures RMS AC current for a proscribed range of frequency and waveform types: the easist to quantify is sinusoidal current draw and a moderate frequency like 60Hz. CFLs have more than a transformer: their circuit could contribute other components besides 60 HZ to the load, and the load profile may not be sinusoidal. Maybe they rate the current component drawn only at 60 Hz (the name plate says rating 14W @ 120VAC 60Hz) and ignore higher-frequency components for "rating purposes." (To verify this: osciloscope across the current-measuring device would give a dv/dt profile of the di/dt).

The utility meter on my outside wall is the ultimate judge of how much current I pay for. Does the utility meter measure AC current the same way the DMM (RMS) does? If not, would the utility really "give away" up to 50% of my residential electric usage, and over-tax their network by delivering more power than they are metering? I doubt it.

If the utility meter responds to these loads in a manner similar to my DMM, then I simply need to take the packaging in CFLs with a grain of salt, a 150% grain to be exact. I still will wonder what consumer-law loophole allows manufacturers to advertise a rating that is inflated an average of 50%. Because CFLs justify a higher cost by claiming a savings in electricity: that savings may or may not be all it's chalked up to be. The utility meter is the master of my dollar.

Wal Mart recently announced (NY Times Dec 2006 - Jan 2007) a major initiative to sell CFLs in its stores, at lower cost, and "push" the devices out to consumers who, to date, stubornly continue to buy incandescent bulbs. This is often a "high initial cost" issue for their customers, so Wal Mart wants to bring the ASP of CFLs way down from current levels. So I think however this rating thing pans out, it is worth getting to the bottom of the "ratings game" that seems to be going on here. If customers step up to pay for a higher-cost CFL, it would be nice to know the true savings.

Friday, February 2, 2007

More: power consumption of household devices

If you actually care about how much electricity you use, you want to know much electricity each device draws. Most civilians just look at the "rating plate" that says what the device, in theory, draws. But not all devices have a rating, and rating plates on some devices are at best an approximation, or a maximum. So, since I have the tools, I measured.

This is a follow-on from the Jan 22 post. I have measured more household devices.

My household "almost always on" total now runs up to:
0.12W neon night light
1W: cell phone charger
3W: motorized lamp timer
4W: weather station
4W: night light
5W: two laptop-computer charger units (while not connected)
5W: three strings of LED christmas lights (OK, it's seasonal)
8W: CFL 10-watt drawing 16W at 50% usage
10W: phone answering machine
12W: alarm system control box
25W: Comcast/Motorola cable box (digital, no DVR, no HD)
77W total thus far.

More Measurements:

Device: tiny orange-neon night light.
Current: measured about 1.02 mA AC.
Power: 0.12 W.
Usage: This typically is plugged in 24/7/365.

For old-timers, look up the NE-2 bulb with series resistor that lets it run directly off the wall voltage. Still a very efficient and long-lived device.

Device: Motorola Cable TV Box (Comcast) circa 2005. Model DCT2524/1612. Rating plate: 125V 35W. This is measured, of course, with nothing pluged into the accessory outlet on the back. It receives digital cable, but has no digital video recorder, and is not HDTV-capable.
Current: With the box starting up cold from power-on, I measured 173 mA AC to 192 mA AC as the device initializes. This is the draw while the "power" switch is off, and the device still runs.
After the device is powered on and running: 198 mA AC with the TV turned off, and 207 mA when the TV gets turned on and presents a load.
Power: 23.5W off, 23.8W on with TV off, 24.8W on with TVon.
Usage: This is plugged in 24/7/365, so call it 25 W.

Note: When I get a meter with more than 2A range, I'll measure a newer Motorola box with DVR and a nameplate rating of 450 watts. Ouch!

Device: Night light, 4W incandescent bulb candelabra base, GE.
Current: 34.7 mA AC, with a high inrush of 150mA for less than a second.
Power: Gee, no surprise: 4.2 W. At least this runs at the rated wattage: but look at the CFLs below.
Usage: This is plugged in 24/7/365.

Device: Compact Fluorescent "10 W" GE Helical, FLE10HLX/2/SW. Rating 120VAC 140mA.
Current: 132 mA AC, with a high inrush of 520mA for less than a second. (This appeared to be a stable reading over a half hour "warm-up." Some CFLs seem to drop current during the first few minutes.)
Power: At 122V line, this is 16.1W, not 10W claimed in bold letters on the package, but in-line with the rating that appears in fine print on the base of the bulb.
Usage: This is just used at night, sometimes left on all night.

This is interesting. How can this be "advertised" prominently on the package a 10 watt bulb rating, when the product's real consumption is 160% of the advertised rate? Even on the bulb itself, in fine print, there is a rating that says it will draw 16.8 watts (120V x .14A); yet the bold print on the package says it draws much less.

I've looked at a couple other screw-in CFLs and they have similar "deflated" rating properties, promising to save energy when in fact the savings are much less than claimed. Sounds like a letter to a consumer organization is in order.

Monday, January 22, 2007

Power consuption of household devices (U.S.)

I'm curious about the "wall wart vampires" throughout my home. These household applicances drain ever-so-little power (or so it seems), but when you look at how many of these are plugged in, you wonder... will that power-drain add up to something significant over time?

1W: cell phone charger
3W: motorized lamp timer
4W: weather station
5W: two laptop-computer charger units (while not connected)
5W: three strings of LED christmas lights (OK, it's seasonal)
10W: phone answering machine
12W: alarm system control box
40W total thus far.

This total is the tip of the iceberg, because it does not yet include the big vampires: CATV, TV, stereo. Of course, the lights, furnace, fridge, range, laundry, and outdoor lights are the big consumers.

The term "wall wart" is something I heard once a long time ago: the little plastic plug-in transformers are like warts growing out of the wall outlet. and they are vampires, because they suck juice from the wall outlet, often doing so even if they are not plugged into their companion devices.

Experimental equipment.
I have my ancient trusty Fluke 8060A digital multimeter. In years and years I can't recall that I ever measured AC current with it. Until now! I rigged up an AC plug and AC receptacle cord, and sacrificed an old Fluke lead-set to put the meter in-line with the hot lead (the lead coded black). The OEM leads provide good insulation on the plug ends. Sadly, this DMM's AC current limit is just 2 amps AC. But for wall-warts, that's enough!

Method:
Plug the device into the wall using the Fluke DMM AC current measurement scale, and adjust the scale multiplier from high to low scale to obtain the most precise three-digit measurement. This Fluke DMM is very precise, more than needed for this type of measurement. AC current is measured using RMS (root-mean square) measurement. Power (watts) consumed is computed as the current times voltage. Either the line voltage is measured and recorded or is assumed to be similar to earlier readings. (My house it was measured as 121 and 122 VAC 60Hz, assume nominal 120V)

It is important to measure the AC power actually consumed at the wall outlet connection, rather than measuring it on the consumption (DC) side. I'm curious to see if some of these devices are "inefficient." Name-plate ratings for power (wattage) and current usually tell some, but not all, of the story.

Template:

Device. Manufacturer, model, nominal name-plate rating of the wall transformer, other data.
Current: the measured current on the AC line side, 120 Volts AC nominal, 60Hz.
Power: computation: watts = current x voltage.
Usage: describes how much time in each day this device is plugged in, drawing power.

A note about rating versus actual. The "name plate" of the device specifies the "rated" voltage and current. This is usually much higher (one hopes) than the actual current the devise uses; the manufacturer's stated rating is the devices' maximum capacity and it's actual intended usage is lower than that, perhaps half or less. If you run one of these wall transformers at its actual rated output, you might not see smoke but it will be warmer than you want it.

Measurements.

Device: Nokia cell phone charger: Bought in 2006 for model 6162i cell phone. Wall transformer rated as 100-240VAC 180mA. Output is 5.7VDC 800mADC.
Current: measured without the phone plugged in, draws 5.7mA. While a quiescent, almost-charged cell phone is plugged in, the draw varies every few seconds 65mA - 85mA AC.
After unplugging the phone, the quiescent no-load current dropped inexplicably to 2.4mA AC. Clearly there is some sort of "smarts" in the charger.
Power: 0.70W quiescent (no phone).
Usage: This typically is plugged in 24/7/365 even if the phone is not charging.

Device: Wireless weather station: Davis Vantage Weather-Pro 2. Wall transformer rated as 120VAC 5VDC 200mA.
Current: measured while in use28.8mA AC line, with "lamps off." If the "display lamnps" are turned on, it goes up to 31.2mA AC.
Power: 3.54W (or 3.83W with lamps on)
Usage: This runs 24/7/365. Only reverts to battery if the power fails.

Device: VTech answering machine with wireless phone: base stations. Wall transformer rated as 120VAC, ??? mA AC.
Current: 85mA AC with no phone sitting in the charger base. When a (mostly) charged phone was placed in the charger it pulled 91mA AC but this draw depends entirely on the charging cycle. AC line assumed to be 122VAC.
Power: 10.4W (no phone in base).
Usage: This is plugged in at my house 24/7/365.

Device: mechanical plug-in timer, Intermatic "Time All" model SB811. This is the type that is motorized, not electronic.
Current: 22.7mA AC.
Power: 2.8W.
Usage: I have one plugged in 24/7/365 to control a lamp.

Device: Charger for Sony laptop computer, model VGP-AC16V8. Ratings 100-240VAC@1.5A; output 16VDC @ 4A bought around 2005. It has a little green LED that is lit as long as it is plugged in.
Current with no connection: 18 to 22mA AC.
Power: 2.2W to 2.7W.
Usage: I have one or two of these plugged in, idle, a lot of the time. OK, almost all the time: one in the famliy room, one in the bedroom.
Current when plugged into half-charged laptop, while streaming a vdeo program, fan is operating, display (12" LCD) brightness is set to 3 of 8:
approx 500mA AC.
Power: approximately 60W.
Usage: perhaps a few hours each day.

Device: LED christmas light string (3). Each string has 35 bulbs, total 105 lamps. The box says Westinghouse; the label says "Holiday Creations Ltd, item 34417." The "bulbs" are little plastic globes about 1cm diameter in red, orange, yellow, green and blue.
Current: an amazingly low 42 mA AC (for three light strings)
Power: 5.12W for three thrings (105 bulbs)

This is phenomenally low current: this is 0.048W per tiny bulb, and 0.4mA per bulb! This is such a low draw I don't feel bad leaving these lights on all night, and they make a great night light. I can remember, back in the day, ah yes, that the bias current on old LEDs was normally 10mA to 20mA. This new stuff is similar light output for a phenomenally small current draw !

Device: Alarm system: a home alarm with wireless sensors, call center monitoring, professionally installed by one of the top three nationwide alarm monitoring companies. The alarm control box is about 15 years old, and is powered by a chunky wall-mount transformer. Rating primary 120V 0.14A, secondary 14VAC 1.4A. This is not a DC converter, just a raw transformer.
Current: measured with system idle: 100mA AC. While alarm is "talking" and reporting status: 120 mA.
Power: 12W to 14.4W.
Usage: This is plugged in all the time 24/7/365

Up next for measurement, things at my place and other people's homes: Compact fluorescent lamps, night light, light bulbs, light bulbs on dimmer, Comcast cable TV box, TV set, stereo, alarm clock, computer, computer monitor (CRT).

What's this about?

I've got to get my geek on.
Bobo Rojo