The Dirty Little Secret. Pixels are Power Hogs.

The Dirty Little Secret.  Pixels are Power Hogs.

Pixel trees.  Good gosh they’re amazing to watch.  They make a show ‘pop’ and get people talking.

But…

I built three pixel trees for my Christmas display.  I’ll admit it was a bit odd when walking around them during the day and hearing the fans on the low voltage power supplies.  If the fans were spinning that meant something was using electricity.  As I remember it was about 30 degrees Fahrenheit  that day.  I checked the enclosure holding the power supplies and electronics for the pixel tree and it was warm in there.  How strange since there were no lights illuminated on the tree.

After the season I setup one of the pixel trees in my lab and added a gizmo to measure the power being consumed.

Pixel Power

My test pixel tree had 16 pixel ribbons, each with 170 pixels.  Each ribbon required 5VDC (using WS2812B LEDs).  To control everything I used a standard 16 port E1.31 board.  Low voltage power was delivered by two Mean Well SE-350-5 power supplies.

PixelTree at idleHere is where things got interesting.  I plugged everything in but told all the pixels to stay off.  Look carefully at my test meter in the image.  At 124.0VAC, the tree was drawing 32 watts of power.  No big deal, right?  I figured the power supplies weren’t 100% efficient converting 124VAC down to 5VDC.  In fact, Mean Well says the supplies I’m using are only 78% efficient (http://www.meanwell.com/search/se-350/SE-350-spec.pdf). There’s probably also a power factor issue since power supplies aren’t a pure resistive load.

Pixel Closeup of RGB LED and support electronicsWhat else was drawing power even when nothing was turned on?  Look carefully at a pixel ribbon, you’ll see electronic components soldered around many of the pixel LEDs and those components consume a bit of power even though nothing is lit.  You can hold a ribbon and sometimes feel the warmth of all the electronics cooking.  Maybe the pixel tree drawing only 32 watts sitting idle isn’t so bad.

I told the pixel tree to turn white.  That meant all of the (170 x 16 =) 2720 individual RGB pixels were told to turn on at full intensity.  Look VERY closely at a pixel and you’ll see it’s made of three individual (and very small) LEDs (one red, one green and one blue.)  From an electrical consuming viewpoint, my white tree was equivalent to (2720 x 3 =) 8160 individual LEDs… though the human eye only sees 2720 points of light.

PixelTree current draw full on361 watts.  That’s a lot of watts.  To put it in perspective, an 800 lumen LED based light bulb (equivalent to an old fashioned 60 watt incandescent light bulb) consumes about 10 watts of power.  My pixel tree when full on was using the same as 36 full sized light bulbs scattered around my house.

Now when it comes to paying for the electricity, 361 watts isn’t a big thing.  Your local electric utility sells a kilowatt hour of electricity for about 12 cents.  In English that means running a 1000 watt hair dryer for one hour costs 12 cents.  Running my pixel tree at full-on costs about (361 x .12 / 1000 =) 4.3 cents/hour.  It’s not much of your typical monthly electrical bill but multiple trees can start adding up and don’t forget that 32 watt drain during the off-hours when the tree is doing nothing.

I’ll admit the 361 watt number bothered me.  LED use is suppose to be good for the planet   I could probably buy more efficient power supplies and maybe better pixel ribbons but it’s amazing how much current all this fancy technology (aka: bloody cutting edge) can consume.

Old Fashioned LED Power

I then plugged my tried-and-true workhorse called the Light-O-Rama CTB16PC controller into my fancy power measuring gizmo.  It drew so little power nothing registered.  I plugged eight controllers in and they consumed 12 watts total so I figured each controller used 1.5 watts of power when sitting idle.  With one controller I plugged an old fashioned 100 watt light bulb into each of the 16 channels.  With all the lights turned off, the controller still used only 1.5 watts of power.

I dug up a few strings of 100 count LED mini-lights.  Each string averaged 4 watts of power consumtion.  Hmmmmmmmmmmmmmm.  I did a little math on how many LEDs I could drive with 361 watts using my LOR controller.  Ends up it was (361-1.5/4=) 89 mini-light strings or 8900 individual LEDs.  Since the LED strings are wired to work at 120VAC, there’s no inefficient power supplies needed.  When all the lights are off the only power consumed is that 1.5 watts needed by the Light-O-Rama controller.

Pixels of Points?

What do people respond to the most in Christmas displays?  I’ve been asking that question for decades.  The consistent answer seems to be points of lights.  People can never see too many Christmas lights.  With any luck they might remember the lights were dancing to the music in a show but what initially grabs their attention is all those points of lights.

Where am I going with all this?  Don’t become too enamored with the over-the-top pixel trees.  They’re expensive to build (the pixels, controller and power supplies get expensive in a hurry), not to mention challenging to hook up, sequence and maintain in a hostile environment of rain, snow and ice.  For considerably less money and the same amount of electricity you can have three times more points of lights by staying away from the bloody cutting edge pixel technology.

Pixel Balance

What am I going to do?  One pixel tree is perfect and I’ll put the controller on a mechanical timer so it’s drawing zero watts during the off-hours of the display.

 

 

 

 

 

Related posts