Experiment Home
Home
 
Email John
Last update 12/08/2010

Experiments with Compact Fluorescent lights.
Custom Search

  since 07/28/07

  
I've been fascinated with Compact Fluorescent lights (CF) ever since they appeared on the market.  I mean, how do they get that much light out of that small bit of glass and phosphor?  How do they make 'em so cheap?  Being of an experimental nature, I decided to find out.

The first thing I decided to do was to cut open a CF tube, attach some neon electrodes and see how the tube operated as a neon tube.

The candidate device is pictured to the left.  This is the largest, brightest unit available as of this writing.  The tube, liberated from its mount is shown above.

The top arrow points to the mercury releasing device.  I though it was a blob of mercury but a closer exam shows it to be a solid.  The bottom arrow points to a thick buildup of phosphor.  Sloppy.  So thick that it impedes light production.

  
This is a close-up of the mercury release device.  I don't yet know what it is made of, though I suspect it is an amalgam.     Click to Enlarge
  Here I'm cutting off the existing electrodes with a Kitco hot wire tube cutter.  I hold the switch on the Kitco until the wire loop reaches a dull red.  Then I remove it and touch a wet finger to the glass.  It parts at the heated line.  Like magic.    Click to Enlarge
  This is what the inside of the tube looked like.  This unit had failed after an electrolytic cap blew out from overheating.  This caused the electronic ballast to malfunction and over drive the filaments.  One had burned out completely.  Both had sputtered the inside of the tube black.    Click to Enlarge
  Here is one of the original electrodes.  No wonder fluorescent lamps burn out so quickly!    Click to Enlarge
  Here's the tube with neon electrodes attached and the whole thing installed on the bombarder, ready to be bombarded.  Unfortunately the first shot of bombarder power resulted in one of the bends cracking.  Therefore the rest of the photos will show one electrode sticking out the side.    Click to Enlarge  
  In this photo we see the tube filled with about 4mm Hg of argon and illuminated at 60 ma.  It is quite curious that the phosphor glows this gorgeous magenta even though the phosphor is a dull orange under a UV color checker.  The arrow points to the thick layer of phosphor on the bottom of the tube.  The light output is greatly reduced here.    Click to Enlarge  
  In this photo I have just fired the mercury capsule contained in the Masonlite QuickSilver electrode.  The mercury is just starting to migrate through the tube.  It is evident here just how much the mercury contributes to the brightness of the tube.    Click to Enlarge  
  One of my objectives was to see if I could operate the tube at its original intensity using neon electrodes which will last essentially forever.  This requires more current than neon uses so I decided to try a fluorescent ballast.  Specifically this one.  This is an electronic ballast designed for 2 4ft tubes.  It was the only small electronic ballast I could find at Home Depot. 
 		    Click to Enlarge  
  Here's the result.  The tube is lit from the fluorescent ballast.  The Seaquest tube in the background was attached to the ballast to load the other output, though the instructions say it isn't necessary.

I used a lot of fill flash in this photo to try and capture the tubes accurately.  These tubes are BRIGHT.  Too bright to look at comfortably.  This ballast can't drive the tube as hard as the original electronics. 

As of this writing this setup has been running for about 3 weeks.  The ballast is just barely above room temperature which means it is happy.  The tubes run warm as is to be expected, but not excessively so.  Most important, there appears to be no sputtering of the electrodes even though they are running at a current density much higher than spec.

BTW, that large lamp in the background is a 2500 watt theatrical lamp.  Rated at 25 volts, 100 amps.  It lights up nicely with my stick welder!

    Click to Enlarge  
  Speaking of the original ballast, here it is. 
The red arrow points to the capacitor that blew up.  Note the general signs of overheating.  The thermal management in this light leaves a lot to be desired.  It only ran for about 3 months before the cap blew out.		    Click to Enlarge  
  My conclusion is that this experiment is a success.  CF tubes can be refitted with "permanent" cold cathode electrodes, processed as a neon tube and then be operated from a fluorescent ballast.

My next experiment will be to find a higher current ballast, perhaps for high intensity lamps, and give that a shot.