DC CFL lights, volts, tricks

[nothing to lose]

Late last year a few of us took apart the standard CFL 120Vac screw in bulbs and connected the tubes to 12Vdc ballasts from other devices. Scanners, battery CFL lights, battery powered flouresent droplights, etc...

 My 12Vdc droplight ballast (8watt tube) worked pretty well lighting a standard 120Vac 26watt CFL tube using 24Vdc untill I reversed polarity and fried a mofset or power transitor or something.

Well tonight I was playing again with CFL and DC.

GE soft white 100 ( 00262 package number for 3 pack) 26watt CFL supposed to be equal to 100 watt,  1700 lumens work. However, basically same "Light Of America" CFL does not work so far.

This bulb lights up well on 60Vdc and appeared to put out some good light! NO MODS, I just screwed it into a normal table lamp and wired the plug to the 60Vdc battery bank, flip switch and it turned on fine. This was a low amp bank (about 1.3amp) and I had already been playing with a 100watt incadesent bulb so the bank was not full charged.

 I don't have any power use tests or figures, but it worked if others want to try it. Also I don't have anyway to compare light output from DC and AC bulbs other than use them both and look at them.

At 35VDC it does not light well. Somewhere around 34-40Vdc or so I have to play with it a bit to light it. This is a bulb that resembles a corkscrew, and at low volts the bottom turns of the tube light a little. If I wrap my hand around the bulb it lights up most the time. Sometimes I grab the bulb and it tries to light but not quite, touch my finger inside the tube turns and it lights. Once lit it has stayed lighted down to 30VDC although you can see lower volts is making less light of course. I never ran the batties low enough that the light actaully went out yet.

I thought this was very interesting myself and that others may also want to try various things with this light! My battery packs are dead now and I am charging them from a high volt low amp genny, nice winds tonight here. $5 motor, $40 in magnets, getting 0.25-0.5amps into 100V battery bank, 25-50watts kinda disapointing. Not very good at 48V and it stalls out bad at 24V, but at least it runs good at 60Vdc and above.

My thought on this was if the voltage were bumped up from 48Vdc to 60Vdc or higher to start the light then it may work well at 48Vdc. For a $2-$3 bulb it may make a nice 48Vdc CFL light. Perhaps something simple like charging a capacitor in a lamp fixture to give a voltage boost when first turned on just to get it lit?

Anyway just thought I'd pass on that the GE CFL is working on the 60Vdc fine sofar with no mods if anyone wants to play with it. I don't have any CFLs less than 26watts, a 15watt may work better or on lower volts or may not?

[TomW]

NTL;

DonC figured this out awhile ago when we were messing with boosters. Almost any standard ac cfl can eat dc as long as it is above 60 volts or so like you used. Not sure if we posted about it but we did talk it up on IRC at the time. Problem is not many use high voltage DC.

Anyway just wanted to pass that on.

Cheers.

TomW

[richhagen]

NTL, since you last stopped by, I've got a 100 amp hour 48V bank set up, and have mounted my remaining panels.  It is too bad that it takes 60VDC to startup, as 48V is fairly common, and would be of use to quite a few folks, myself included.  Rich

[nothing to lose]

Thanks TomW

I may have missed it or just forgot if I read it before.

Here I thought maybe I was on to a new find with the GE CFL, and the LOA CFL doesn't do anything at the same volts that light the GE. Everything the same, I just remove the GE and install the LOA then put back in the GE. At least we can say the GE works better or at lower volts anyway since the LOA does not light at all and the GE does.

[RP]

Neat work!

A few thoughts:

Some standard (cylindrical) flouresent lamps require a metal shield or sometimes a ground wire run down the length of the tube to help them start.  You mentioned it started easier with a hand on the tube.  You might try putting a grounded wire (22-24 gauge should be fine) in, on or around the tube to see if that helps it strike at lower voltages.

As for measuring light output:  One thing that comes to mind is using either a CDS (Cadmium Sulfide Sensor)  or a small solar cell you could do some qualitative measurements.  CDS sensors have the little squiggly line that acts a variable resistor and you find them in nightlights, or anything that turns on after dark.  You also find them on the shelf at Radioshack.  :-)  In either case I'd wire a 1uf or so capacitor across the leads to filter out the high frequency ripple you might otherwise see from the DC to DC convertor in the lamp.

As to measuring:  The light energy from a point source of light falls off as the square of distance.  

Set your lamp at one end of a table with a measuring tape, yard stick, meter stick, etc. stretched out across the table.  

1. Put your sensor at say 36 inches away and pointed directly at the lamp.  This is your "reference distance"
   
2. Measure the output of your "sensor" with 120VAC applied to the lamp.
   
3. Power the lamp from 60vdc and determine the distance required to get the same reading.
   
   

The light output of the lamp will be related to the new distance by the following formula.

Relative output = (test distance / reference distance)^2

For instance if 60vdc requires you to move the sensor in to 18inches to get an equivalent reading as 36inches with line voltage then you're lamp is now putting out (18/36)^2 or 0.25 as much light as normal.

In reality this is not a point source of light but as long you use fairly long distances (10-20 times the width of the bulb) it shouldn't matter too much.  If you want to be fancy and if your sensor has the range you could put a diffuser around the bulb (like a piece of paper and then cover that with aluminum foil with a pencil size hole in it to get more of a point source.

Oh, turn off the other lights in the room of course.  ;-)

[nothing to lose]

Glad you got it set up.

 Maybe try a CFL on your 48V bank and see what happens. Mine was weak, still is. Right now shows 59V no load but when I turn on the CFL it drops fast to 35V so my little bank is dead! It's only about 1.3 amps anyway. I just tried my CFL again and it had trouble lighting, I had to touch the bulb then it lit, meter showed the the bank droped to the 35V right away, not good for 60V bank! I am geussing a good strong 48V bank might light them up decent, don't know how bright they will burn though. Did not seem to harm mine so nothing to lose I geuss. Also on a side note, the desk lamp I am using has about a 9watt  C9 type incadescent bulb for a night light and it barely glows when I turn it on, so my set is pretty dead but almost lights the CFL anyway. Wind stopped right after I posted so it's not charging. I do have 4 other 19.2V packs charging on a cordless drill charger right now though.

 I don't have a 48V bank I can get to easy, Matrix bank would be great but I can't get to it right now. I'll probably wire about 50V on the sub-C cadium packs and try that when they charge up full.

Sorry I haven't made it back up yet, just everything going wrong here. Leaky roofs, vehicles breaking one after another. I fix 1 and 2 break. I had 5 I was driving, 4 nice and the ugly van, none are running drivable right now. I just put new tires and alternator on a real nice Jeep last week (#6), last night it sounded like the muffler was gone. Have to fix that today!

 I don't remember breaking any mirrors!

[nothing to lose]

Thanks for the tips there.

Not sure what the CFL will do, but I have a few things I can try that with and see what works. A solar pannel for a 6V electric fence charger is handy. Now to get the battery packs charged up. It takes awhile with a 1.3amp rated cordless drill charger. Over an hour per pack, I think it puts out less than the rating says. I should check that also with the next pack I do.

[Gordy]

NTL,

I'll have to try the GE cfl's. I've been using the LOA cfl's on my standard 120 ac, but am very disapointed with the life I've been getting from them, 4 to 12 month for the one's that get used regularly. I realy liked the light output & quality of light from their 1.5 watt night light. But the longest one lasted was 3 days past the 1 year warenty, Not even close to their (FALSE) advertising of a 5 year life. I'd say that's expensive lighting. Luckly I kept the sales receits, and could take them back as I have gone through 4 of these.

Rp, beat me to the suggestion about the grounded wire to the bulb. Would be easy to do or undo if it does'nt work.

I was playing with one of those dollar store ($5) battery operated Tesla ligthning globes, and some (1,2,& 3) burnt out flor. bulbs from a 12v lantern. It seamed like I got almost 2x the output with a ground wire hooked to the - of the battery pack and to one end of the bulb. With the other end of the bulb duct taped to the globe. Only got enough light out to make a good night light though. But then again the globe unit has a 6v wall wart adapter plug, needing 250 milli amps.

Gordy

[ghurd]

What if the 26W tube were changed to a 11W tube?

If it needs to supply less power, maybe it will work at a lower input voltage.

Sort of the same idea as a 9V wall-wart making 18V open.

G-

[whatsnext]

Any chance you could check the current draw next time you get your battery bank recharged?

Thanks, John....

[nothing to lose]

Good News/Bad News, mostly bad

I dug into the Matrix battery bank, 48V. It was kinda hard to get too.

I connected the meter to one post and one leg of the desk lamp and a wire from the other post and the other leg. A fully charged 48V bank lit the bulb!

Yea, good news.

 It was cold in the house and at the battery bank so the bulb was cold. At first it took a bit of time to light and I did not think it was going to work at all then it flashed on, nice and bright. Meter showed 0.23amps and it climbed to 0.27amps over a few minutes time. After the bulb lit the first time, it lit instant each time I turned it off and back on.

 I did not have a second meter handy to mesure volts but I'll geuss at 50V roughly since it was fully charge Gell cell 48V bank. So 0.27amps x 50V = 13.5watts roughly? This is a 26watt bulb so I was using far less power than rated. Even at 60V which I was well below for certain it would only be about 16watts.

The light was nice and bright.

Well that was the good news, now for the bad news.

Life expectancy of about 10 minutes

While I was playing with it everything was great. I left it turned on and left the room for about 10-20 minutes and when I went back it was off. Meter read 0.02 amps. I played with it, checking for a loose wire turning it off and back on etc.., it never tried to light on the 48V bank again. I put the bulb in the ceiling 120Vac light fixture and it barely glows at the tube ends now. It does not light up at all.

DARN, it was actaully working well enough I was planning to run a 48V line and use one for awhile. I have not taken bulb apart yet to see if some little part fried or if it's in the tube itself.

Now when I ran the same type tube on a 12V ballast board from a 12v 8watt flouresent drop light I did not have this problem. I also ran that 12V board at 24V though I had to add a little heat sink to a part. Maybe it is not the tube that fried on this 48V test?

Since the bulb does light up on a 48V bank and seems to work well (untill it burned out) perhaps someone with more electronics knowledge than me would like to test and fry one to see what actaully went bad. If lucky maybe it's an easy part to change, just solder in a new $1 part? That would be nice it we lucked out that way. $1 for a part and $3 for a bulb, a few minutes time, maybe have a good 14-16watt 48V CFL for $4.00 ??