FL Inverter schematic

[RobC]

Here's the schematic feel free to make suggestions to improve the circut. For the transformer I suggested using a ferrite E core but I think the core from a large flyback transformer from a tv would probably work as well and be a lot easier to wind. I will try it as soon as I have time.RobC

[Jerry]

Hi Rob.

What is the voltage accross the lamp while lit?

                    JK TAS Jerry

[RobC]

I'm not sure, Jerry.  My meters go wacky if I try to measure it. The 8ft bulb I was using is an F96T12CW single pin on each end cold cathode type and are still common everywhere they just don't start well in an unheated building. Temp in my shop is probably 50 degrees when I tryed it an I can tell you it lit instantly without hesitation. I will see If I can find a ballast at my hardware store and see what starting voltage is I think its around 900 volts open circut. In the mean time I will try to figure out how to measure the voltage when the bulbs are lit.RobC

[johnlm]

Rob,

Just curious, Did you try the ckt without the totempole buffer/driver?  I would have thought the 555 could drive a mosfet directy since there is practically no gate current.  Also that type of complimentry symmetry emitter follower driver stage (npn/pnp)is typically used if you are driving bipolar voltages + and - . It looks like in this ckt the pnp is not really doing anything unless the gate of the fet wants to float up above ground when the npn is turned off (output of 555 is at 0 volts), then it would allow the pnp to get turned on and pull the gate down to around +0.6 V.  But this could be also accomplished with a resistor to ground of say greater than 10 K ohms instead of the pnp.  Probably worth a try.  Ive looked at several 555 application books (signetics, who invented the part, put out alot of app notes back in the 70s and 80s on this part) and it looks like in similar applications , they drove the transformer primary directly off the 555 if very low power, or drove a single npn power transistor operating as an emmiter follower buffer between the 555 and the transformer if medium power.  Seems like I remember reading a post by you or someone where you tried using a 2n3055 in this circuit and had problems with them overheating or failing.  That might have been caused by inductive voltage kick back from the primary when the transistor turned off and the transformer field collapsed.  Usually a normally reversed biased catch diode across the emitter to collector is used to steer those kickback voltages to ground so they did not exceed the collector to emmitter breakdown voltage and destroy the transistor.  Most power mosfets have one of these diodes built into them (between the S and D) for that very purpose.

I suspect this ckt is more stable and sure to start than the transformer winding feedback single transistor oscillator type that was on the site BT Humble wrote about a few weeks back.

I may need some florescent lighting in a hut up in the mountains this summer and might build one of these for that purpose.  Thanks for sharing the effort and your ckt.

John

[deerslayer660]

   im playing with a 555 ckt to now and just wondering may be going to try  winding say 500 t for a secoundary and 18 to 50 dont know exactly what to use on a rod like hb humbles ckt any think this wont workbefor i try and let out thr preety blue smoke george

[RobC]

The 555 will definately drive the mosfet directly. The turn off turn on may not be quite as positive. But it will work. The extra 50ns diode across the mosfet just takes some load off of the built in diode. RobC

[RobC]

I would say 20-25 turns of 16awg wire for the primary if your going to use an antenna rod. However its just a guess on my part. Use plenty of tape between layers to prevent arc through. An amp meter will really help you to keep your circuit from frying. Good luck let us know how it works out. RobC

[RobC]

If you want experiment with different value try this.

[RobC]

George, I took an antenna rod transformer and removed all but the 450 turn secondary. Next I wound on 35 turns of 17 awg wire. Guess what, it also lights up the 8ft tube. After 5 minutes it was a little warm but not bad. It doesn't look like I got that much more out of the E core transformer and it was a pain to wind. Is this fun or what! RobC

[commanda]

I agree that the totem pole is not necessary, as the gate current into the FET is almost non-existent. Possibly a series resistor from pin 3 of the 555 to the gate of the fet, and another pull-down from gate to source. Make sure that fet is really off when the 555 output goes low. Still probably over-kill.

To estimate the open circuit volts, place a diode & capacitor in series across the transformer secondary. Use really high voltage types. The dc voltage across the capacitor will equal the peak voltage of the secondary. Connect meter with alligator clips before turning on. After turning off power, discharge the capacitor. You'd be surprised how long those volts will hang around just waiting to bite you. As the frequency is pretty high, you'll only need a small capacitor, certainly less than one microfarad.

With circuits like this, it's also a good idea to use a pot in the RC side of the 555, to vary the frequency. Pick that frequency which gives the highest output voltage.

With those 2 diodes in there, and your choice of resistors, I'll bet the mark/space ratio from the 555 is nowhere near 50%. I think you'll find it will work better if your square wave is more square, rather than a pulse.

Amanda

[johnlm]

Amanda,

Nice to see you back.  I missed your comments / posts as you always seemed to know the electronics side of things pretty well.  I almost wrote a post a few weeks back asking if anyone knew what happened to you.

Regards

Johnlm

[RobC]

Just wanted to say thanks John and Amanda  for the  advice on circuit improvements I shall try them soon.

[commanda]

Haven't gone far. Just snowed under. Started with Xmas, and getting the car fixed up for rego (January 5th, bye-bye xmas holiday pay). Then kids, spouse, work, ferret, kids, cricket, it never ends. Did get to drive up North and have a look at our 2 acres of retirement bliss though. Still counting the decades till then.

Back to the FL inverter. Anybody thought of using the transformer out of an old computer power supply. These things get dumped by the millions. Here in Australia, they run 340 volts DC on the primary, with 5 & 12 volt secondaries. Seems to me, if you ran it backwards, got to be in the ballpark. Just a wild thought.

Amanda

[monte350c]

Here's a question - is there any danger in exciting a fluorescent tube with voltage or frequency other than standard - ie can the tube put out ultraviolet or other kinds of light that may be harmful?

I really don't know and would like to before getting too far into this!

Ted.

[johnlm]

Rob,

Might I ask what is the purpose of the diodes around the resistor between pins 6 and 7?  In all the applications Ive seen, Ive not seen diodes used in that location.

John

[jimjjnn]

Ultra-violet Fl are normally clear glass. The phosphorescent white coating changes it to visible light filtering out all the uv elements

[RobC]

Hi John, The diodes are needed when you want less than 50% duty cycle. My output is more of a fast pulse than a square wave. Probably not optimiun but it works in my setup. RobC

[TERRYWGIPE]

Yes if you go to a Sceince fair with a testla coil yuo can light a fluorescent bulb juse by putting the bulb near the tesla coil.(works great with neon too.), just screws arround with RF.

      Terry

[RobC]

What I'm working on is something that can be used 24 hours a day. And doesn't get you in trouble with the FCC. And the light output is the same as a standard ballast.RobC