DIY CCFL inverter?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Jump to Latest
S

Stocker

Member
Joined 2004
The point of making my own power supply is to test my free, unknown condition CCFL tubes. Buying one sort of defeats the DIY point, at least to me. Besides, I can make a transformer put out whatever combination of Vs and As I like. More power from a stock part = more money for more parts!

Any special ideas on current limiting (high-power resistor comes to mind)

Why use a choke?
 
Sch3mat1c

Sch3mat1c

Member
Joined 2003
seventransistorlabs.com/
Choke is lossless (well...sorta), whereas a resistor only burns heat.

If you want an inverter specifically, you might try a quickie 1:10 or 1:20 ferrite core transformer and a MOSFET tied to a 555. The flyback pulse is inherently current-limited and will be capped to the tube's forward voltage. Mind it's also DC, if the tube reacts fast enough (since the voltage is assymetrical).

Tim
 
S

Stocker

Member
Joined 2004
Rectified = 120Hz or I have missed something in your idea.

No, I *have* missed something in your idea b/c I don't know what you mean by "on" voltage.

370H... :bigeyes: YGBK! Small, only the size of a beer can! How many miles of 22AWG would that take! LOL, not poking fun at you just at the idea of an inductor that big, as I am a low-volts, solid-state type fella.

Last night I had the idea of a variable voltage supply, from say 800 to 300 varied with the duty cycle on the output FET. I'm thinking a good HOT from an old TV would do the trick. High voltage... :att'n:
 
Sch3mat1c

Sch3mat1c

Member
Joined 2003
seventransistorlabs.com/
Stocker said:
Rectified = 120Hz or I have missed something in your idea.
Click to expand...

Uh, that wouldn't work very well, since chokes don't limit DC current.

Fluorescent tubes also need AC, otherwise one electrode wears faster and life is shorter.


No, I *have* missed something in your idea b/c I don't know what you mean by "on" voltage.
Click to expand...

All plasma discharges have a certain ignition voltage and a certain forward voltage. Forward voltage behaves somewhat like a zener diode, but voltage drops slightly as current rises (negative dynamic resistance), which also prevents the use of a capacitor across discharges (unless you want it to oscillate: see neon light relaxation oscillators). But that's not a concern here.

Anyway, take a clear path between two electrodes, in any gas at any pressure. As you increase voltage, eventually the electric field becomes strong enough to ionize and it becomes a plasma. Before it was ionized, conductivity is almost nil - a good insulator. After ioniziation, conductivity increases and current flows. Given a current-limited supply, voltage drops dramatically.

That plasma is a conductor while plain gas is not is why you need a current limited supply and a high starting voltage.


370H... :bigeyes: YGBK! Small, only the size of a beer can!
Click to expand...


Eh, maybe a quarter or half can. Not bad.

How many miles of 22AWG would that take! LOL
Click to expand...


Um, quite a few, but 22AWG would also have to be wound on a four ton iron core. On the other hand, it *would* be capable of storing 0.74 kilojoules at four amperes.

not poking fun at you just at the idea of an inductor that big, as I am a low-volts, solid-state type fella.
Click to expand...

Well you said 5 miliamperes, not five whole amps. That'd use 40AWG or so, probably 20,000 turns or so.

Last night I had the idea of a variable voltage supply, from say 800 to 300 varied with the duty cycle on the output FET. I'm thinking a good HOT from an old TV would do the trick. High voltage... :att'n: [/B][/QUOTE]

Keep it AC. A chopper style oscillator with a loosely coupled secondary (to limit current) would work best. Which coincidentially...is exactly what a CCFL driver is. :clown:

Tim
 
S

Stocker

Member
Joined 2004
:D I am working round from the wrong way to reinvent the CCFL inverter! I was just about to make a post about the AC vs DC business and you already had covered it somewhat. I love this website.

:xeye: ugh. Too much thinking here to end up with one component and fluorescence. I was going to say BANG, fluorescence, but I suppose the BANG would be easy enough with one component! It is really too bad that I sold my 8000lb iron core just last year... :)

Sooooo... am I back to searching for inverter schematics yet? I think not quite yet. So how do you limit current? With a fat choke I suppose... ( ! ) But how do you use the same current limiting circuit for a tube that needs 800V to strike, as with a 300V tube? (thinking on the keyboard now) At the input dummy!

Forgive my ignorance... I only realized a few days ago, thinking about this project, the way that PWM power supplies work, chopping up an available voltage. Proving once again that book learnin' will be anywhere from helpful to useless, until you figure it out for yourself.
 
Kenshin

Kenshin

Member
Joined 2005
It's better to test it with HOT :att'n: from a old TV horizontal transformer (powered by a MOSFET?) and then buy a COMMERCIAL inverter.

DIY High Voltage transformer arcs easily. It may not blow the circuit if the current is limited. But the transformer may overheat and cause a fire :hot:.

The transformer may run peacefuly for a long time, until it becomes old or wet and then arcs. So the time of the fire is entirely unknown.

Especially dangerous for a lamp inverter -- when it start to smoke, you are probably not watching, and the fire goes big... :dead:

Also, adding fuse can't protect you against this danger.:bawling:
 
S

Stocker

Member
Joined 2004
I have had an idea, and surely it cannot be this simple. I have an old computer power supply laying around. Can I install a high voltage winding on the main transformer, with a current limiting cap. on one wire, and run it like that? Wouldn't the 5 and 12 volt regulating circuitry keep the thing relatively stable, by PWM'ing the transformer's primary?

And is there a good way to run about 18 inches of lead wire to an inverter from the lamp?
 
S

Stocker

Member
Joined 2004
Well, that's awfully pessimistic of you!

I have lots of small-guage wire. I might have to give this a try someday this year. Watch this space, I guess.

I am back on this project for one reason: I was given a scanner.

That is, I was given a scanner that is considerably chumpier-looking than my own, with a working, always-on CCFL light source. That I have verified will illuminate anything from a 10" to a 30-ish inch 3mm CCFL tube. Now I have a good / bad tube tester anyway. Those long tubes do cool things depending on where you touch them.

And they break like a sunnova :bawling: when you flex them! The project would have been good if I hadn't broken THREE of those things. Maybe I'll get some more. Anyhow, I am planning on running maybe 20, 14" tubes at low current each for a BIG BRIGHT tracing / light table.
 
S

Stocker

Member
Joined 2004
Well, there is some potential, literally and figuratively. I ran five loose turns of thickly insulated wire through the toroid in my AT supply and picked up a significant fraction of a volt. I can't say how much exactly because the leads for my DVM are going out. Anyhow, it did show some VAC there. I took the holder for the three dead tubes (make that 20" BTW) and cut and bent it into a circle that I will try to use for a bobbin to spin some large number of turns onto the transformer once it's out of the power supply.

How much current can you run through 170 feet of 24 or 26AWG without it heating excessively? I guess we'll find out. I need to dust off the oscilloscope and see what frequency this thing switches at so I can figure out a current-limit cap. value.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Top Bottom