Oznog

Certainly not. No transformer can ever create DC without a rectifier, no matter what you do with the primary.

Current in induced in the secondary winding when the mag flux changes. The mag flux increases when the primary transistor is on and creates current in the secondary. When the transistor is off, it decreases and makes current in the other direction. The sum of current over time must always be zero. The shape of the waveform is typically undesirable- when the transistor is on, flux increases rapidly making a high current spike on the output. Then it decreases at a more moderate rate when the transistor is off.

However, the core flux is always in the same direction. It increases and decreases but does not reverse, so the core is essentially biased at all times like an electromagnet. This reduces the capacity of the core. Furthermore, if you do not allow enough time for the flux to be drained by the load during the "off" period, current will increase even higher the next "on" period and so on until the core flux reaches a saturation point. At that time the current during the "on" period is limited only by the total resistance in the primary circuit and will typically draw enough current to smoke.

It is not a very appropriate design. He needs a split primary pretty badly.

__________________
I thought what I'd do was I'd pretend I was one of those deaf-mutes.

Hero999

Alright I'll be a bit more acurate doesn't create steady DC, it makes very small positive pulses and very large negitive pulses.

Try connecting the primary of a small mains transformer to an AA battey in series with a switch, connect a neon lamp to the secondary. When you hold the button the neon might flash very dimly as the field builds but it will flash even brighter when you release the button as the field collapses and notice the corrona is only generated around one electrode.


True but not much current is generated as the field builds slowly.

When the transistor suddenly shuts off, the current in the primary decay more rapidly and generates a huge back EMF, it explains why you can generate 1kV from a 1:20 transformer with only 12V at the primary.

You've got that the wrong way round, the current increases slowly when the transistor turns on, remember inductors don't like sudden change in current and it's for this reason that a huge spike is generated when the transistor turns of; it's like turning a relay off.

That's true in this case and it's why large negitive spikes are generated.

It does turn off (minus the small current in the feedback).

No, it's because the core is only ever being magnetised with one polarity.

That will only happen if the tube is disconnected.



Either way I agree.

But it does use pulsed DC, read the link I posted before. I've bought cheep DC fluorescent tube fittings before and one end of the tube always goes black and it doesn't illuminate uniformly

__________________
What's so bad about Microsoft?

Get Opera it's simply a superb browser.

Nigel Goodwin

There's too much rubbish on these forums about 'pulsed DC', there's not really any such thing - it's AC, as simple as that!.

The reason one end blackens is because the AC used isn't symmetrical, nothing to do with 'pulsed DC'. It's simply because the inverters used are crude, and waveshape isn't of much concern.

__________________
PIC programmer software, and PIC Tutorials at:
http://www.winpicprog.co.uk

Oznog

Running a tube with too little current will also blacken the ends in short order. I assume doing it with too much current will do the same thing.

Waveforms which heavily deviate from a sine wave are a problem, but so is failure to regulate tube current. IIRC those great compact flourescents actually require a bit of starting cycle algorithm to get long life. Take one apart sometime, the 110v ones- even the cheap ones- still have like two dozen components inside.

There are some quality 12V DC devices out there:
http://www.oksolar.com/n_cart/search...xtures%2012VDC
And these are some excellent 12v DC lamp bulbs. They're really good. I've used them.
http://www.oksolar.com/n_cart/search...=Light%20Bulbs

They're certainly expensive for a flourescent, but those are not the "crude" solutions mentioned.

The simple answer is to simply use a power inverter.

__________________
I thought what I'd do was I'd pretend I was one of those deaf-mutes.

Hero999

I did explain this a couple of posts a go.

None the less you do have a point, but how else should I explain it?

How about this:?

This inverter is no good because the AC waveform it produces is very asymmetrical (mostly negitive). Current only flows in one direction through the tube the electrons are only emitted from one end of the tube, this causes one end to be brighter than the other and excess electrode splutter also occurs at one end cause it to go black.

__________________
What's so bad about Microsoft?

Get Opera it's simply a superb browser.

Sig239

First this guy revives a year old thread with rubbish, then the next guy comes in two months later......next thing you know...A fine tale of two newbs triggering a frenzy over an old thread is all but written in the history books.
I have begun to profile different threads. The first thing I do when I see a thread with more than 600 views is check the dates. If you see a thread with like 6 replies and 1500 views it seems to almost always be revived.
EDIT: In this case 19 replies and 2023 views LMAO

Hero999

Anywaym powering a fluorescent tub is a very common request, and it' easy to do but harder to do properly. If you have any more questions then I can point you in the right direction. Really the only hard part is the transformer which can be a bit experimental but it's not too bad really.

__________________
What's so bad about Microsoft?

Get Opera it's simply a superb browser.

Overclocked

Its more efficent to go and buy a CCL for computer modding. You can get em cheap for 2 bucks at SVC when they have deals. All things for computers work for 12V, all you would have to do it modify the connector.

http://svcompucycle.stores.yahoo.net/12vb2ccflwht.html
White, they go for 6 bucks.

Hero999

They won't power standard fluorescent tubes though but I see what you mean, the idea is to use the CCL instead.

__________________
What's so bad about Microsoft?

Get Opera it's simply a superb browser.

Overclocked

Right, The whole thing comes with an Inverter. But still it does make a good HV circuit, and they do produce nice arcs

EDGE

This is true about the differance of driving flourescents withpulsed d.c bias . It is also an apparently complex and overkill cicuit for the application which could be =effectivly implimented with 2 transistors driving a center tapped transformer with one hihg volt secondary and one intermediate winding to alternatly toggle the bases / gates . this scheme can be point to point wired withuot any p.c. board . Oh yea ,I forgot that you need 2 bias resistors and 1 capacitor.

Hero999

That's exactly what I was talking about.
http://www.repairfaq.org/sam/samschem.htm#schmp1

__________________
What's so bad about Microsoft?

Get Opera it's simply a superb browser.

delta06

Built this one but I can't get it to work somebody please help.

Hero999

Have you tried swapping the base connections to the transistors?

__________________
What's so bad about Microsoft?

Get Opera it's simply a superb browser.

IQBALSOUND

Hi hsab...nice tricky oscillator but is this circuit is reliable?