Flyback transformers - how and why.

This Ible was written to answer the questions of fellow Ibler, Electorials. I hope it helps.

The flyback circuit is a classic method of generating moderately high voltages. It is not, by any means, the ONLY way, but it was once a very common circuit, and was employed in EVERY CRT based TV or monitor.

It is a VERY clever circuit, because it solves two design problems in the system with one circuit.

The scanning electron beam in a CRT needs to be a.) created and b.) scanned.

All pictures bar two are taken directly from Wikipedia.

Step 1: Creating an electron beam

The electron beam is created in an "electron gun". A wire filament is heated strongly in a vacuum, and in the presence of a high voltage. Electrons liberated by the heat and voltage are accelerated down the tube and attracted to the screen, where they stop suddenly and cause a phosphor to glow.

Wikipedia

SO, we need a high voltage.

Comments

[-max-]

So how do I limit the power to the flyback when the core does saturate? would running a higher freq. help? or lowering the duty cycle? but all that would decrease the output, and I don’t want that. adding a resistor between the coils and the transistor would fry just about any resistor, would a inductive ballast/and capacitor in a tank circuit config in series with the flyback/mosfet help?

[steveastrouk]

You get NO output when the core is saturated, and, like the formulae say, energy stored in the field = 1/2 L i²

If L = 0, because your saturated, you get no energy storage

Running a higher frequency will help greatly - remember they are DESIGNED for 64kHz typically. 
Carefully watching the current in the core, and then cutting it off at the maximum value, before saturation is what is needed. This 'ible has the maths you need to work out the F for a given V and L.

[a226263]

I bought a brand-new flyback a couple weeks ago. I'm trying to determine the pins to the primary coil on my flyback.

My flyback has 10 pins. I found out which pins are connected by testing if there was a resistance between two pins with a multimeter.
Pins 1, 5, 9 are connected; pins 2, 8 are connected; pins 3, 4, 6 are connected; pin 7 is by itself; pin 10 is by itself.

This seems problematic because there are only two coils on a flyback, but I have 5 different groups of connected pins. Does this mean that something is disconnected inside my flyback?

Do you have any other ways to determine the primary coils besides this http://lifters.online.fr/lifters/labhvps/tht.htm

[steveastrouk]

Many other coils are frequently wound on a flyback - for other power supplies, for oscillators driven by the coil itself. It can be hard to work out the best pins.

[waldosan]

i have a big old transformer dug out from a junked stereo, it isn't specifically a fly-back transformer but could i use the same circuit you showed in step 4 to get a higher voltage?

[steveastrouk]

Not really. The turns ratio won't be anywhere near right, and the core will saturate very quickly.

Steve

[iceng]

Great point about saturation of the core and the resulting conflagration
where the only limit on current is the small DC current path resistance.

A path that includes the switch, the coil and the power source resistance, which will fuse a silicone semiconductor in an instant and heat the wire insulating encapsulation into a chard coil scrap.

A

[steveastrouk]

Hi A.

Seen anything you'd change ?

Steve

[Electorials]

Oh hey, look! You have a comment even before your instructable was published.
Thanks for this, it explains a lot.
Is this picture somewhat correct?
it's what I now understand about how it works.

http://www.instructables.com/files/orig/FCU/X6WR/GUQRR77M/FCUX6WRGUQRR77M.jpg

[steveastrouk]

Not bad, but flux is proportional to current, and if it didn't change there'd be no terminal voltage....

[Electorials]

At school they told me it differently: (not saying you are wrong, but it confuses me)

The flux is not proportional, but more like di/dt.
This means it will be zero if the current doesn't change. (a constant 200mA current is also a non changing current)
It's negative if the current Decreases, and positive if it Increases.

If the current increases Linear, the flux will have a constant value
If the current increases quadraticly, the flux will be linear
(Each time one power lower)

[wannabegeek]

Flux can be thought of in non E&M terms too.

Flux is the amount of stuff moving past a point, or more typically through a plane, or through a Gaussian surface.

Guass' Law = Divergence Theorem.

Current , I, is a rate A/s, that is a flow and thus as long as the electrons move throw the surface you have chosen, there is flux.

Flux is positive by convention, if the flow is OUT from the origin and negative if the flow is IN .

Since the flow can be a function of time or some other parameter, we typically integrate to find the total flux. Flux is also a function of space, and we often want to add up all the field lines passing through a plane or sphere.

We almost always use symmetry arguments when integrating flux because fields get too complicated with funny shapes or edges of planes.

good luck !

[steveastrouk]

Charge, not electrons.

[wannabegeek]

[:dunce cap:]

yeah duh...field lines ....

[wannabegeek]

dang...can't delete or edit my comment....

[:dunce cap:] again

charge not electrons ...I still get this confused obviously...

Is charge carried by the holes ?

[steveastrouk]

He he

[steveastrouk]

How old are you ? What level of school are you at ?

NO !! FLUX is ALWAYS present if I is present !!


Are you happy with differential calculus ?
V = L di/Dt or V = n d (phi) /Dt

It all comes from that really. But flux is effectively proportional to NxI

dI/Dt produces dPhi/Dt.