[sci.electronics] Flyback Transformers

hgw@MATH.UCLA.EDU (07/27/88)

Help!!  Does anybody know of a supplier of flyback transformers?
I am in need of them for a Freedom 100 terminal.  I've tried to call
Liberty Electronics but they don't seem to exist anymore.  The number on the
transformer is 1212-00006 YGCT.  I've been using a KFS-20708 DHAK as a
substitute and it seems to work well.  Any suggestions would help.

How are these transformers described?  Are all the part numbers unique to
each manufacturers.  Are there cross reference books?
-------------------------------------------------------------------------------

Harold Wong         (213) 825-9040 
UCLA-Mathnet; 3915F MSA; 405 Hilgard Ave.; Los Angeles, CA 90024-1555
ARPA: hgw@math.ucla.edu          BITNET: hgw%math.ucla.edu@INTERBIT

larry@pdn.UUCP (Larry Swift) (07/27/88)

So what's a flyback transformer, and how did it get its name?

Larry Swift                     UUCP: {peora,uunet}!pdn!larry
Paradyne Corp., LF-207          Phone: (813) 530-8605
P. O. Box 2826
Largo, FL, 34649-9981           She's old and she's creaky, but she holds!

rsd@sei.cmu.edu (Richard S D'Ippolito) (07/29/88)

In article <3844@pdn.UUCP> larry@pdn.UUCP (0000-Larry Swift) writes:
>So what's a flyback transformer, and how did it get its name?

There were sort of like passenger/carrier pigeons -- they were kept in wire
cages and used for carrying messages from screen to screen and always
returned reliably. But, alas, they are now extinct, having been mercilessly
slaughtered with electron guns...

Rich

dya@unccvax.UUCP (York David Anthony) (07/29/88)

In article <3844@pdn.UUCP>, larry@pdn.UUCP (Larry Swift) writes:
> So what's a flyback transformer, and how did it get its name?

	This comes from the darker ages of television engineering,
and unfortunately continues to this day.  The original "flyback"
term has to do with the topology and operation of the horizontal
sweep in a television receiver.

	Exceedingly basically, you have a switch (transistor, 
tube, whatever) strobed at the horizontal line rate.  This switch
is connected in series with an appropriate inductor. Now, assuming
you have no current initally in the inductor, and you close the
switch (the inductor connected to your handy-dandy perfect 
voltage source) you have the dreaded relationship

	v(t) = L (di/dt)  (has it been that long :-) :-)

	Neat. Now, the net.magnetics experts can show you that
the changing magnetic field inside a solenoid inductor can be
used (with a little help from some more nonlinear stuff, but who
cares at this point) to deflect the CRT electron beam and make
a raster.

	However, when we want this beam to "fly back" we want
it to do so instantaneously. We write the beam across the CRT at
a constant surface velocity (which is not the same as a constant
angular velocity on the deflection axis--why?? :-)) but would
like to start the next line instaneously after the end of the 
first.  In reality, you can't change the magnetic field or 
current in the inductor instantaneously, but for TV purposes,
fast enough is simply to open the bloody switch. The magnetic
field collapses damn fast, causing the terminal voltage across
the inductor to increase drastically, among other things, etc.

	If we make the switch the horizontal output transistor,
provide a damping diode to recover the negative excursion of
the inductor current (it will oscillate at some natural frequency
because there is a physical capacitance); make the inductor a
transformer coupled to a deflection yoke, and put a big fat
autotransformer coupled to the same transformer with a half-wave
rectifier (tripler, or other such ilk) you have a nice high voltage
supply made from the collapsing field (and rising terminal voltage,
multipled through the autotransformer) during retrace. The
high voltage rectifier doesn't conduct during scanning, so
no big deal.  Instead of the collapsing magnetic field doing
everything in the circuit in (because that voltage will flash
over somewhere, eventually), you've harnessed it and made the
accelerator voltage for the CRT. Hence, the "flyback" transformer.

	Now, in reality, we tune this turkey to ring at the
(third, fifth, seventh) harmonic, and we rearrange the windings
on the transformer so that (in the vacuum tube case) the damping
diode provides the necessary boost in plate supply voltage to
the horizontal output tube. As a consequence, we get all kinds
of other neat voltages from this transformer.  Virtually every
CRT monitor today gets not only acceleration voltage, but 
all other supply voltages from the "flyback" transformer. It
is a very efficient switching power supply.

	Anyone over the age of 25 will tell you what a drag it
is to change a flyback transformer in a consumer TV set.

	Really, folks, it's just like the points (whatever) in
your car, with a rectifier on the end :-) :-)

York David Anthony
DataSpan, Inc.

PS. There were some really crummy flyback designs out there, too.
How many of you have changed the transformer in an RCA CTC-10
through CTC-51 (the one where the plate cap for the 3A3 is 
moulded into the secondary winding.) I thought so. :-) :-)

myers@hpfclm.HP.COM (Bob Myers) (07/29/88)

A flyback transformer is that little odd-looking transformer that makes the
high voltage (CRT anode voltage) in your TV set - at least, that's one
example.  It transforms the high-voltage (several hundred volts, typically)
spike produced by the retrace, or "flyback", of the horizontal coils in the
deflection yoke.  (This spike results from attempting to quickly reverse the
current in this inductance - V = L di/dt, right?)  Hence the name.  Used quite
often in situations like this, where you need a high voltage at relatively low
current.

Unfortunately, I can't help the original poster in his search for a replacement
part - there a literally hundreds of companies, many quite small, making
magnetics for consuumer equipment.  If the transformer you tried seems to 
be working OK (not running too hot, sufficient brightness, and the anode
voltage to the CRT is not over nominal - excessive anode voltage produces
excess X-ray emissions), then I wouldn't be too worried about it.


Bob M.

hplabs!hpfcla!myers

rpw3@amdcad.AMD.COM (Rob Warnock) (07/31/88)

Couple of other trivia notes about flyback transformers:

The high-voltage rectifier -- back in the vacuum-tube days -- was of
course a vacuum tube, and needed filament [a.k.a. heater] current.
But if you consider that the cathode was at the output voltage (look
at how you have to use a diode), that meant that the cathode-to-filament
insulation would have to withstand 15,000-35,000 volts... if they were
separate. So they were connected.

But now you had the task of making a filament transformer that would
withstand 15,000+ volts, and handle a few hundred milliamperes at 60
Hertz. No way! It would be *huge* (and exPENsive). So instead they ran
the filament on the 15 KHz horizontal frequency: the "transformer" was
typically 2-3 turns of (heavily) insulated wire wrapped around one side
of the ferrite core of the flyback transformer. (The flyback was normally
made of a couple of large U-shaped pieces of ferrite butted together
to make a big open "box" shape, with the actual autotransformer coil
all on the top side, and the mounting bracket on the bottom side.)
Worked like a charm.

More trivia: While color sets needed an explicit output capacitor to
keep the voltage stable enough, many cheap black & white sets simply
used the capacitance between the coated inside and the coated outside
of the picture tube as the "filter". Since the high-voltage feed point
was covered with a rubber "nipple", and since the vacuums in the picture
tube and the high-voltage rectifier diode were *good* insulators (as were
the well-insulated wires), it was not uncommon for the high-voltage to
persist for *weeks* after a set was turned off. You could get a *nasty*
shock from a set you pulled out of a trash pile (for example). And in
doing service work (which I did in high-school), you learned (sometimes
the hard way) that the glass of a picture tube had some sort of "memory"
for the voltage (or maybe it was a electret effect), because it wasn't
sufficient to discharge the high-voltage once (which you usually did
by slipping the end of a grounded screwdriver up under the H-V nipple
on the picture tube) -- to be safe, you had to discharge it 4-5 times
at first, and then maybe 2-3 times again over a few minute period.

The safety rules around the shop said that if you were going to leave a
picture tube lying out on the bench unattended for more than a few minutes,
you had to clip a drain wire between the H-V nipple and the ground strap
on the tube.  (Yes, I got zapped at least once by a "discharged" picture
tube that had regained some of its charge. You *don't* want that to
happen while you're carrying it!)


Finally, I once had a need for some high-voltage in an easily portable
package (to power a surplus aircraft strobe light being used as a
"come home" beacon for some campers... uh... don't ask). I took a
flyback transformer -- including a 1A2 (?) diode tube with the filament
"transformer" mentioned above -- and wrapped an additional few turns
of heavy wire around the unused side of the ferrite core (are you
still counting?), and kludged up simple oscillator from a car-radio
audio output transistor. With a 12 volt lantern battery as input,
it took about 15 seconds to charge up the output capacitor (surplus
power company "power-factor" adjusting cap) enough to let the strobe
fire.  (I got enough over-voltage I didn't need a trigger for the strobe.)

Ah, the good (?) ol' days...


Rob Warnock
Systems Architecture Consultant

UUCP:	  {amdcad,fortune,sun}!redwood!rpw3
ATTmail:  !rpw3
DDD:	  (415)572-2607
USPS:	  627 26th Ave, San Mateo, CA  94403