[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Some transformer theory



Original poster: "Jim Lux by way of Terry Fritz <teslalist-at-qwest-dot-net>" <jimlux-at-earthlink-dot-net>

 > >
 > >Can the root of the impedance ratio also be applied to flyback
 > >transformers - are they "classical" transformers, or are they in the TC
 > >bracket, where turns ratios cease to be entirely relevant?  I didn't have
 > >one with a known turns ratio to test, but measured one and the ratio
 > >*looked* sensible.
 > >
 > >Cheers
 > >
 > >M
 >
 > Hi M ...
 >
 > normally flyback transformers are made with a ferrite core...   they do
 > behave differently. One of the "neet" things about a flyback ( or a
ferrite
 > cored Xfmr) is that they can be driven to a resonate state.  You can apply
 > a voltage to them over a broad range and then all of a sudden you will see
 > a tremendous peak in voltage on the secondary....

This has nothing to do with it being a flyback, or a ferrite core... It's
just the inductance resonanting with the stray capacitance.  Auto ignition
coils show this behavior, and they have an iron core.

 >
 > I found this out by experimenting with a smallish ferrite cored Xfmr from
a
 > photoflash ( a tad bit bigger than a sugar cube) ...   I was feeding the
 > primary with a sine wave at about 2p-p volts while varying the frequency
of
 > the sine wave ...
 >
 > from 50 Hz to 70KHz the output was on average 1:100...  at 80KHz I started
 > to see a rise in the secondary....  at 100KHz  the 2 volts p-p sine
 > wave  went to 600 volts secondary side  ( 1200 V p-p) ...   a 1:600
 > ratio   now mind you...  the windings never changed :) obviously the
 > frequency determines the output voltage :)

Unloaded output voltage, I take it?
Also, was the source voltage actually held constant or just the signal
generator left at the same output setting?  If the generator has some finite
non-zero output impedance (i.e. like most...), then you might also be seeing
an effect as the load impedance varies.  For a lot of lab sig generators
(e.g. a HP3325) the output amplitude setting is what voltage the output will
be, into a resistive 50 ohm load.  Essentially they set the internal output
(zero impedance) at twice the setting, and feed through a 50 ohm series
resistor (making the source impedance 50 ohms).

What was the measurement method on the output? An oscilloscope?  Terminated
in 50 ohms? Straight into the  1 Meg -at- X pF input? A standard 10x probe?
You'll get different results with each.

 >
 > at 120 KHz the voltage dropped back down to the 1:100 ratio again...

So, a peak at 100 kHz, and 20 kHz on either side you're back to
non-resonant?  If you measure where the voltage is .707 * 600, you can
measure the Q of the circuit (and the loss will be mostly in the ferrite
core and copper wire). If we just use the "resonant rise" of 6 as a starting
point, we'd expect that the bandwidth would be on the order of 100/6 kHz,
or, 16-17 kHz... hmmm.. pretty close to what you observed, I'll bet...
 >
 > so  in effect its safe to say that at 100KHz for this particulary Xfmr
 > there is a resonate rise which produces its maximum voltage...
 >
 > hmmm ...  I wonder of iron core Xfmrs have a simular reaction....

yes they do..