# Re: How do you measure coupling

```Original poster: "Antonio Carlos M. de Queiroz" <acmq-at-compuland-dot-com.br>

Tesla list wrote:
>
> Original poster: "Mike Tucknott" <michael.tucknott-at-virgin-dot-net>

> How do I measure the coupling between the primary and the secondary.
> I can get hold of a sig gen and a scope, I`v no idea how to operate these
> two bits of kit, I need the basics and some theory..

A method that takes everything into account, and also tunes the
coil:
Assuming a conventional Tesla coil (magnifier below):
Assemble everything, primary coils and primary capacitor,
secondary coil in place with top terminal and ground, spark gap,
but do not connect the power supply (that would be across the
spak gap).
Connect the signal generator across the gap, and connect a small
resistance across the gap too (as 1 Ohm). This is to lower the
output impedance of the signal generator to the order of the
effective resistance of the gap.
Set the signal generator to a square wave with a frequency around
500 Hz, with the highest amplitude.
Connect the oscilloscope across the primary coil, with the trigger
input connected to the output of the signal generator (or to a
trigger output if it has one).
At each transition of the square wave, you will see a waveform as
this:
http://www.coe.ufrj.br/~acmq/tesla/tefpreg1.jpg
This is the same waveform that would appear at each discharge of
the primary capacitor, scaled down.
Adjust the primary tap position until you see full notches, as in
the picture. The system is then correctly tuned.
You can take streamer loading into account if you connect a wire
with the expected shape to the terminal, or simply approach your
hand to the terminal.
Count the number of cycles b in each of the beats, between the
notches (about 8.5 in the picture). Let a=b-1. The coupling
coefficient is then k=(b^2-a^2)/(b^2+a^2), and the coil is
operating in mode a:b, meaning that it resonates at two
frequencies with the ratio b/a.

If the output of the signal generator is too small, making the
measurement too noisy, you can try to make a "tuner" circuit:
http://www.coe.ufrj.br/~acmq/tesla/tuner.gif

For a magnifier, follow essentially the same procedure. But a true
magnifier has another point to tune, that is the capacitance
across the secondary coil. If the system is properly tuned,
the voltage across the secondary coil when the voltage at the
primary coil is zero is zero too. See the waveforms at:
http://www.coe.ufrj.br/~acmq/tesla/mag345.html
This voltage can be measured with a 10x probe in the oscilloscope.
Increase or decrease the capacitance across L2 until you see
coincident notches at the primary and secondary voltages. This is
not difficult if the magnifier is designed considering this from
the start. If you will use my formulas, choose a mode of the
type a:b:c, where b=a+1 and c>b+1. This results in smaller
capacitance across C2, that is easier to obtain with distributed
capacitances. The program mrn6, at:
http://www.coe.ufrj.br/~acmq/programs
can evaluate the formulas for you and plot the expected waveforms.

Or, if you just want k, simulate the structure. Programs as Acmi,
Fantc, or my Inca (same site above) can calculate k from the
geometry better than you can measure it.

Antonio Carlos M. de Queiroz

```