Experiments with coupled circuits
In the next measurement I measure the coupling between my antenna unit1 and detector unit1.
For measuring the coupling between the two LC circuits I use this schematic.
From left to right we see:
 A sweep signal generator
 A dummy antenna circuit
 The antenna tuner unit
 The detector unit
 A measuring amplifier
 An oscilloscope
The detector unit is loaded with a resistor of 1.5 M.Ohm.
The signal generator gives a output voltage of 0.2 Volt peakpeak (unloaded),
loaded with the dummy antenna it is 0.1 Volt peakpeak.
The frequency is constant varying between two set values, on the oscilloscope
screen you get a frequency spectrum with the response curve of the two circuits.
The antenna tuner unit has two tuning capacitors, there are much possible combinations
of settings to get resonance in the antenna tuner unit.
There is however only one combination which gives the highest voltage across the
coil.
In the table (click here) you can find
the value of the (series) inputcapacitor which will give the highest voltage
across the coil.
These values are used in the following measurements.
In the first measurement, the response curve of the receiver is measured at
600 kHz.
The distance between the coils is 20, 30, 40, 50 and 60 cm.
The horizontal scale is 10 kHz / cm. The total screen width (10 cm) is
equal to 100 kHz.
So the left side of the screen is 550 kHz, the right side is 650 kHz, and
the middle is 600 kHz.
Horizontal: 10 kHz / cm vertical: 1 Volt / cm
600 kHz 10 kHz / cm Distance: 20cm 
600 kHz 10 kHz / cm Distance: 30cm 
600 kHz 10 kHz / cm Distance: 40 cm 
600 kHz 10 kHz / cm Distance: 50 cm 
600 kHz 10 kHz / cm Distance: 60 cm 
If the distance between the coils is too small, overcoupling will occur, and
we see two peaks in the response curve.
With increasing distance between the coils, the distance between the peaks will
get smaller.
At a certain distance we have only one peak, this is critical coupling between
the coils.
At a further increasing distance, we keep one peak with decreasing height, so the
receiver will get less sensitive.
A piece of theory about the coupling between LC circuits you will find here.
Next the response curve is measured at 1500 kHz.
Again at 20, 30, 40, 50 and 60 cm.
The frequency range is from 1450 kHz (left side of the screen) to 1550 kHz
(right side of the screen).
Horizontal: 10 kHz / cm vertical: 1 Volt / cm
1500 kHz 10 kHz / cm Distance: 20cm 
1500 kHz 10 kHz / cm Distance: 30 cm 
1500 kHz 10 kHz / cm Distance: 40cm 
1500 kHz 10 kHz / cm Distance: 50 cm 
1500 kHz 10 kHz / cm Distance: 60 cm 
In the next measurements the distance between the coils is each time 42 cm.
We see at the different frequencies, the coupling between the coils is about
critical.
So there is no need to change the distance between the coils when tuning the
receiver to another frequency.
Also we see the sensitivity of the receiver (height of the curve) is about
constant for the different frequencies.
The input voltage of the receiver is 0.1 Volt peakpeak, the voltage
across the detector circuit is about 5 Volt peakpeak, so we have a voltage gain
of 50 times.
Horizontal: 2 kHz / cm vertical: 1 Volt / cm
600 kHz 2 kHz / cm Distance: 42 cm 
900 kHz 2 kHz / cm Distance: 42 cm 
1200 kHz 2 kHz / cm Distance: 42 cm 
1500 kHz 2 khz / cm Distance: 42 cm 
Instead of increasing the distance, we can also reduce the coupling between the coils by increasing the angle between the coils towards 90 °.
Top view.
The angle between the coils is 90 ° here.

The distance (centre to centre) between the coils is in this case 14 cm.
For critical coupling between the coils, the angle had to be 88 ° at 600 kHz.
And 89 ° at 1500 kHz.
The response curve is comparable with the method described above (with large
distance between the coils).
The adjustment of the angle however must be very accurate (a fraction of a
degree).
Also at different frequencies, we need another angle, probably at such a small
distance between the coils the capacitive coupling also plays a part, and this
is depending on frequency.