FREQUENCY STANDARD FOR ALLOUS 162KHZ
(MODIFIED DROITWICH STANDARD)
(2012)

KLIK HIER VOOR DE NEDERLANDSE VERSIE


Frequency standard locked at the transmitter Allouis on 162 kHz.
Aand the radio receiver for 162 kHz, a modified medium wave receiver.

Modification of the Droitwich frequency standard for Allouis 162 kHz.
The frequency standard locked at the transmitter Droitwich (BBC on 198 kHz) can be modified very easily for use with the French transmitter Allouis on 162 kHz. We only have to change the wiring between the various divider ic's. And of course, the receiver has to be tuned to 162 kHz instead of 198 kHz. The transmitter Allouis can also be received in Southern Europe. For nothing we can lock our frequency standard at this very accurate signal!
It is not necessary to make a special receiver, a normal longwave broadcast radio can be used that is tuned on 162 kHz.


Location of the long wave transmitter Allouis.
It can also be received in Southern Europe.

The differences between the block diagrams
The only difference is the dividers, but it are the same 2 and 5 divider ic's as in the circuit diagram for Droitwich. We only have to change the wiring between the various divider ic's.
The 1 kHz output can be changed into 10 kHz or 1 MHz or other values. I needed the 20 kHz output, it is used for the calibration of the QRSS beacon receiver.


Block diagram of the original frequency standard for Droitwich.


Block diagram of the modified frequency standard for Allouis.
Only the dividers are different, but it are the same ic's.

How does it work?
The signal on 162 kHz is received with an ordinary broadcast receiver.
The 10 MHz crystal oscillator is divided by 125 and by 5000. From the signal after the 125 divider (80 kHz), the 2nd harmonic is used. Now we have a 160 kHz and a 2 kHz signal. The 160 kHz signal is injected into the ferrite antenna of the broadcast radio. Together with the 160 kHz signal, the 162 kHz signal of Allouis will cause an interference tone of 2 kHz in the loudspeaker. This audio tone of 2 kHz is compared in a phase detector to the 2 kHz signal from the divider. When both 2 kHz signals are not exactly the same, the frequency of the 10 MHz oscillator is controlled.
The 2 kHz audio signal is filtered to suppress speech and music of the broadcast transmitter.
I did need an accurate 20 kHz signal for the calibration of my QRSS beacon reception receiver on 10140 kHz and an accurate 1 kHz signal. But you can change the 1 kHz signal to 10 kHz, 200 kHz or 1 MHz and even 5 MHz. And also 10 MHz if you do use the remaining NAND as a buffer. But of course you can choose your own desired frequencies from the outputs of the various dividers.


Longe wave broadcast receivers that can be used
for the reception of Allouis 162 kHz and Droitwich 198 kHz.

Calibration of external 10 MHz oscillators
To calibrate external 10 MHz oscillators, the switch S1 is set to the other position and the internal 10 MHz oscillator is replaced by the external 10 MHz signal. This oscillator is not controlled of course, we have to adjust it ourselfs. On the screen of an oscilloscope the filtered 2 kHz audio signal from the broadcast receiver is displayed. The oscilloscope is triggered by the 1 kHz signal from the dividers. The only thing we have to do now is to adjust the 10 MHz oscillator so that the 2 kHz sine of the audio signal does not move to the left or to the right on the screen. When the sine moves one period in 1 minute, then the error of the 10 MHz oscillator is approximately 1 Hz.


Calibration with an oscilloscope. The audio signal on the screen does
not move to the left or right when the frequency is exactly 10 MHz.
Triggering is done by the 1 kHz signal on channel 2.

Phase modulation of data signals
Unfortunately, we can not use the 162 kHz signal without any problem. It has phase modulation for the transmission of data. You can see that on the screen of the oscilloscope here below. The sine has jitter. This phase modulation has to be suppressed in the loopfilter of the frequency control, otherwise it will disturb the accuracy of the 10 MHz signal of the frequency standard. This can be done by making the loopfilter very slow. By doing so, the control loop will become also less sensitive for radio interference. The loopfilter in the circuit has two positions: a fast one to lock the oscillator and a slow one for when the crystal oscillator is locked.


Phase modulated data signals have to be filtered out
in the loopfilter by making it very slow!

The diagram
Right above, you can find the 10 MHz crystal oscillator. With S1 you can select the internal oscillator that is controlled or for an external 10 MHz oscillator that has to be calibrated (adjusted). The internal 10 MHz oscillator is controlled by a varicap BB212 of which only one half is used. Both 74HCT390 are the frequency dividers, it are /2 and /5 dividers.
The 160 kHz signal is injected into the receiver with a short piece of wire that is mounted close to the ferrite antenna. So not a small loop around or close to the ferrite antenna, but capacitive coupling with a piece of wire, that does work much better. Play a little with the length and place of the wire and the value of the 10k resistor. I had to change the 10 ohm into 2200 ohm, as more signal was required than for Droitwich. The 2nd harmonic of 80 kHz does have a lower level than the signal of 200 kHz in the Droitwich diagram.

Other standard frequencies
It is very simple to change the 1 kHz standard frequency to other values. For that, we need th connect pin 1 of U3a with another divider as given in the diagram:

Wanted standard frequency Connect pin of U3a with
1 kHz pin 13 of U3b
10 kHz pin 13 of U2b
200 kHz pin 9 of U2b
1 MHz pin 7 of U2a


Diagram
The wiring has been changed an the 10k ohm resistor changed to 2200 ohm.
big diagram

The 2 kHz audio signal of the receiver (headphones output) is connected to "lf in" and filtered in 2 active filters with a LM358. What remains is a good useable 2 kHz signal without interference of music and speech. Both filters should have a Q factor of 10 and a gain of 2x2=4x. With the 500 ohm potentiometers, they are exactly adjusted to 2 kHz. The filtered audio signal can be monitored with an oscilloscope by connecting it to "2 kHz audio".

The NAND U1C does make a square wave of the 2 kHz audio signal. U1B is the phase detector. The average value of the output voltage of this phase detector varies between 2.5 V and 5.0 V, perfect to control the varicap. Phase modulation and the 2 kHz square wave are filtered out in the loopfilter.

The loopfilter has two capacitors, one of 10 uF and one of 100 uF. The 10 uF capacitor removes the 2 kHz signal. With S2, the time constant of the filter can be selected for "locking" or "locked" or fast or slow. The loopfilter is too slow with the 1M ohm resistor to get the oscillator locked. When the oscillator is locked, the switch is set to the position "locked" and the 10 MHz oscillator is controlled very slowly. Phase modulation, the 2 kHz square wave and interference do have hardly any influence anymore on the stability in this position of the switch.
The stability of the control loop can be adjusted with the 10k potentiometer in series with the 100 uF capacitor. At 0 ohm, it will take much time before the circuit is stabilized, the frequency goes up and down very long. Also after interferences, stabilization takes a long time. With the potentiometer adjusted to 5k ohm, stabilization goes much quicker.
With the led you can check if the circuit is locked. If not, then the led goes slowly on and off with the frequency difference between both 2 kHz signals.
One NAND is not used, you can use that for example as a buffer after the crystal oscillator when you do need a 10 MHz signal.


Instead of a long wave broadcast receiver, a modified
medium wave receiver is used for the reception of Allouis 162 kHz.

Modified medium wave receiver
The frequency standard worked perfect with the portable radio. But on a flea market, I did buy a small medium wave radio for only 0.50 Euro. It was modified for the reception of Allouis on 162 kHz by placing capacitors in parallel with the two sections of the tuning capacitor. To the section for the ferrite antenna, it was 1650 pF (1500 pF + 150 pF) and the oscillator section 150 pF. The 3 volt supply is taken from the 5 volt stabilizer of the frequency standard via two diodes to lower the voltage. The radio is permanently connected with the frequency standard by means of a 4 wire screened cable with a length of 1 meter. Supply, audio and the 160 kHz injection signal go through this cable. The loudspeaker can be switched on or off with a switch.

Results
A good impression of the accuracy can be obtained by comparing the 2 kHz audio signal with the 2 kHz (or 1 kHz) signal from the dividers with an oscilloscope. After that the control loop has stabilized, the variations during a few seconds are approximately 1/50 period. Or, the 2 kHz sine moves 1/50 period to the left or to the right during a few seconds. This is difficult to measure because of the phase modulation. Variations are caused by instability of the 10 MHz oscillator, interference (fading) in the reception and modulation of the transmitter with speech and music. So the frequency error of the 10 MHz oscillator is less than 1 Hz. For us, radio amateurs, more than sufficient!


Inside view

BACK TO INDEX PA2OHH