DCF77 is a transmitter in the VLW band with a
frequency of 77.5 kHz at 50 kW, see
Wikipedia
for more details. It is located at Mainflingen in mid-west
Germany (near Hanau/Frankfurt) and transmits the
"official" time in Germany by
reducing its amplitude down to 15% at the start
of each second for a duration of either 100 or
200 milliseconds, by that encoding zeros and
ones bits,
a missing 59th bit that signals the end of a
minute and the start of the next amplitude reduction
as the begin of the next minute,
encoding time and date information as well as several
parity bits in those 59 bits.
With that information a clock can be synchronized to display
exact time and date.
2 Why a home-brewed receiver for DCF77?
Besides that it is fun to build an own receiver for those
Very-Long-Wave signals there are some rational reasons
to do that:
Contrary to commercially available DCF77 receiver
modules this receiver has a very high amplification
of the signal and so can receive the signal over a
very large distance, beyond the 2,000 km limit
that DCF77 has been designed for. Two amplification
regulators reduce the amplification level in the
near-field so that it works well in my 30 km
distance to DCF77 (with very small amplification)
as well as in the far-field (at a very high
amplification level). Even enclosed within an iron
cookie box the AM receiver still works fine.
Several LC resonators and a ceramic filter increase
the immunity to noise. Noise in this frequency range
comes from switching regulators (e.g. power supply
devices, switched lamps, etc.) that can have very
large amplitudes and can tamp any receiver that has
less selectivity. E.g. my energy-saving lamp in my
shag or the power supply of my laptop produces so
much noise that any commercially available DCF
receiver module in less than 2 m distance is
overwhelmed and produces nothing.
Just because you can do it by yourself: receiving
VLW waves via a ferrite rod antenna, decoupling
of this already very selective LC resonator with an
FET, preamplifying this signal in a regulated HF
amplifier, mixing the HF with a (77.5+455) =
532.5 kHz oscillator, selecting the 455 kHz
mixer result with a High-Performance IF filter,
feeding this into a high-gain regulated amplifier,
rectification of the resulting IF and to filter the
amplitude in an RC filter to detect the duration of
the second's amplitude reduction can be done within
a single chip and some external components.
But why use such an old IC TCA440 instead of more later ones?
Because more modern IC combine AM and FM in one chip,
but the FM part is useless here. The same with audio
amplifiers on board: useless for a DCF77 receiver.
Because more modern ICs only offer a very small part of
the overall amplification of a TCA440, they are designed
to cover only very large signal strengths with that low
amplification factors.
Other commercially available DCF77 receivers are of the
direct amplification type and filtering resonators
(of the ceramic or crystal type) for 77.5 kHz are
rarely available, if at all. By mixing the signal
High-Performance IF filters can be used to increase the
selectivity. And the TCA440 has anything on board that
is necessary for that.
To increase the selectivity of more modern AM receiver
ICs a large number of external components would be
necessary. The TCA440 has anything on board for that.
3 Schematic of the receiver
This is the electrical scheme of the receiver.
The VLW signals are received with a 10 cm ferrite rod
with enameled copper wire. The capacitors are to be selected
for the measured impedance of the rod (see below) and a
trim capacitor of 90 pF allows for adjusting the
antenna resonator for a maximum signal.
The N channel FET BF245 (any other N channel type can be used)
decouples the resonator and generates a symmetrical
(differential) signal for the HF input of the TCA440,
which is coupled to the input with two capacitors.
The HF amplifier of the TCA440 is connected with pin 16,
where an LC filter with a Germanium diode and a 15 k
resistor and a 100 µF electrolytic capacitor
produces a regulator voltage for the gain of the HF amplifier.
This regulator voltage controls the HF amplification in the
near-field. The LC filter is self-made on a Neosid coil body
05 9539 00 (9.5 µH).
The LC resonator on pins 4, 5 and 6 of the TCA440 generates
an oscillator signal of 532.5 kHz to be fed to the
mixer. This LC coil is also self-made and produces a stable
oscillator signal. As a potential alternative I present an
untested Xtal oscillator for replacing the coil with a
temperature stable oscillator
here.
The output of the mixer on pin 15 is fed into a commercial
SFT006H IF filter consisting of a combination of two LC
resonators and a 455 kHz ceramic resonator. The filtered
signal enters the IF amplifier on pin 12 of the TCA440.
IF amplifier output on pin 7 of the TCA440 is filtered
with a commercially available 455 kHz resonator and
rectified with a germanium diode. A RC filter network filters
the fast seconds pulses, to be fed to the positive input
of the CA3140 FET opamp, and averages those very slowly to
be fed to the negative input pin of CA3140. The slow average
voltage is also fed to the IF regulator voltage input on
pin 9 of the TCA440 and to the display driver input.
The display driver output voltage drives a 288 µV
instrument and can be read on an AD converter input pin of
the controller.
The output of the opamp CA3140 drives a transistorized-driven
LED and provides the seconds pulses for the controller
readout and DCF77 signal analysis. The output signal follows
the amplitude of DCF77: high amplitude = high, low amplitude
= low. A good controller software should be able to detect
the signal no matter if it is high- or low-active (like the
one described here).
All external components, including the power supply of
5 V, are tied to a 6-pin connector.
Simulation of the voltages
These are the voltages that occur on the capacitors of the
RC network when a logical zero and one occur in the DCF77
signal by decreasing the amplitude to 15%.
The voltage on the first capacitor, after rectification of
the IF, decreases and increases very fast. Its small
capacity of 1 µF has no relevant influence and
just smoothes down the 455&nbps;kHz peaks from the rectifier
diode.
The voltage on the fast integrator with R3 (2k2) and C2
(1µF) decreases relatively fast with the reduced
amplitude. It just buffers the signal during the time
when the rectifier does not load the capacitor.
The voltage on the slow integrator with R4 (39k) and C3
(220µF) remains uninfluenced by the second's
amplitude reduction (very small decrease). It takes many
seconds until the voltage reaches its average value.
The digital output signal on the CA3140 is very stable,
but very short spikes can occur from the 455 kHz
rectified signal. Those can easily be filtered by the
controller's software.
4 Mounting
4.1 Mounting the antenna
The VLW receiver antenna is build on a ferrite rod, where
a 0.25 mm copper wire is wounded on. Litz wire is
not necessary due to the low frequency.
On both ends the coil is fixed with tape. The antenna
is mounted in two plastic holders onto the PCB.
To determine the inductance of the antenna coil a capacitor
of 10 nF is added to form a resonator and a signal
generator (can be a digital one, such as described
here)
feeds its signal over a small capacitor such as 10 pF
into that resonator. Either the signal on the resonator is
viewed with an oscilloscope or the shown rectifier produces
a DC voltage from that, which can be measured by a usual
voltage meter. By that the resonance can be identified.
From the resonance frequency fres the inductance
of the coil can be calculated:
L(H) = 1 / 4 / Π2 / f2 / C.
The inductance shall be approximately 3.5 mH.
From the measured inductance the capacitor C, which is
necessary for a resonance frequency of 77.5 kHz,
can be calculated:
C(F) = 1 / 4 / Π2 / 775002 / L
4.2 Mounting the LC filters
The LC filters are mounted as follows.
4.2.1 HF preamplifier circuit
Primary coil 91 windings, secondary 45 windings, 0.1 mm
enameled copper wire on Neosid coil body 05 9539 00
(9.5 µH).
4.2.2 Oscillator circuit
Primary coil 30 plus 76 windings, secondary 13 windings,
0.1 mm enameled copper wire on the same Neosid coil
body.
4.2.3 IF filter
This filter is a commercially available TOKO SFT006H.
4.2.4 IF output filter
The output filter is a commercially available filter
7MC-A2549HM.
4.3 Mounting the PCB
The receiver can be build on a small PCB and wired with
0.25 mm solder-able enameled copper wire. The TOKO
filter does not fit into the 2.54 mm pattern and
larger holes have to be made to fit that into.
4.4 Adjustment
When adjusting the receiver's circuits the ferrite rod
should be directed to Mainflingen and this direction
is not to be changed during adjustment (because it has
a strong influence on signal strength's).
Note that adjusting is easy because none of the
adjustments affects other adjustments as well. So
bringing all to a maximum is easy. But also note
that the temperature also has a strong influence on
the result, so do not be disappointed if an evening
adjustment yields other result than an afternoon
adjustment.
If you have an oscilloscope start with it connected
to the source of the BF245. If not, connect a voltmeter
to pin 5 of the 6-pin connector. You can also
use a field strength instrument, if you have one
attached.
First adjust the trim capacitor parallel to the antenna
coil to a maximum signal strength. If you are very far
away from Mainflingen you might not be able to see the
signal yet on your oscilloscope. In that case use the
voltage measurement or the signal strength indicator
instead. If that does not work either come back to this
step later on, when the rest of the filters has been
adjusted.
If you own a frequency counter, such as the one described
here, you
can first adjust the oscillator frequency to 532.5 kHz
on pin 6 of the TCA440. If not, adjust to maximum.
Adjusting the other resonators goes all around and in
direction of the maximum.
5 Operation
The signal strength is varying mainly due to the
sensitivity of the ferrite rod antenna to directional
changes. When changing direction it might takes rather
long until the signal level stabilizes due to the
slow averaging. The LED can be used to identify a
successful direction change: a flickering signal on
that shows unsuccessful changes. But in any case leave
enough time for averaging.
Good luck with receiving DCF77.
Praise, error reports, scolding and spam please via the
comment page
to me.