DCF77 receiver => Interfering signals
Diese Seite in Deutsch:
AVR single chip controllers AT90S, ATtiny, ATmega and ATxmega
DCF77 interfering signals
9 Interfering signals on DCF77's frequency
There are many opportunities why DCF77's signal can be masked by noise:
- Self-generated interference signals: the microcontroller near the reception
coil generates rectangles, which are themselves or of which their harmonics are
on or very near to the reception frequency.
- Other electronic equipment, such as laptops, power supplies or current-reduced
lamps generate an interfering signal on or near to 77.5 kHz.
9.1 DCF77 interferences by PWMs with the ATtiny25/45/85
When experimenting with DCF77 receivers, that used PWM channels, I realized
interferences. Those can be so strong that the whole reception of DCF77 fails.
The following PWM generators can interfere with DCF77:
Other PWMs, such as the signal former of the 110kHz-oscillator, do not interfere
because their base frequency is already far above 77.5 kHz.
- the AM rectifier with an ATtiny25, that reports the strength of the received
signal to the decoder controller by using a PWM, or
- from the DCF77 controller, that uses two PWM channels to adjust the
frequency and the gain.
Both PWM generators identified above work with PWM frequencies of 31.25 kHz.
Because rectangles of this frequency consist of the ground wave and all odd
harmonics (with decreasing amplitude), also their harmonics are interesting.
I have measured the potential interfering signals and played with diverse
properties of the PWM generators. Here are the results.
9.1.1 Synchroneous- and asynchroneous modes with the ATtiny25/45/85
TC1 is, unlike all other ATtiny, not a 16-bit but an 8-bit-TC. It can be clocked
from the system clock with a prescaler divider. Other than in all other TCs, it
allows prescaling not only by 1, 8, 64, 256 and 1,024, but with all powers of two:
1, 2, 4, 8, ... , 16.384. Already this makes the ATtiny25 to a mighty tool.
Clocking the TC1 with the system clock and by the prescaler value is called
The ATtiny25/45/85 has additionally an asynchronous mode. If you switch that on
then the TC1' clock source is an on-board 64-MHz-oscillator (alternatively: a
32-MHz- or an 25.6-MHz-oscillator (in ATtiny15 compatibility mode). With those
oscillators PWMs at higher frequencies can be ran.
This is a bit confusing, because there a are sooo many opportunities available
to clock the TC1-PWM. I have setup a LibreOffice calculation sheet that allows
to play with all those modes and properties for download
In sync mode the TC1 clock depends from the system clock, which can either be
generated from the internal RC oscillator, from an external crystal or resonator
or from an external crystal oscillator. I have integrated all commercially
available frequencies in those drop-down lists, so you can fit it exactly to
your needs. In sync mode the PWM clock can also be prescaled with all named
multiples of two. The port register OCR1C can be set to any value between
0 and 255 to select the PWM resolution. This third port register allows to
use both channels, OC1A and OC1B, as PWMs.
The latter is also the case in async mode. In async additional
High-Speed-Oscillators can be used as clock source, as described above.
As in all cases used here 8-bit-PWMs are required, the following is limited
to this case.
9.1.2 Initing the modes
For the experiments I have build a test generator, which consists of an ATtiny45
and a two-stage RC network of 10kΩ and 1µF on the OC1A- and
OC1B output pins. In the standard experiment I have clocked the ATtiny with
8 Mhz (Clock source: internal RC oscillator 8 MHz, CLKPR to 1). Both
PWM channels were inited in synchronous mode with a prescaler of 1 and a
PWM resolution of 256. As PWM value I used one half, because this produces
the largest noise. With that the PWM frequency is 8 MHz / 256 = 31.25 kHz.
The following sub-chapters show how those modes of the ATtiny25/45/85 are
set-up in assembler.
18.104.22.168 Setting the sync mode
In sync mode the clock signal stems from the system clock. In the standard case
the system clock comes from the internal oscillator with 8 MHz, that can be
divided by the CLKPR precaler (1 .. 128). This does the following code sequence:
The synch mode for the two PWM channels is inited with this sequence:
; Init clock prescaler
ldi rmp,1<<CLKPCE ; Enable prescaler change
ldi rmp,(cClkPsr-1) ; Load prescaler
I added the interrupt to enable a countdown of a 16-bit counter value that
blinks the green LED attached to PB2. The counter value reflects the speed
of the PWM, so blinks in the same rhythm under all PWM speeds.
; Init sync mode
ldi rmp,cTestPwmA ; PWM start A
out OCR1A,rmp ; to channel A
ldi rmp,cTestPwmB ; PWM start B
out OCR1B,rmp ; to channel B
ldi rmp,255 ; Set 8 bit PWM
ldi rmp,(1<<PWM1B)|(1<<COM1B1) ; Positive PWM
ldi rmp,(1<<PWM1A)|(1<<COM1A1)|(1<<CS10) ; Prescaler to 1
ldi rmp,1<<OCIE1A ; Enable interrupts
22.214.171.124 Selecting async mode
The initiation of the async mode is more trickier. First the
high-speed-oscillator has to be started by setting the PLLE bit
in the PLL control register PLLSCR. The bit LSM forces the
frequency low to 32 MHz (LSM=1).
Then you'll have to wait until the PLL has reached lock state. This can be
seen from the lock bit PLOCK. If this bit is set, the PLL is ready
and the modes of the two PWM channels can be set.
This is the init procedure in async mode:
Tc1Presc is the prescaler between 1 and 15, in our case 4.
ldi rmp,(1<<LSM)|(1<<PLLE) ; Switch PLL on, Low-Speed-Mode
WtLock: ; wait until PLL is locked
in rmp,PLLCSR ; Read lock bit
sbrs rmp,PLOCK ; wait for lock bit set
rjmp WtLock ; not yet set
ldi rmp,(1<<LSM)|(1<<PCKE)|(1<<PLLE) ; switch PCK on
ldi rmp,(1<<PWM1B)|(1<<COM1B1) ; Enable Compare Match B PWM
ldi rmp,(1<<PWM1A)|(1<<COM1A1)|cTc1Presc ; PWM1A-en, Prescaler,
out TCCR1,rmp ; OC0A aktive, clear on compare match
9.1.3 Software for seeting modes
In this assembler source code
all modes and prescaler values can be adjusted. By changing the .include
line, ATtiny25, 45 or 85 can be selected, all other properties can be
changed in the section Configuration.
9.1.4 Measuring the interferences of the PWM
To measure the interference by the PWM I have used a VLF receiver that I had
build in the Eighties. It is equipped as follows:
shows the displayed characteristics of the receiver and the frequency counter in the
- Ferrite antenna 20 cm long, with enamed copper wire over the whole length,
- Variable capacitor with 500+365 pF, parallel,
- Direct receiver with thre transistors,
- HF rectangle output for frequency counter,
- Reception range: standard ca. 60 to 210 kHz, with additional capacitors
down to ca. 32 kHz.
Like to be expected, the third harmonic of the base frequency shows up at ca.
96 kHz with a high amplitude. An unmodulated carrier signal at 129 kHz
stems not from the PWM generator. The fifth harmonic has not enough amplitude,
shows up with a small resonance and has no interference potential.
Another hint: switch off your laptop after you programmed the ATtiny and before
switching the receiver on. Such a laptop produces serious interferences, as
discussed in Chapter 9.3.
Conclusion: better select your PWM frequencies with the harmonic list, as provided
in the calculation sheet. Do not operate PWMs with 75 to 80 kHz or their third
or fifth harmonic at that frequency band.
9.2 Interference by an energie-conserving lamp
The energy-conserving lamp on my tinker shelf produces a strong interfering signal
between 40 and 220 kHz. DCF77, even though very near, is only realized with
a very small signal on top of this noise, but definitely cannot be decoded by any
Conclusion: Such an interferer kills each experiment with DCF77.
9.3 Interference by my laptop
To identify signals from my laptop, I placed the receiver near to my Lenovo T430S.
The laptop produces a medium random noise spectrum between 40 and 200 o;kHz.
At 70 kHz an unmodulated carrier can be heard. The DCF77 signal can be identified
clearly, but with a relatively high random noise background. If I would be in a
larger distance to DCF77, I wouldn't be able to hear that signal.
Conclusion: Better switch off the laptop or send him to sleep before experimenting
with DCF77, or place it in a larger distance.
9.4 Interferences with the laptop's power supply
This power supply device is a phenomena: it produces a very strong signal over the
whole band from 40 to 220 kHz. At 215 kHZ a maximum of interference is
seen. The relatively strong DCF77 signal cannot be discrimated from the strong noise,
no reception signs can be seen. The distance over which this spectrum spreads is
relatively short, in 1 m distance anything is fine.
Conclusion: Far, far, away.
©2020 by http://www.avr-asm-tutorial.net