roberts: Comparison_MSF60_DCF77

Martin Renz Ernst Erb Vincent de Franco Martin Bösch Mark Hippenstiel Bernhard Nagel Otmar Jung Heribert Jung Eilert Menke 
Cliquez sur le bouton d'information bleu pour en apprendre plus sur cette page.
Forum » Radios, TV, Appareils techniques (Physique) etc. » MODÈLES PRÉSENTÉS » roberts: Comparison_MSF60_DCF77
Karl-Heinz Bradtmöller
D  Articles: 374
Schémas: 61
Images: 11
24.May.17 13:02

Remerciements: 1
Répondre  |  Vous n'êtes pas connecté, mais en mode visiteur   1


on special request find english language version below:

However see for pictures and links german version (thus to avoid duplicating)


leads directly to german language version containing further diagrams and external links...

Dear friends of  Radiomuseum,
There is hardly anyone to whom the German DCF77 time signal transmitter is not a well-known term (77.5 kHz, located at Mainflingen near Frankfurt/Main, Germany).
However, even in the north of Germany, the English transmitter MSF60 (60 kHz, Anthorn Cumbria, formerly located at Rugby) can also be received with a reasonable level is probably not so well-known.
 (The search for keyword "MSF 60" provides only few answers.)
So I would like to try to close the information gap with the following text a little bit.
In short, the principle of the time signal transmissions, and what  - roughly sketched - markes the difference both transmitters from each other.
In the minute n, the time and date code is broadcast for the minute n+1.
For DCF77 (77.5 kHz) the carrier oscillation is reduced to 15% of the maximum level once per second (with the exception of second 59).
Where the start of the second’s blankings is always at a distance of seconds, the recovery of the transmitter amplitude to 100% is delayed by 100 or 200 milliseconds.
In the length of these blanking times the logic states are represented,
100 milliseconds for "zero", 200 milliseconds for "one", which can then be decoded by the radio clocks.
The MSF 60 (on 60 kHz) Scottish time signal transmitter does no more use  the "fast code" in second 1, simultaneously to the “slow code”,  thus the "fast code" has been totally replaced by the “slow code” since a couple of years ago. That “slow code” looks like the code transmitted by DCF77.
However, a carrier reduction of MSF 60 exists at almost 0%, and there are more blanking durations, in addition to the 100 and 200 milliseconds, namely 300 and 500 milliseconds. The 59th second blanking is also transmitted.
If you listen to this station for the first time, the "double pulses" in the time interval between second 1 and second 15 are particularly noticeable.
This makes the use of a DCF77 decoder for MSF 60 simply impossible, which can only interpret a carrier reduction within one second interval and only two time intervals, and also requires the extended "pause" for the minute synchronization, with the absence of the 59th minute blanking.
The question will be asked why it is so complicated with MSF 60.
Well, in reality, two independent codes, code A and
Code B are transmitted simultaneously.
In Code A, there are only 100 and 200 millisecond blankings,
in those there are hidden the "pure" time and date informations in the BCD code.
For code B, 100 and 200 milliseconds are equal to logic "zero" and 300 milliseconds equal to logic "one".
With code B the DUT1 code - which has no more been transmitted by DCF77 since the late seventies of the last century - is transmitted in seconds 1 to 15, which contains the information about relative deviations of the actual earth rotation speed time to the atomic frequency normal time.
Furthermore, the parity bits and the british summer time bit appear in the B code.
Since a series of fixed bits - always an equal pattern (01111110) - are transmitted on the end (seconds 52 to 59) in the A code and then the 500 millisecond blanking in minute zero to detect the start of the minute, one also could achieve without a parity bit check a reasonable time and date display on radio controlled clock.
This means in practice that one does not necessarily have to decode the B-code, if one waives the representation of DUT1 and parity check.

The diagrams show the blanking times in MSF 60
And the comparison between DCF77 and MSF 60

If the two timecodes of DCF77 and MSF 60 are compared with each other, one also recognizes that the BCD coding is reversed (LSB interchanged with MSB).
Also, the bit assignment of the information from year to minute is interchanged.
Then, at MSF 60, the parity bits are not spread over, but are sent sequentielly all together at the end in bits for seconds 54 to 57.
The sum of the bits from 17 to 24, 25 to 35, 36 to 38 and 39 to 51 in the parity bits 54 to 57 is supplemented to an odd number of ones, ie to "odd parity", contrary to the "even parity" on DCF77.

There is another special feature in the assignment of the Weekdays to sent digits.
Although also in the UK the calendar week starts on Monday with counter number 1,
the Sunday will not be the same as for DCF77 consequently with the number 7
but with the number 0 posted.

Acoustically obvious in MSF 60 is also the missing random number phase modulation, which is seen as "noise" during DCF77 reception, which only disappears completely during the blankings.


Pièces jointes:

Cet article a été édité 28.May.17 13:40 par Karl-Heinz Bradtmöller .

Michael Watterson
IRL  Articles: 964
Schémas: 649
Images: 2489
29.May.17 14:30

Remerciements: 1
Répondre  |  Vous n'êtes pas connecté, mais en mode visiteur   2

In Limerick, Ireland, the MSF signal eventually updates the time on a very cheaply made alarm clock. Reception is helped by orientating the set (separate internal ferrite rod to that used for MW).