[time-nuts] Droitwich Frequency Error

[Kasper Pedersen]

On 18.10.2020 16.07, Andy Talbot wrote:
> 
> This morning, mainly to prove something it to myself I made a plot of the
> off-air UK Droitwich LF transmitter whose carrier is supposed to be a
> national standard - although I believe it uses a Rb source that is
> periodically updated with a Cs one - manually.
> 

The stability of UK Droitwich has had its good times and its bad times.
In 2008 I did a similar measurement, and got behaviour like this (from
Silkeborg, Denmark):

https://n1.taur.dk/timenuts/droit.png
https://n1.taur.dk/timenuts/droit2a.png

(apology for unlabeled axes, Y is offset in seconds, X i sampletime in
seconds)

Compared to those, your plot is very well behaved, and I recall talking
to a retired BBC engineer who had the company they had outsourced the
operation to in rather low regard. So performance has improved greatly.

The phase change of about 20 degrees is around 280ns - that's really
good. The 680km to DCF77 goes up and down by about a microsecond over
the same period. On the other hand, unless your antenna and amplifiers
are really broadband, the phase change could come from temperature drift
in your antenna/filers/amplifiers/.. The way you test for that is to
build a little test transmitter - maybe another DDS clocked from the
same reference.

Considering that you are comparing an error in microseconds to the
GPSDO, unless the GPSDO is unlocked or broken, the error from it will
not be significant. (this is also where you realise that you need at
least two more clocks:-)

Careful, building coherent VLF receivers is a gateway drug.

/Kasper Pedersen
OZ2KKP