Peter
Peter Martinez wrote:
Stuart:
The design of an MSF receiver, intended to recover the carrier
frequency to any great accuracy, needs to be able to reject the
existing 'unwanted' signals either side of the carrier. These are only
about 15dB down and at +/-1Hz. I am referring to the 1 sec timing
pulses (gaps).
Building a narrow band filter is only one way of doing this.
Some receivers - for example the Decca MSF design - use a coherent
technique.
They know when the is due pulse and put the phase comparitor into
holdover for the
duration of the pulse.
Indeed if you are trying to get time and frequency out of the same
receiver you have
to open up the bandwidth enough to resolve the pulses, and then the coherent
mechanism is probably more attractive than the ultra-narrow filter method.
The disturbance from this 1Hz source has to be attenuated by a very
large amount if it isn't to affect the output of the frequency
standard. If we suppose that MSF receivers in service can achieve
this, they shouldn't suffer any problem from unwanted products which
are somewhat further away, at 6 and 8Hz, and are already a long way
down. I cannot detect them here which probably means they are more
than 70dB down.
The issue that I have with this is trying to figure out is how to
convert the
voltage and amplitude of the interference into a time error. 70db is
1e-7 power and
only 3e-3 voltage. This seems big compared to 5e-11.
This clearly needs some more thought.
I know it would be nice to be rid of the LORAN interference, but this
line of reasoning may not be a means to that end, and in any case it
wouldn't help amateurs close to the other LORAN transmitters.
I was certainly told by NPL at ITSF 2005 that they did not know whether
there
was an effect or not, and that there was some concern that users near the
LORAN transmitter might not be getting the expected accuracy from the MSF
service. Hence my line of reasoning.
I understand the Rugby site must close in a year or so and they are
looking for an alternative site for the 60kHz MSF transmitter. The
question becomes whether the LORAN transmitter will also move to that
new site or whether it will close down. The long term future of LORAN
has always seemed shakey but it doesn't look as if it will be
killed-off until there is a European second-source for satellite-based
position-finding, and that seems some years away yet.
I am not sure what is going on in Navigation circles, but in timing
circles LORAN
is being sold in the US (by the LORAN providers) as an alternative
source of
time reference/frequency reference.
Contra to the US position, European service providers do not want to rely
off-air frequency service for reference - they are very vocal on this. The
telecomms industry is therefore persuing: synchronous ethernet, IEEE-1588,
timing over packet and possibly NTPv5 to address this and the signs are
that we can get to the required 10ppb using packet network technology.
One can but hope that the absence of interest in LORAN by the
telecommunications industry in Europe will limit the economic viability
of LORAN over here and assist its early demise.
73
Stewart G3YSX
73
Peter G3PLX
The accuracy of MSF is 2 parts in 10E12, so systematic interference
even a long long way down
may well have an effect. It will depend on the design of the
receiver and whether it has a front
end filter that has high rejection at +6Hz, -8Hz, but prior to the
LORAN transmitter.
|
