> Yeah, IIRC it is carrier off/on. But I was figuring that one
> could detect
>
>Alright, enough is enough. ;-) The carrier is never turned off,
>it is reduced in power by 10dB at the start of each second. It
>is then restored to full power at either 200mS, 500mS or 800mS later.
>This conveys a binary 0, a binary 1 or a position marker respectively.
Hmm, it seems this varies depending on the transmitter. Just browsing some
info that says the DCF77 in Germany reduces to 25% at beginning of each
period for 100mS to signal a binary 0, and 200mS to signal binary 1. For MSF
in the UK the carrier is indeed switched off at the beginning of each second
for 100mS to signal a 0, and 200mS to signal a 1. For WWVB in the USA it is
reduced by 10dB as you stated, and for JJY in Japan it is reduced to 10%
with timing methods different again to those above.
The document I have from MSF also indicates the carrier is switched off
So one station does switch carrier off, giving 100% AM modulation, three
others do not with varying percentages of AM modulation. The time slices
also vary, with MSF giving 100mS, 200mS 300mS and 500mS carrier off times.
The MSF document is available at http://www.npl.co.uk/npl/ctm/index.html go
to "MSF 60kHz Signal" on LH menu, and then scroll about half way down the
page for the PDF document. Also has a number of other interesting looking
bits in the three items under the MSF one of the menu.
The Useful Links also looks good. Includes one for the International Earth
Rotation Service - now I thought the earth rotated itself without needing
any servicing, but you live and learn ... ;)))))
Alan B. Pearce wrote:
> I wrote:
>>> Yeah, IIRC it is carrier off/on. But I was figuring that one
>>> could detect
>>
>> Alright, enough is enough. ;-) The carrier is never turned off,
>> it is reduced in power by 10dB at the start of each second. It
>> is then restored to full power at either 200mS, 500mS or 800mS
>> later. This conveys a binary 0, a binary 1 or a position marker
>> respectively.
>
> Hmm, it seems this varies depending on the transmitter. Just
> browsing some info that says the DCF77 in Germany reduces to 25% at
> beginning of each period for 100mS to signal a binary 0, and 200mS
> to signal binary 1. For MSF in the UK the carrier is indeed switched
> off at the beginning of each second for 100mS to signal a 0, and
> 200mS to signal a 1. For WWVB in the USA it is reduced by 10dB as
> you stated, and for JJY in Japan it is reduced to 10%
WWVB was the only one that I was referring to. Until you posted
this, I knew nothing about the other transmitters. Thanks for the
info. :-)
> with timing methods different again to those above.
>
> The document I have from MSF also indicates the carrier is switched
> off
That sure seems extreme. Since the carrier is the locked to the time
base, again I wonder how they key and unkey the transmitter without
causing phase/frequency distortion/shift? That is unless they don't
guarantee the accuracy of the first couple of cycles.
> So one station does switch carrier off, giving 100% AM modulation,
> three others do not with varying percentages of AM modulation. The
> time slices also vary, with MSF giving 100mS, 200mS 300mS and 500mS
> carrier off times.
>
> The MSF document is available at
> http://www.npl.co.uk/npl/ctm/index.html go to "MSF 60kHz Signal" on
> LH menu, and then scroll about half way down the page for the PDF
> document. Also has a number of other interesting looking bits in the
> three items under the MSF one of the menu.
Sure enough, they turn it on and off. Seems like that would make it
very difficult to use for a PLL reference.
BTW, I was wrongly thinking that MSF was some kind of generic
worldwide acronym for the time systems. Funny, it's a callsign with
M being a UK prefix and the SF randomly allocated making it very
specific to the UK system.
> The Useful Links also looks good. Includes one for the International
> Earth Rotation Service - now I thought the earth rotated itself
> without needing any servicing, but you live and learn ... ;)))))
I'm sure it won't be long before someone (make that some govt)
decides that it does need servicing and promptly sets about devising
a method to "correct" it.
>> The document I have from MSF also indicates the carrier is switched
>> off
>
> That sure seems extreme. Since the carrier is the locked to the time
>base, again I wonder how they key and unkey the transmitter without
>causing phase/frequency distortion/shift? That is unless they don't
>guarantee the accuracy of the first couple of cycles.
Well it is a time standard, not a frequency standard, so I do not see why
the phase/frequency etc distortion should matter. They do state in the
document that the rise/fall times at the transitions are dependant on the
transmitter, and I would guess they would have to tailor those to minimise
unwanted sidebands anyway. Same must happen on any of the others as well to
some degree.
Alan B. Pearce wrote:
>> That sure seems extreme. Since the carrier is the locked to the time
>> base, again I wonder how they key and unkey the transmitter without
>> causing phase/frequency distortion/shift? That is unless they don't
>> guarantee the accuracy of the first couple of cycles.
>
> Well it is a time standard, not a frequency standard,
I don't know anything about MFS, but all the WWV stations, including WWVB,
are also intended as frequency standards. The carrier frequency is derived
from the same atomic clock that the time is. The WWVx carriers are highly
accurate and traceable frequency standards. The only problem is that there
can be significant phase distortion due to multi-path propagation,
especially at the higher frequencies. If you live close enough to reliably
receive the ground wave, these carriers are the most accurate frequency
standards you likely have access to.
*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com
> michael brown
> Sent: 06 April 2005 12:33
> To: Microcontroller discussion list - Public.
> Subject: Re: [EE] Using clock mechanisms (was High sleep current on
> Sure enough, they turn it on and off. Seems like that would make it
> very difficult to use for a PLL reference.
Not if you use a very long time constant and only let the PLL work when
there is a carrier to work on... The bit of an anologue TV which detects the
colour sub-carrier works on a very short bit of subcarrier (colour burst - I
think it's about 10 cycles but it might be more than that) just after the
line sync pulse. The "burst locked oscilator" is good enough to keep in
phase with the original colour burst for the duration of the picture line.
> BTW, I was wrongly thinking that MSF was some kind of generic
> worldwide acronym for the time systems. Funny, it's a callsign with
> M being a UK prefix and the SF randomly allocated making it very
> specific to the UK system.
I always took the SF to stand for 'Standard Frequency'. I doubt it was
random.
> Alan B. Pearce wrote:
>
>>> That sure seems extreme. Since the carrier is the locked to the time
>>> base, again I wonder how they key and unkey the transmitter without
>>> causing phase/frequency distortion/shift? That is unless they don't
>>> guarantee the accuracy of the first couple of cycles.
>>
>>
>> Well it is a time standard, not a frequency standard,
>
>
> I don't know anything about MFS, but all the WWV stations, including WWVB,
> are also intended as frequency standards. The carrier frequency is derived
> from the same atomic clock that the time is. The WWVx carriers are highly
> accurate and traceable frequency standards. The only problem is that there
> can be significant phase distortion due to multi-path propagation,
> especially at the higher frequencies. If you live close enough to reliably
> receive the ground wave, these carriers are the most accurate frequency
> standards you likely have access to.
IF you live close enough. A GPS disciplined frequency standard
will have far better accuracy since the 'reference' signal
is available all the time rather than the limited hours of VLF
signals at long distances. All digital PCS cellular systems
use GPS for synchronization between nodes. Your cellular phone
has a VERY accurate crystal timebase because it needs it.
Now THERE is a great use for old cell phones.
www.absoluteastronomy.com/encyclopedia/R/Ra/Radio_clock.htm
"Radio clocks synchronized to terrestrial time signals can usually
achieve an accuracy of around 1 millisecond relative to the time
standard, generally limited by uncertainties and variability in
radio propagation."
On Wed, 2005-04-06 at 14:12 -0600, Robert Rolf wrote:
> All digital PCS cellular systems
> use GPS for synchronization between nodes.
Are you sure about that? While I know that CDMA based technologies have
GPS antennas at each site, I've never seen GSM, TDMA or iDEN sites with
those antennas, and as I understand it GPS is not necessary with those
techniques.
