[time-nuts] GPS Modulation and 10 MHz Delay Lock (Jim Lux)

[Jim Lux]

On 10/2/12 8:40 PM, johncroos at aol.com wrote:
> Hi Jim -
>
> Thanks for the update on the modern GPS receivers. I was aware that the
> modern ones
> do not have a classical analog tracking loop, much less a bunch of them.
> However it is a useful concept for purposes of explanation that you do
> not need the 1 pps to
> lock up the 10 MHz VCXO - which was my main point.
>
> The Tbolt block diagram in the manual Figure 5-10 shows the 10 MHz VCXO
> output going to
> the receiver and also to the output. The 1 pps comes from the "cpu and
> support" circuit.

yes.. the Tbolt is an example of a receiver that shifts the frequency of 
the reference so that it can generate an "on the mark" 1pps (and 
generate an accurate 10 MHz as well...)


>
> - On jamming - maybe so, but the effect of the receiving correlator is
> to spread the energy of a CW interferer
> and concentrate the energy of the signal with the matching PN modulation
> is it not????

Yes, if the receiver is linear (e.g. say you do a sliding code 
correlator and slide until you get the peak, with the correlator using a 
multiplier)...

But since most (inexpensive) receivers have hard limiters/1 bit 
quantizers in front of the correlator, what goes into the correlator is 
a square wave at the jammer frequency, and you can slide your PN code 
all day and not get a peak.

There's some analysis out there that tells you how many bits you need 
for a given Jammer/Signal ratio, but in general, if the jammer is 20 dB 
over the signal, you need 3-4 bits.

Back in the day, when bits were very expensive, that's why narrow band 
excisers were popular.. basically you'd run something like a PLL to 
recover the tone jammer, and subtract it out.