[time-nuts] Thunderbolt versus Home made
not.again at btinternet.com
Sat Feb 24 19:10:18 EST 2007
On Sat, 17 Feb 2007 08:50:56 -0600, you wrote:
>> On Fri, 16 Feb 2007 17:44:14 -0600, you wrote:
>>> Richard W. Solomon wrote:
>>>> I looked at the ACE-III on Trimble and do not see mention of a 10 KHz output.
>>>> Maybe I miised it, but without it, the N1JEZ board will not do unless you modify
>>>> it for 1 pps.
>>>> I am trying to find more info, but since it reached EOL, Trimble may have deleted
>>>> 73, Dick, W1KSZ
>>> You are correct, you need to patch another set of dividers in the chain.
>>> Keep in mind that the only benefit of a 10 kHz output on the Jupiter is
>>> to save 4 decade dividers. You should not use the fact that the Jupiter
>>> outputs 10 kHz to speed up the loop in a GPSDO. The loop bandwidth is a
>>> function of the crossover point between the VCXO and the GPS, which in
>>> most cases will be between many minutes (cheap VCXO) and a few hours (HP
>>> 10811). Whether you use 1 Hz or 10 kHz comparison frequency into the
>>> phase detector will not affect the filter.
>> (As this is my first post here, I'll try not to mess up too badly :)
>> I think that for a simple controller like the N1JEZ board, a 10KHz
>> output on the GPS really is needed.
>> I tried a circuit like that with 100pps once out of curiosity, but as
>> I remember, the lower error sensitivity made things more difficult.
>> BTW, with a standard GPSDO, would using the 100Hz/10KHz the pulse not
>> reduce the need for a fast comparator clock, and also reduce sawtooth
>One of the problems with the 10 kHz output us that it is only updated
>once per second, i.e. the period of the 10 kHz signal is the same for 1
>second, and every second, there is a small phase jump to update the
>phase of the 10 kHz to match the 1 PPS.
I don't know whether the outputs on a Jupiter exhibit any granularity
or not (I've not seen anything to suggest that they do), but I think
that with receivers that do, changing from 1PPS to a faster PPS output
would normally change the granularity effects that are seen - assuming
things are not in sync.
As for possibly reducing the need for a fast clock; being able to
measure a number of pulses in a second and then take the average count
can allow a rather better resolution than just taking one measurement
- so in some circumstances this could be useful.
I've not really looked into it much, but was really just interrested
if anyone here had come accross the 100/1K/10K, etc pulses being used
>That means the 10 kHz has a one Hz component that is the correction
>signal. It must be filtered out. But even filtering the 1 Hz component
>is not sufficient, since the 1 Hz component has the same short term
>noise that is present on the 1 PPS output of any GPS receiver.
>So while it's a lot easier to filter 10kHz down to a level where 10 kHz
>ripple does not appear on the OCXO EFC input, such a filter would be
>grossly insufficient to make a GPSDO. You would simply have an OCXO
>phase locked to a poor reference.
Although that's not what I was talking about doing above, I think that
it's pretty much what some of the hardware GPSDO's actually do using
various types of oscillators (and to good effect too). As in these:
the loop filter is basically just an RC circuit. It's not quite raw
GPS, but is quite different to using a digital filter as used in a
Thunderbolt, Shera, etc., which can run to hours.
When the N1JEZ board was mentioned, I assumed that it was this type of
controller that was meant - and I think it does need 10KHz unless it
gets much more of a mod than a couple of counters.
>You must filter the output signal from
>the phase comparator so that the short term instabilities in the GPS
>timing signal are also filtered out.
>The 10 kHz is useful for experimentation, since you could artificially
>speed up the loop with one second or so time constant or even less so
>that you could easily verify with an oscilloscope that the OCXO is
>actually phase locked to the GPS, but you would not want to run the
>system that way because it would have terrible performance.
>With a time constant of 20 minutes to an hour, it can be very
>frustrating to verify that the system phase locks and it would be best
>to have another hobby to attend to in the mean time :-)
>Regarding the sawtooth effect, since the 10 kHz is only updated every
>second, it has no effect on the possibility of hanging bridges, and I am
>not sure what you mean by reducing the need for a fast comparator clock.
>time-nuts mailing list
>time-nuts at febo.com
More information about the time-nuts