[time-nuts] HP 5061Cs reference question
magnus at rubidium.dyndns.org
Sat Dec 6 11:22:08 EST 2014
On 12/06/2014 04:51 PM, paul swed wrote:
> Great but I am looking for very specific detail. If you pick a wrong peak
> especially if you can't see peaks on a very weak tube then I think that
> translates into an actual offset.
Yes, there will be a significant offset unless you adjust your
synthesizer frequency accordingly, which few clocks allow you to do.
Then, if you have a weak tube, then the center pidestal is the most
likely to lock onto.
> If thats a true statement. Given the modern GPS boxes we have today that
> are stable. Wouldn't you see that as a always constant drift. Say 7-10ns
> over 20 minutes?
It would be significant yes, and it depends on the C-field, but in
general, the C-field is so strong as it separates of the -3, -2, -1, +1,
+2 and +3 peaks from the central 0 peak so that it can be observed
without interference. The C-field will offset the 0 peak un a much lower
degree than it offsets the closer -1 and +1 peaks (which is also
weaker). The nominal C-field setting is reflected in the synthesizer
offset and balances out. That way most cesiums isn't really as "primary"
as being advertized.
> If you restart the system it magically comes back to the same offset.
> Also there is a second harmonic reading and control. If you very the fine
> phase on the 5 MHz indeed the ctl voltage readout adjusts and the system is
> locked. Green light.
It is actually the strength of the 1st harmonic which is interesting,
since if you have strong 2nd degree with almost zero 1st degree, then
you are pretty much on the mark (i.e. lock).
> Here is the real issue the tube has always been so weak that you simply
> can't look at the i meter and see humps. What I had done a long time ago
> was add in another meter that was very sensitive and then use a magnifying
> glass to see the peaks.
I wired up a network analyzer with a mixer to re-modulate the response
onto the 12.6 MHz signal. Worked like a charm, but slow scanning rate is
recommended. I think I did some photos of that.
Anyway, on a weak tube, you are most likely to lock to the 0 pedistal,
because it is the strongest of all 7. If you have a GPSDO you can
pre-tune your OCXO to help locking.
> On Sat, Dec 6, 2014 at 10:41 AM, Magnus Danielson <
> magnus at rubidium.dyndns.org> wrote:
>> Hi Paul and Bob
>> On 12/06/2014 04:04 PM, Bob Camp wrote:
>>> Sure you can set them on the wrong peak. If you really get confused, you
>>> can also set them on the wrong transition … (gulp).
>>> Peak wise, the one you want is the highest Q / best SNR. Set it to one of
>>> the others and your ADEV degrades.
>>> Transition wise … not a good idea at all.
>> It's worse than that. Of the 7 peaks, the middle one has significantly
>> least sensitivity to the C-field, as well as having the strongest response.
>> The "new" digital cesiums actually measures the near-by peaks to sense the
>> C-field and servo the C-field and then use the center peak for servo the
>> frequency. This is a key to increase the stability of frequency and reduce
>> a systematic effect.
>>> On Dec 6, 2014, at 9:11 AM, paul swed <paulswedb at gmail.com> wrote:
>>>> I have a curious question that really applies to all Cs references.
>>>> Its possible to set them on to the wrong peak.
>>>> Typically in the literature it will speak to at least 3 peaks and you
>>>> to select the highest central peak.
>>>> However if you select the wrong peak, how much would the output frequency
>>>> be off?
>>>> I had read a tech note for the airforce that seems to indicate its pretty
>>>> easy to get on to the wrong peak.
>>>> time-nuts mailing list -- time-nuts at febo.com
>>>> To unsubscribe, go to https://www.febo.com/cgi-bin/
>>>> and follow the instructions there.
>>> time-nuts mailing list -- time-nuts at febo.com
>>> To unsubscribe, go to https://www.febo.com/cgi-bin/
>>> and follow the instructions there.
>> time-nuts mailing list -- time-nuts at febo.com
>> To unsubscribe, go to https://www.febo.com/cgi-bin/
>> and follow the instructions there.
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts