[time-nuts] HP 5065 Rb update

Tue Dec 8 01:48:54 UTC 2020

Hi

Since the magnetic “response” of the field around the physics package is 
non-linear, reversing the field on a package with magnetized mu metal
*should* produce a delta frequency. 

More or less:

	C field at 4 ma is +0.10 gauss ( no it’s not, but you need a number here …)
	Mag of the mu metal is +0.15 gauss ( it’s the ratio that matters ….)

So starting out you get 0.15 + 0.10 = 0.25 gauss 

Reverse the field and you get 0.15 - 0.1 = 0.05 gauss

If there *is* a significant static field, ( = enough to shift things out of pot range), 
then you should see a very significant frequency shift when you reverse the
C Field current. 

For all the fun with math that backs this up see:

http://www.wriley.com/Rubidium%20Frequency%20Standard%20Primer%20102211.pdf <http://www.wriley.com/Rubidium%20Frequency%20Standard%20Primer%20102211.pdf>

Figure 36 on page 46 gives you a rough idea of what’s going on. There is more 
detail in the text. 

Bob

> On Dec 7, 2020, at 7:35 PM, Dana Whitlow <k8yumdoober at gmail.com> wrote:
> 
> Rheostats are not necessarily stable because the current is 1st-order
> dependent
> on the contact resistance.
> 
> I had missed the part about the nearby lightning strike.  Could it be that
> the mu-metal
> shielding got partly magnetized by the high current pulse of the
> lightning?  If so, there
> ought to be some stability benefit from performing a degaussing operation.
> If not
> degaussed, the magnetization of the mu-metal will likely change over time
> due to
> physical knocks, changes in orientation of the Rb in the Earth's field,
> temperature
> changes, etc, all of which will make the Rb less stable than it could be.
> 
> Dana
> 
> 
> On Mon, Dec 7, 2020 at 6:18 PM Bob kb8tq <kb8tq at n1k.org> wrote:
> 
>> Hi
>> 
>> The existing c field winding is *inside* the magnetic shield on the
>> physics package.
>> Simply getting a second winding inside that package involves some (scary
>> to most
>> of us) tearing apart (an re-assembly) of the physics package.
>> 
>> Next up, the mag field needs to be reasonably uniform. That means that the
>> “new”
>> winding needs to overlap pretty much all of the “old” winding. Again a bit
>> of a chore.
>> 
>> Finally, that “crude rheostat” circuit is pretty stable. The current needs
>> to be stable into
>> the << 100 ppm sort of range to keep the device happy. Ideally something
>> that does
>> a couple ppm ( 1 to 3ppm) would be the “best case” in terms of device
>> stability.  Noise
>> is (obviously) an issue as well.
>> 
>> Since the sensitivity to mag field decreases with C Field current ( also
>> with the assumption
>> there is no stray field ….) there is a tradeoff between “tuning range” and
>> required stability.
>> 
>> One alternative is to brew up a fixed C field regulator board. Set the
>> field to the lowest
>> practical value using that board. Then tune the device by other means.
>> 
>> Bob
>> 
>>> On Dec 7, 2020, at 2:31 PM, Dana Whitlow <k8yumdoober at gmail.com> wrote:
>>> 
>>> Paul,
>>> Would it be feasible to add a smaller winding around the outside of
>>> the existing C-field winding, and drive it from something other than
>>> the crude rheostat circuit?  Then you could make finer adjustments
>>> than the existing driver permits, and would rely only on the existing
>>> driver to provide only a coarse tuning field.  Admittedly this would
>>> still involve a "dangling pot", but now entirely on your own terms.
>>> 
>>> Dana
>>> 
>>> 
>>> On Mon, Dec 7, 2020 at 12:37 PM paul swed <paulswedb at gmail.com> wrote:
>>> 
>>>> Ed
>>>> Like you I don't like dangling pots either. But since I had the system
>> open
>>>> I took a look to see how hard that would be to accomplish. The pots on
>> the
>>>> digital counter and its actually quite small. The 3 leads all fit into
>> an
>>>> area about the size of a pencil eraser. To work on it requires removing
>> the
>>>> front panel. Decided to leave that alone. To risky.
>>>> The systems all re-assembled stable and working about the way it always
>>>> has. I did go with Bobs suggestion using a different synth count and the
>>>> lower cfield current.
>>>> Regards
>>>> Paul.
>>>> WB8TSL
>>>> 
>>>> On Mon, Dec 7, 2020 at 1:12 PM ed breya <eb at telight.com> wrote:
>>>> 
>>>>> OK, I see now what the circuit looks like, in the manual. The pot is
>>>>> used in rheostat mode, so the specific value is needed, and it would
>>>>> take some more modification of the circuit to increase the setting
>>>>> resolution.
>>>>> 
>>>>> I would recommend making a simple change to the existing pot hookup
>>>>> though, which is to tie the "NC" end of the pot to the wiper - I have
>>>>> this thing about not leaving dangling ends on pots used as variable
>>>>> resistors. The most likely possible failure mode in a pot is an open or
>>>>> intermittent wiper. In this circuit, if the wiper should open for some
>>>>> reason, the circuit will go open-loop, and self-limit at around 13 mA
>>>>> through the coil (or somewhat less if the coil's resistance is higher
>>>>> than 700 ohms or so) - twice the normal maximum setting, or four times
>>>>> the power dissipation. I don't know if this matters here, but for
>> higher
>>>>> reliability, tying that pot's loose end to the wiper will ensure that
>>>>> this can't happen - the circuit will limit at the maximum of the normal
>>>>> adjustment range.
>>>>> 
>>>>> Ed
>>>>> 
>>>>> _______________________________________________
>>>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>>>> To unsubscribe, go to
>>>>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>>>>> and follow the instructions there.
>>>>> 
>>>> _______________________________________________
>>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>>> To unsubscribe, go to
>>>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>>>> and follow the instructions there.
>>>> 
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at lists.febo.com
>>> To unsubscribe, go to
>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>>> and follow the instructions there.
>> 
>> 
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at lists.febo.com
>> To unsubscribe, go to
>> http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
>> and follow the instructions there.
>> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.