[time-nuts] HP5061 Cesium ion pump question

Wed Dec 2 15:21:42 UTC 2020

When I found my HP5061A lingering in a skip, soaked in rainwater some 20
plus years ago, I thought I'd be lucky if it was only good for the OCXO,
and never even contemplated it might be fully working.  I first left it in
a warm room for a week to dry out then, without having a clue about how to
get it going, tentatively plugged it in with the switch set to 'Caesium
off', and just left it alone to stabilise.   With no instructions that
seemed instinctively the right thing to do.   To my delight 5MHz appeared,
so I borrowed a Rb standard from work to check, and the OCXO was clearly
working well - very well - at first sight, just looking at traces on a
scope, the OCXO seemed as stable  and accurate as the Rb was supposed to
be.   I already had an MSF reference at that time, so comparisons were made
against that too.

Over a period of a couple of weeks I just left teh 5061A running and put
out an appeal on the LF chat group on the  (still quite new then) Internet
for anyone who could help me with a manual.   SOme kind person trusted me
enough to actually post me their hard copy on loan, which I copied,
returned the original and studied thoroughly.  And discovered....

THE FIRST THING TO DO when powering up from cold is to leave it in standby
for several days / weeks to pump out the ions.   EXACTLY what I'd already
done! Without thinking.  So as soon as I'd read that paragraph, following
the instructions, tentatively switched to 'open-loop' and watched the meter
showing beam current rise.  After an hour or so when it appeared to have
settled,  (I was too excited to wait the recommended three) switched to
'OPER' and pressed the reset button.   The joy when the green light came on
!   A working Cs standard, recovered as scrap from a skip.   The light went
off after an hour, but pressing the button made it come back on, and it
stayed on from then on.

Further studying the manual, and vaguely knowing that unit's history and
politics surrounding the original owning organisation, (don't ask, I won't
say)  suggested there could actually  be a fair bit of life left in the
tube - beam current certainly suggested so.    The internal NiCd batteries
were completely dead and had a few green crystals on them, so I removed the
whole battery and its charging circuitry, substituted an LM317 regulator
set for 27.2V and used a pair of (old second hand) lead acid gel cells as
backup instead.   And there it sits to this day, on standby
continuously for over 20 years with the Cs switched on when wanted.
 Typically it only needs 35 - 45 minutes warm up, so I allow 1.5 - 2 hours
before any serious tests that need full accuracy.   Ironically, it's been
running for so long the only things to have failed have been my backup gel
cells - floated in series at 27.2V, one set bulged alarmingly after some 13
years, and another pair died such that it couldn't hold-up even standby for
more than about 15 minutes when mains went off.   I replaced those only two
weeks ago.

The only thing I've never tried is calibrating out local magnetic field
using Zeeman lines.   Not having the Cs running continuously, I suspect
this would be a pointless exercise  anyway, and the loss of a few parts in
10^-13 don't over-worry me.

Andy
www.g4jnt.com

On Wed, 2 Dec 2020 at 14:43, paul swed <paulswedb at gmail.com> wrote:

> Morris
> I do see that of the 3 HP5061s the results are different and do know each
> tube leaks or internally emits stuff. But as you say maybe the metering is
> off. It definitely trips out at 40ua.
> I have my units that pulse like that and then settle down as the tube
> clears.
> But its hard to explain that if a unit is pumped down how quickly it
> pollutes again.
> Wondering if as an example the getter just releases its captured crud when
> the power is removed. That way you are always pumping down mostly the same
> stuff.
> Just very curious.
> Regards
> Paul
> WB8TSL
>
>
> On Wed, Dec 2, 2020 at 6:23 AM <vilgotch1 at gmail.com> wrote:
>
> > I think there's quite a bit of variation in leakage rates between
> > individual
> > physics packages, and there's also variation in the accuracy of the
> > metering. I recently started up my spare 5061A after over a year in
> > storage.
> > The Ion current meter showed zero with the Cs off but about 40 in the
> Open
> > Loop position. I tried connecting an external +3500V supply through a
> > microammeter as described in the manual and that showed less than 10 uA.
> > The
> > unit kept trying to start and pulsed for about 12 hours at a peak reading
> > of
> > 40 and then settled down nicely to a zero reading and hasn't missed  a
> beat
> > since.
> >
> > I have a spare tube sitting here that hasn't been powered up for several
> > years. It was removed from a junked 5061A that wouldn't start at all but
> > after the unit was parted out some failed caps were found in one of the
> > modules that could have explained the fault. I plan to try it on the
> bench
> > with the +3500 V supply and microammeter but I don't know whether I can
> > trust the reading. If the tube in one of my functioning units fails one
> day
> > then I'll give it a try.
> >
> > Morris
> >
> > ------------------
> > From: paul swed <paulswedb at gmail.com>
> > To: Time-nuts <time-nuts at lists.febo.com>
> > Subject: [time-nuts] HP5061 Cesium ion pump question
> >
> >
> > During my HP 5061Cesium testing last week I was watching the ion pump
> > current and am curious.
> > What is the typical behavior people see on the tube after say 3 months of
> > being turned off.
> > With 3 units I see one that pumps doen from 8 to 2 in 40 minutes. Another
> > maybe at 25 and takes 28 hours to get to 18. Just seems to be all over
> the
> > place.
> > Regards
> > Paul
> > WB8TSL
> >
> >
> > ------------------------------
> >
> > Message: 10
> > Date: Tue, 01 Dec 2020 06:19:25 +0000
> > From: "Poul-Henning Kamp" <phk at phk.freebsd.dk>
> > To: Discussion of precise time and frequency measurement
> >         <time-nuts at lists.febo.com>, paul swed <paulswedb at gmail.com>
> > Subject: Re: [time-nuts] HP5065a cfield issue
> > Message-ID: <86443.1606803565 at critter.freebsd.dk>
> > Content-Type: text/plain; charset="us-ascii"
> >
> > --------
> > paul swed writes:
> >
> > > Guessing far less than 3 ma. by adding up resistors.
> >
> > More like typical 4 mA: there is a factory select resistor in
> > parallel with the current-setting resistor.
> >
> >
> > --
> > Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
> > phk at FreeBSD.ORG         | TCP/IP since RFC 956
> > FreeBSD committer       | BSD since 4.3-tahoe
> > Never attribute to malice what can adequately be explained by
> incompetence.
> >
> >
> >
> > ------------------------------
> >
> > Message: 11
> > Date: Tue, 1 Dec 2020 10:40:54 +0000 (UTC)
> > From: ew <ewkehren at aol.com>
> > To: time-nuts at lists.febo.com
> > Subject: Re: [time-nuts] EOL Motorola Oncore Remote Antenna
> > Message-ID: <1292998389.3097482.1606819254766 at mail.yahoo.com>
> > Content-Type: text/plain; charset="utf-8"
> >
> > Jim, I agree. As part of my OSA 8607A GPSDO project I planned to place
> the
> > OCXO in an aluminium box to eliminate pressure change. Step one included
> > buying a separate unit to make sure the rubber gasket did do the job.
> > Attached first results. Temperature does change pressure more than
> ambient.
> > The vertical steps are the result of gentle tightening the lid, did not
> > know
> > how much lid had to be tightened. Now I know and have a 10 day test
> > running.
> > With AC running, the AC control at the other end of the house lab
> > temperature stays within ! C. Today a cold front hit, no AC and
> temperature
> > will drop more than 1 C. Taking advantage of the forecast I am extending
> > the
> > test for a couple more days.? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
> ? ?
> > ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?Doe to the low
> power
> > dissipation of the 8607 fan cooling is not an option but I will use the
> box
> > , seal it and characterize the 8607 for pressure and temperature and use
> > external pressure and temperature sensors for compensation once combined
> > with a new generation GPSDO.? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?
> ? ?
> > ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ??
> >
> >
> >
> > I'll bet pressure changes inside the "sealed" radome due to temperature
> > changes are bigger than those due to local barometer changes.
> >
> > But an interesting thing - water vapor will go through cracks, porosity,
> > that liquid water will not. The commercial success of GoreTex is an
> > example of this, but cracks, o-rings that aren't quite right, etc. are
> > also ways it can happen.
> >
> > Making a truly hermetic box is hard.
> >
> > >
> > > -----Original Message-----
> > > From: Poul-Henning Kamp <phk at phk.freebsd.dk>
> > > Sent: Sunday, November 29, 2020 11:19 PM
> > > To: Discussion of precise time and frequency measurement
> > <time-nuts at lists.febo.com>; Art Sepin <art at synergy-gps.com>
> > > Subject: Re: [time-nuts] EOL Motorola Oncore Remote Antenna
> > >
> > > --------
> > >
> > >> It's obvious from the photo that the O-Ring seal failed its purpose
> > >> over its many years of service. Has the unit totally failed or does
> the
> > electronic portion still function?
> > >
> > > No, the electronics is stone dead.
> > >
> > > To me it looks more like water ingress through micro-cracks in the
> > plastic-dome, and the O-ring did its job and kept that water in.
> > >
> > > The microcracks are uniform and seem to follow the molding flow, and
> that
> > is probably to be expected in our climate:? We have a lot of humid
> > freeze-thaw cycles.
> > >
> > > I wonder if buffing the radomes with car-wax would help ?
> > >
> > >> I said lucky because I found some GSynQ parts here in an engineering
> > >> storage cabinet that we? can send to you at no charge to revive your
> > unit.
> > >
> > > Thanks for the offer, but dont bother: I had a spare on hand, and I may
> > still have third one lying around somewhere.
> > >
> >
> >
> > _______________________________________________
> > 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.
> > -------------- next part --------------
> > A non-text attachment was scrubbed...
> > Name: 11-21 druck.pdf
> > Type: application/pdf
> > Size: 171559 bytes
> > Desc: not available
> > URL:
> > <
> >
> http://lists.febo.com/pipermail/time-nuts_lists.febo.com/attachments/202012
> > 01/0a9b642a/attachment.pdf
> > <
> http://lists.febo.com/pipermail/time-nuts_lists.febo.com/attachments/20201201/0a9b642a/attachment.pdf
> >
> > >
> >
> > ------------------------------
> >
> > Subject: Digest Footer
> >
> > _______________________________________________
> > time-nuts mailing list
> > time-nuts at lists.febo.com
> > http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> >
> >
> > ------------------------------
> >
> > End of time-nuts Digest, Vol 197, Issue 1
> > *****************************************
> >
> >
> > _______________________________________________
> > 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.
>