[time-nuts] Fwd: HP5061B Ion Current

[Donald E. Pauly]

https://www.febo.com/pipermail/time-nuts/2017-March/104374.html

Paul:

I don't think that I made myself clear.  This power supply is a rare
example of HP engineering incompetence.  The transformer is EASILY
capable of 10 Watts out at 3500 V.  Instead they made it where it
can't even put out 100 mW without sagging.  This supply was designed
in the middle 80s when IC technology was well developed.

>From what I am able to determine about ion pumps, current is
proportional to pressure and voltage is proportional to pumping rate.
Paragraph 4-349 of the HP5061B manual states that 2 μA ion current
corresponds to a pressure of 5x10^-7 Torr.  At 760 Torr, the mean free
path of an oxygen molecule is 93 nm.  Assuming that a cesium atom is
the same size, this gives it a mean free path of 141 meters.  This
means that there is a 50% chance of a collision in a 141 meter travel
thru such a vacuum.  At an ion current of 1 mA, the pressure would be
500 times greater and the mean free path 500 times less or 0.282
meters.  This is 11" or about the length of travel for the beam in the
tube. A beam current of 1 mA would only cause a loss of 50% in beam
current and should still allow lock.

That same paragraph claims that cesium is turned on at 40 μA.  That
was not true.  Another place claims 20 μA (Appendix A-1 b) as well as
30 to 40 μA.  The exact trip point depends on how much sag occurs in
the HV supply and is very complex to calculate.  It is stupid to have
to wait for weeks for a reduction in ion current to allow the cesium
to come on naturally.  Lock can be achieved instantly.   I think the
final vacuum improvement can be achieved more quickly if the cesium
oven is on.  It should cook off cesium that has condensed on its
outside.

See page 8-51 schematic.  We measured the properties of the power
supply as it is.  For 18.7 V supply, 357 μH and 20 μs, energy stored
is in T1 primary is 108 μ Joules.  Frequency can be adjusted from 524
to 2002 cps.  Power supply with no load was set to 704 cps for +3,500V
out with no load except for the internal 200 Meg bleeder.  Input power
was 76.3 mW and output power was 61.25 mW or 80% efficiency.  At
frequencies above 704 cps, output voltage increases above safe levels
for no load.

With gassy beam tube for load, voltage sags to 2296 V with 39 μA ion
current 11.4 μA bleeder current.  Total power is 115.9 mW.  With
frequency raised to 55 μA ion current, voltage was 2460 V or 165 mW
power.  With frequency turned all the way up to 2002 cps, voltage was
2562 V and 71.4 uA ion current or 215 mW power.

We are working on a voltage regulator and current limiter that will
provide at least 1 mA before voltage reduction from +3500V.  This is
an improvement of at least 20 to 1. Output power is limited by the 21
kc self resonance of the transformer due to secondary winding
capacitance.  We don't know its saturation current yet. We plan to use
the two unused meter switch positions to monitor output voltage and
current.  It will likely mean that no external 3500 volt supply will
ever be required for tubes that have been in long storage.

I made a typo on WB4BPP's call in my first post but he reports using a
5,000V external supply.  He didn't provide current figures or whether
this voltage actually stayed there.  I am concerned about arc over at
the ion pump at that voltage.  Ion pump manufacturers caution about
overheating at high pump currents.  HP claims in Appendix B A-2 h that
+3500 V at 5 mA for no more than 15 minutes is permissible.  This is
17.5 Watts which sounds like a lot for a small cathode.

When we overrode the cesium lockout at 29 μA or so of ion current, we
needed only minor front panel adjustments for beam current of 20 μA.
(We shorted across A15 R-4.) Our last ion current before power supply
modifications at risen to 39 uA.  Beam current has been stable.

WB0KVV

πθ°μΩω±√·Γλ

---------- Forwarded message ----------
From: paul swed <paulswedb at gmail.com>
Date: Mon, Mar 20, 2017 at 7:08 PM
Subject: Re: [time-nuts] HP5061B Ion Current
To: Discussion of precise time and frequency measurement <time-nuts at febo.com>

Donald welcome to the group. If a units been off a long time and it sure
sounds like thats the case it may take quite a while like a month or so for
the unit to remove all of the "Stuff" that has out gassed. So be patient
and let the pump do its job. After it does lower and my fingers are
crossed. Then you only need to run it about every 6 months.
The fact that it actually locks and you found a simple fix is pretty good.
What was the beam current?? That gives you a hint on the quality of the
tube.

Not sure I would run the defeat on the HV supply for to long. That may
stress the supply if I had to guess.
Regards
Paul
WB8TSL