Hiya Bryan :)

>         We noticed poor precision of organic carbon and nitrogen standards
> in continuous-flow mode (using a Carlo Erba elemental analyzer), so we ran
> a 24-hour time scan on CO2 from the EA's reference gas injector.  All beams
> and ratios OSCILLATE synchronously with a period of ~30-60 minutes; the
> amplitude of the major beam oscillations is ~1E-10 A (with a trap
> current=200 uA tuning).  The story is similar for injected N2.  This
> problem showed up suddenly, within a 6-day span of unuse.  We have since
> found the same oscillations (with smaller amplitude) in a time scan of CO2
> in the dual inlet.

My guess would be that your filament is about to bite the dust.  The
thoria coating tends to thin out as the filament gets older, and the
higher temperature needed for thermionic emission results in sag,
which changes the required current, so the filament "unsags" and so
on.  If it's any older than about a year old, this could well be it :(

First:  Try taking off the amp head, and cleaning the gold contacts
both on the amplifier feedthrough, and on the amp head itself.  Use
something nice and non-abrasive, and a bit of acetone (a "good"
grade).  Sometimes a little bit a dirt/grease can cause a miniature
capacitance effect across one of the pins, which /can/ show up as
short time-scale oscillations.

As a final note, the electro-magnet balancing resistor (mounted on the
bench under the amp head) is extremely temperature sensitive,
especially at ambient temperatures about about 27°C/81°F.  Is your
peak shape as sharp as it should be?  If not, you could be seeing
drift across a "nasty" part of the peak shape.

>         We have tested everything we can think of in order to track this
> problem down, without success.  If anyone has suggestions or has seen this
> problem before, I'd love to hear about it.

I've seen this kind of thing a couple of times, on a SIRA10 and a
PRISM (II or III), and it was filament death both times.  :( The dirty
feedthroughs I've seen on all the older (pre-Optima) machines, SIRAs,
602s, PRISMII etc, but the manifestation is more "sawtooth" than
"sine", as the capacitance discharges.  But it could be??

Good luck  :))

Andrew.  <8^)