Joe,
    I had similar patterns which I traced to the 
pressure fluctuations from the ref gas.  If the 
flow out the vent capillary (or another split) or 
through the main split is too high, and the 
pressure on the ref tank regulator too low, the 
needle regulators on the conflo have a hard time 
adjusting and "flutter".  You can see the 
amplitude/frequency  change by changing the pressure on the ref tank regulator
Greg

   At 04:05 PM 7/18/2008, you wrote:
>Hi Everyone,
>
>         Just by chance we have noticed a cyclic pattern of noise with our
>Gasbench-IRMS system.  We are delivering CO2 from a tank through REF 1 on
>the Gasbench, which is interfaced with a Delta Plus equipped with the
>CF-needle valve(s), dual inlet, multiport, and freezing finger.  I’ve given
>a link below to a JPEG image to illustrate the machine’s behavior.  We
>simply plotted the beam intensities through time.  As you can see there is
>this rhythmic pattern of changes in all 3 cup/mass intensities.  The pattern
>is cyclic but does not repeat itself perfectly.  The entire system has a
>power conditioner that has been in line for years.  What is strange is that
>the relative “jump” in beam intensities 
>appears to be about the same for all
>3 cups despite the different amplifiers.  If REF 1 is OFF the background is
>stable with no cycles whatsoever.  Delivering CO2 from the dual inlet alone
>(CF-closed) the beam is also stable, which points to the problem being with
>the Gasbench/He flow.  Has anyone noticed such a pattern?  Is it possible
>that either the He carrier gas or the CO2 reference gas is pulsing in
>pressure/flow rate to produce this?  Our Gasbench setup is pretty straight
>forward, as we have not tinkered with settings set forth by Thermo.
>
>http://www.geo.ua.edu:16080/asil/Images/CO2REF1on.jpg
>
>         Can anyone recommend the best way to 
> check the flow rate throughout the
>Gasbench?  We have a couple bubble flow meters but we need a way to properly
>go from a capillary to a large tube for accurate numbers.  Maybe a digital
>flow meter exists that can look at flow rate through time?
>
>         Let me add one last bit of information 
> to the mix.  Our next step is to add
>water/oxygen/hydrocarbon filters to the He line.  I could see the lower
>purity He (5.0) giving bad precision, etc. but I do not see a way this could
>give the mentioned behavior.  The same pattern has been present with 2
>different He tanks and we have a 2-stage regulator on each of the tanks.
>
>I appreciate any comments or recommendations you might have.
>
>Thanks!
>
>Joe Lambert
>Alabama Stable Isotope Laboratory

Greg Michalski, Assistant Professor
http://www.purdue.edu/eas/michalski/
Department of Earth and Atmospheric Sciences
Purdue University
550 Stadium Mall Dr.
Lafayette IN.47907