Reference Manual
748384-C
September 2003
Rosemount Analytical Inc. A Division of Emerson Process Management
Introduction 1-15
NGA2000 Reference
You should take the above warnings very
seriously if you are trying to measure in the
I ppm or less band.
If you think that you are not getting good re-
sults at these low levels, the reasons are
almost invariably due to sample contamina-
tion. Make sure you rule out any possible
sources of contamination before deciding
that you have an analyzer problem.
Network variables: CALTIME, CALTIME-
OUT, ZERO, SPAN, AMFN, CALCHKLIM-
ITS, ZERORNGS, CALRANGES,
CALFAIL, CALFPC
d. Troubleshooting
Zero drifts down over time…
If a FID, or trace 0
2
, analyzer, you have
sample contamination. In the case of the
FID, you must determine where the con-
tamination is coming from, possibly by con-
necting zero gas directly to the sample inlet,
using new ultra-cleaned stainless steel tub-
ing, and verify that the analyzer itself is not
contaminated. If it is, it will clean itself out in
time if you leave it, but if this is not practical,
you have to send it back to the factory for
disassembly and cleaning. Otherwise, re-
place your tubing with stainless steel, and
clean all the sample components, preferably
with a blow torch while flowing argon gas
through it.
For oxygen analyzers, an elevated zero
means a leak. Remember that oxygen dif-
fuses through any kind of plastic. Also re-
member that it is the partial pressure
difference that controls the diffusion rate -
the fact that your sample is at 3000 psig
means nothing to the oxygen atoms. To
them, a bottle of pure nitrogen is a vacuum,
and they whistle into it if given a chance. It
is normal for an oxygen analyzer to take a
day or two to come to a good zero reading,
leave it for a while and see if it stabilizes be-
fore worrying.
You can test for leaks by shutting the sam-
ple flow off both upstream (first!) and down-
stream of the analyzer. The oxygen reading
should stabilize at a low level of a ppm or
so, if it goes up to a significant level you
have a leak.
After a zero, the reading is very erratic…
Put a span gas in and see what it reads. If
you have spanned on a zero gas, with the
calibration limits disabled, you will have
made the span setting much too sensitive.
In this case, verify you do indeed have span
gas in the analyzer, and span it again (dis-
abling the limit checking first), then zero it
again.
Similarly, if you have zeroed with a span
gas, zero it with a real zero gas and re-
span, then zero again. Now enable the limit
checking so this doesn't happen again.
It refuses to zero…
Verify you do have a real zero gas in the
analyzer, and disable limit checking. Then
zero it again. Also, zero on all ranges inde-
pendently.
After a span, it won't go back to zero…
Leave it on zero gas for a long time, at least
a hundred times the response time of the
analyzer. See if it made it back to zero
again. If not, check for leaks of span gas
into the zero gas. If the zero is now lower
than it was before, i.e. negative, re-zero
and then re-span. If it is higher, and the
previous zero happened within a few min-
utes, something is wrong with the zero gas,
the sample system, or possibly the ana-
lyzer. If the analyzer temperature has
changed, such as by opening a door and al-
lowing cold or hot outside air to blow on the
analyzer, let its temperature stabilize again
and try again.
It may be that the analyzer span is now very
different from before. Check your span gas
for correctness, and that you did indeed
have flow of gas during the span. If all is
well, re-zero and then re-span.
