Teledyne 9110TH - Nitrogen Oxides Analyzer - Manual (Page 243)

Troubleshooting & Repair Model 9110TH NOx Analyzer
Teledyne Analytical Instruments 223
 The standard gases may be mislabeled as to type or concentration. Labeled
concentrations may be outside the certified tolerance.
 The sample delivery system may be contaminated. Check for dirt in the
sample lines or reaction cell.
 Calibration gas source may be contaminated (NO
2
in NO gas is common).
 Dilution air contains sample or span gas.
 Ozone concentration too low because of wet air in the generator. Generator
system needs to be cleaned and dried with dry supply air. Check the Perma
Pure dryer for leaks. This mostly affects linearity at the low end.
 Sample inlet may be contaminated with NO
X
exhaust from this or other
analyzers. Verify proper venting of the pump exhaust.
 Span gas overflow is not properly vented and creates a back-pressure on the
sample inlet port. Also, if the span gas is not vented at all and does not
supply enough sample gas, the analyzer may be evacuating the sample line.
Make sure to create and properly vent excess span gas.
 Diffusion of oxygen into Teflon-type tubing over long distances. PTFE or
related materials can act as permeation devices. In fact, the permeable
membrane of NO
2
permeation tubes is made of PTFE. When using very
long supply lines (> 1 m) between high concentrations span gases and the
dilution system, oxygen from ambient air can diffuse into the line and react
with NO to form NO
2
. This reaction is dependent on NO concentration and
accelerates with increasing NO concentration, hence, affects linearity only
at high NO levels. Using stainless steel for long span gas supply lines
avoids this problem.
7.3.7. DISCREPANCY BETWEEN ANALOG OUTPUT AND DISPLAY
If the concentration reported through the analog outputs does not agree with the value
reported on the front panel, you may need to re-calibrate the analog outputs. This
becomes more likely when using a low concentration or low analog output range.
Analog outputs running at 0.1 V full scale should always be calibrated manually. See
Section 4.13.6.2 for a detailed description of this procedure.
7.3.8. DISCREPANCY BETWEEN NO AND NO
X
SLOPES
If the slopes for NO and NO
X
are significantly different after software calibration (more
than 1%), consider the following two problems