EPA-600/2-76-210
July 1976
Environmental Protection Technology Series
          AMBIENT  CARBON  MONOXIDE  MONITOR
                                      Environmental Sciences Research Laboratory
                                           Office of Research and Development
                                          U.S. Environmental Protection Agency
                                    Research Triangle Park, North Carolina 27711

-------
                 RESEARCH REPORTING SERIES

 Research reports of the Office of Research and Development, U.S. Environmental
 Protection Agency,  have been grouped into five series. These five broad
 categories were established to facilitate further development and application of
 environmental technology. Elimination  of traditional grouping was consciously
 planned to foster technology transfer and a maximum interface in related fields.
 The five series are:

     1.    Environmental Health Effects Research
     2.    Environmental Protection Technology
     3.    Ecological Research
     4.    Environmental Monitoring
     5.    Socioeconomic Environmental Studies

This report  has been assigned  to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed to develop and
demonstrate  instrumentation, equipment, and methodology to repair or prevent
environmental degradation from point and non-point sources  of pollution. This
work provides the new or improved technology  required for the control and
treatment of pollution sources to meet environmental quality standards.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
               AMBIENT CO MONITOR
    D. E. Burch, F. J. Gates and J. D. Pembrook
           Aeronutronic Ford Corporation
               Aeronutronic Division
          Newport Beach, California 92663
              Contract No. 68-02-2219
                  Project Officer

                  W. A. McClenny
Emissions Measurement and Characterization Division
    Environmental Sciences Research Laboratory
   Research Triangle Park, North Carolina 27711
        U.S.  ENVIRONMENTAL PROTECTION AGENCY
         OFFICE OF RESEARCH AND DEVELOPMENT
     ENVIRONMENTAL SCIENCES RESEARCH LABORATORY
    RESEARCH  TRIANGLE PARK, NORTH CAROLINA 27711

-------
ii

-------
                                     CONTENTS






                                                                         Page




I         Introduction                                                      1




II        Summary                                                           3




III       Conclusions                                                       5




IV        Recommendations                                                   "




V         Optical Layout                                                    8




VI        Sampling and Calibration                                        15




VII       Electrical Circuits and Controls                                18
                                       iii

-------
Number
                                LIST OF FIGURES


                                                                         Page
  1       Two Photographs  of  the Ambient CO Instrument.                     *

  2       Optical Layout of the Components Used to Measure the Concen-
          tration of CO.

  3       Optical Diagram  of  the Entrance Optics.

  4       Exploded View of the Rotating Gas-Filter Cell.                    1°

  5       View of the Images  Formed on Mirrors M4 and M6                    12

  .6       Optical Diagram  of  the H-O Monitor.                               14

  7        Calibration Curves  Relating the Output Signal to the Concen-
          tration of  CO.                                                    17

  8        Electrical  Wiring Diagram of Main Instrument Box.                 19

 •9       Wiring  Diagrams  for DC and AC Power Cables.                       20

  10      Wiring  Diagram of the Battery Cable and the Power Converter
         Assembly.                                                         21
                                      iv

-------
                                   SECTION I

                                  INTRODUCTION
     The concentration of CO in unpolluted air typically varies from 0.1 to 0.3
parts per million  (ppm).  In areas of high automobile traffic the concentration
is frequently several ppm, and in some cases it may exceed 50 ppm.  In order to
study accurately the rates of accumulation and dissipation of CO in the neighbor-
hood of expressways, it is clear that the instrument being used must have a min-
imum-detectable concentration of no more than a few tenths of a ppm.  If the
concentration is to be measured in areas away from heavy traffic where the air
is only slightly polluted, the instrument must be portable and have a minimum
detectable concentration less than 0.1 ppm.

     Several commercially available CO monitors have minimum detectable concen-
trations from a few tenths of a ppm to 10 ppm.  These instruments are not portable
and are not sufficiently accurate for detailed studies of the distribution of CO
in the ambient air.  In 1974, Aeronutronic designed and built an ambient CO moni-
tor for EPA under Contract 68-02-1655.  This instrument has much higher perform-
ance than previously available instruments and has a minimum detectable concen-
tration well below 0.05 ppm.  The instrument can be operated from 115 V ac power
lines or from batteries and is portable so that it can be moved easily to any
location.  This report describes two additional monitors built recently for EPA
to monitor CO in ambient air.  The recent instruments are similar in many ways
to the one built previously.  A few minor improvements have been made in order
to increase the stability and to decrease the time required to flush a sample
from the sample section.  Scientific personnel at the EPA have found that the
sensitivity of the instruments to H20 and the dependence of the zero-setting and
span calibration on temperature are quite acceptable.  Values of CO concentra-
tion obtained with these instruments compare very favorably with values obtained
for the same samples with other instruments.

     Gas-cell correlation methods are employed for the detection and for the
discrimination against other gases in the atmosphere.  The spectral principles
of detection and discrimination have been described previously.*-»2  in ambient
1.   Burch, D. E. and J. D. Pembrook.  Instrument to Monitor CH,, CO and CO.
     in Auto Exhaust.  EPA-650/2-73-030, October 1973.

2.   Burch, D. E. and D. A. Gryvnak.  Infrared Gas Filter Correlation Instru-
     ment for In-Situ Measurement of Gaseous Pollutants.  EPA-650/2-74-094,
     December 1974.

-------
air the concentration of H£0 may be  as  great  as 4  or  5%  in a  humid atmosphere.
This amount of H^O produces enough interference in the measurement of CO that
the interference must be accounted for  in  order to achieve the  maximum possible
accuracy.   An accessory built into the  basic  CO instrument measures the concen-
tration of H20 in the air and automatically accounts  for  the  interference by
this gas.

-------
                                    SECTION II

                                     SUMMARY
      Figure  1  shows  two photographs  of  one  of  the  two  instruments described
 herein.   The upper panel includes  the assembly that contains the 24 V dc to
 115  V ac  inverter and the dc-to-dc converter used  with the battery pack.  The
 interior  of  the  instrument is  shown  in  the  lower panel of the figure; the cover
 plate for the  electronics has  been turned back to  show the electronic cards for
 the  CO channel and the t^O channel.  The instrument is portable and is powered
 either by 115  V  ac line power  or by  a 24 V  battery pack.  Only a few minutes
 are  required to  convert from one type of power to  the  other.

     A multiple-pass  sample  cell provides the  required sensitivity while main-
taining a relatively  small  instrument.  The base length of the sample cell is
approximately  43 cm;  it  is  operated at 28 passes to give a total path length of
approximately  12 meters.  The minimum response time, less than 5 seconds, is
limited by the time required to displace the air in the sample section.

     Both the CO and  H^O  concentrations can be read from meters on the top
panel  or with  a recorder  or voltmeter from  output  jacks.   The peak-to-peak noise
of the output when employing a 3-second time constant  corresponds to less than
0.02 ppm  of CO.  This more  than meets the design specifications for a minimum -
detectable concentration  of 0.05 ppm of CO.  Four different concentration ranges
are available  to read concentrations from the minimum  detectable amount to more
than 200 ppm.  By making  a  simple  gain adjustment, the 1^0 concentrations cor-
responding to  full-scale  readings  of the meter can vary from less than 1% 10
to more than 5%
     The outside dimensions of the main box that contains all of the optical
components are approximately 62 cm x 27 cm x 12 cm.  Switches and control knobs
extend beyond these dimensions on the top and on the front side, which includes
the carrying handle and the output jacks.  The sample exhaust fan extends
a few cm beyond one end.  The weight of the main assembly is approximately
40 pounds.  Threaded holes near the ends of the front and back sides can be
used for rack mounting.  A sunshade can also be attached to the main assembly
by use of these threaded holes.  The instrument can be operated while
sitting on the 4 legs on the bottom side or on the back side.

-------
t:.
'\
$. ,.,., . ,,\."""'1I.t~}
: .-,,~"f',I..) 't'''' ~ ~ ~;.!..,..f:;:~
~,.~><.:."..,.~y, -?"~~{~I~.
~--:;~ - - ~- ,J' .~~.~ ~ - ..,- ...~
"~" ~.... "",,,,f"'~.t' v ._~
:~');G ;I <-)\;to' 't',,;m....#~ ,~._l
6 ~ "':' ~fiIJ' ~:l "''';~ 'Iq :~"4.:--" ."'~ '.
t;~~ ~".J'" ~ ,:.~~..- it!~ \r'" :::;.'.; 'f.-t,;;,: .
.".. . fJ' '. '. ,.. I
:: -~ .:~ ---\ -.. t< -:,.,
. Jn~f'J t,) 1\
'fi ~-', of,' - ;c' ~ ..;'4
f '<".
"
Figure 1.
Two photographs of the ambient CO instrument.
includes the dc-to-ac power supply assembly.
The upper panel
4

-------
                                   SECTION III

                                   CONCLUSIONS
     The results of the tests performed with the two instruments described
herein indicate that they are capable of accurately measuring the concentration
of CO in ambient air.  Interference by 1^0 can be accounted for automatically
by the use of a simple infrared H^O monitor that applies a correction signal
to the main electronics.  The interference by the normal concentration of C02
corresponds to an error of less than 0.01 ppm of CO.  No significant interfer-
ence is produced by any gases other than C02 and 1^0 in the concentrations that
are normally found in the atmosphere.

     Quite good performance can be achieved when powering the instrument by
either 115 V ac line power or by dc power from a battery pack.  Approximately
8 hours of continuous dc operation is possible without recharging the pack of
6 small motorcycle batteries.  The dc-to-ac inverter that powers the synchron-
ous chopper motor maintains the chopping frequency very nearly constant so
that the zero-reading and the sensitivity are quite stable.

-------
                                   SECTION IV

                                 RECOMMENDATIONS
     Air samples can be circulated through the sample section by a variety of
methods.  Two cover plates for the sample section can be removed to allow the
ambient air to move freely through the infrared monitoring beam.  The recom-
mended alternative to this method is to keep the cover plates in place and draw
the air through the sampling section with the exhaust fan provided for that
purpose.  This latter method has the advantage that it is easy to stop the air
flow and flush the sample section with argon to determine the proper zero read-
ing.

     It is sometimes desirable to draw outside air through a gas line to the
instrument placed inside of a building or a portable laboratory such as a
trailer or van.  When this procedure is followed the fan or air pump should be
placed upstream from the instrument so that the sample section is at or slightly
above the local ambient pressure.  This prevents the air surrounding the instru-
ment from leaking into the sample section.  The adjustable outlet vent should
be  left slightly open to allow the air to escape without building up a signifi-
cant pressure in the sample section while avoiding the entrance of the air sur-
rounding the instrument.  The air leaving the instrument can be allowed to es-
cape where it leaves the sample section, or it can be carried through pipe or
tubing to the outside.

     The instruments in their present form provide more sensitivity and accuracy
than are required for many applications.  Several simplifications could be made
to  the instrument that would degrade performance, but not enough to prevent its
use for many types of monitoring.  For example, the thermoelectrically cooled
detector could be replaced by an uncooled detector with the same physical dimen-
sions.  This change would make it possible to eliminate the power supply for
the cooler and would probably decrease the ratio of the signal-to-detector noise
by  a factor of about 8 to 10.  A portion of this loss could be regained by em-
ploying a larger radiant energy source and opening up some of the apertures to
allow more energy to reach the detector.  The small energy source installed in
the instrument was chosen to reduce the power requirement, which is particularly
important when used in the dc mode with batteries supplying the power.  A  larger
source is recommended if only ac power is to be used.  Approximately a factor  of
2 in the amount of energy collected could be gained by replacing the aluminum
reflective coating on the mirrors of the multiple-pass cell with a dielectric
coating with high reflectivity (greater than 99.5%) near 4.6 \m.

     Relaxing the accuracy requirements somewhat would also make it possible
to eliminate the H20 monitor and its associated correction circuitry   The

-------
H20 interference could be partially accounted for by estimating the 1^0 con-
centration or by measuring it by some other means.  Uncertainties of less than
0.2 ppm in the measurements of CO concentration could probably be achieved
without the detector cooler and the EUO monitor.

-------
                                     SECTION V

                                   OPTICAL LAYOUT
DESCRIPTION OF COMPONENTS

     Figures   2  and  3  illustrate the optical layout of the instrument.
The view  in Figure  2  is from above the instrument, and the view of the en-
trance optics  components shown in Figure  3  is from the end of the instrument.
The source is  heated electrically by a wire coiled around a ceramic core; the
useful portion is approximately 0.5 cm long and 0.15 cm in diameter.  The
source is located directly beneath the motor shaft with its long dimension
parallel  to the shaft.  Radiant energy passes through the source window WS
to a toroidal  mirror Ml.  After the beam passes through the rotating gas-
filter cell, an enlarged image of the source is formed on the entrance window
Wl of the sample section.

     The  rotating gas-filter cell is mounted directly to the shaft of the small
1800 rpm  synchronous motor.  A transformer provides approximately 60 volts ac
for the motor, which dissipates approximately 7 watts.  In addition to saving
power, which is important when operation in the dc mode, the motor runs more
quietly and at a lower temperature at the reduced voltage than it would at full
line voltage.  Several seconds are required for the motor to reach synchronous
speed when it  first starts at the reduced voltage; however it easily maintains
synchronous speed after it has been attained.  The primary coil of the trans-
former is connected directly to the 115 V ac power line when operating in the
ac mode.  During dc operation, a dc-to-ac inverter provides the 115 V ac power
for the primary of the motor transformer.  The frequency of the output of the
inverter has been adjusted to 60 Hz so that the motor speed is the same when
operating in either the ac or the dc mode.

     A shield  not shown in either Figure  2  or Figure  3  encloses the rotat-
ing gas-filter cell in order to reduce the flow of heat from the source and  to
prevent dust from accumulating on the windows of the cell.  Window WS transmits
the energy through the shield.

     Much of the good stability and high sensitivity of the instrument is due
to the rotating gas-filter cell, which is illustrated in Figure 4.  Compart-
ment GCl is filled with 0.3 atm of pure CO to form the gas filter; compartment
GC2 is filled with 0.5 atm of non-absorbing N£.  The 5.1 cm  (2 inch) diameter
sapphire windows are anti-reflection coated with a single dielectric  layer  to
provide high transmittance near 4.6 \ua.  Enough CO is introduced in the CO  com-
partment so that its average transmittance, combined with the windows,  is

-------
Gas Inlet
(through
top cover)
                                                     Electronics
                                                         &
                                                      Controls
Adjustable
Shutter
                      Figure  2^.  Optical layout of the components used to measure
                      the concentration of CO.  The optical components used exclusively
                      for the H20 monitor are not included.

-------
                      Rotating Gas-Filter Cell
V




-^
Wl 1


Sample
Section
i





	

f


o

Ml
A




-
-

s

/I
ws






1
1
1
4
•1
•1
|
I
I
I

•












Motor
(1800 rpm)









.








-^_
Source

Figure 3. Optical diagram of the entrance optics.
     Chopper
GC I
Attenuator
Figure 4.  Exploded view of the rotating gas-filter  cell.
                            10

-------
the same as the average  transmittance  of the windows with  the  attenuator in
compartment GC2.  The  cell  body is made  of  stainless steel,  and  it can be re-
filled through a  small port on the outer end of  the cell.  The optical path
through the CO in compartment GC1 is approximately 5 mm.

     A reticle deposited on the surface  of  the rotating cell window that is
located next  to window Wl provides high-frequency chopping of  the beam.  The
chopper has 12 opaque  strips and 12 openings to  produce a high-frequency chop-
ping rate of  360  Hz.   When  CO is added to the sample section,  it absorbs a
fraction of the chopped  energy that passes  through the attenuator side of the
rotating cell.  However, when the gas-filter side, GC1, of the rotating cell
is in the beam, the  gas-filter cell absorbs most of the energy at wavelengths
where the CO  in the  sample  cell absorbs. Therefore, the addition of CO in
the sample section decreases the energy  passing  through GC2 more than it re-
duces the energy  through GC1.  Thus, the 360 Hz  chopped signal contains a
modulation at 30  Hz,  the rotational frequency of the gas-filter cell.  The
amount of modulation depends on the product of the CO concentration and the
path-length in the sample section.

     Toroidal mirror M2  placed just inside  the entrance window Wl to the sample
section directs the  incoming beam to mirror M3.  The radii of  curvature of M2
have been chosen  to  make mirror M3 nearly conjugate to toroidal mirror Ml.
Using the toroidal mirror M2 in place  of a  flat  mirror reduces vignetting by
reducing the  height  of the  beam of light required on mirror Ml to fill the
following optical components.

     Spherical mirror  M3 directs the incoming beam to mirror M4 where an image
is formed of  the  aperture on the entrance window Wl.  Spherical mirror M4
directs the energy of the main beam to mirror M5, from where the beam continues
through the multiple-pass optical system.   Employing mirror M4 with the proper
radius of curvature  to make mirrors M3 and  M5 conjugate to each other avoids
vignetting that would  occur if a flat  mirror replaced mirror J&.

     The optical  principles of the multiple-pass sample cell are explained in
Figure  5  and the legend accompanying the  figure.  Mirror M6  is approximately
43 cm from mirrors M5  and M7.  The mirrors  are adjusted to produce 28 passes
of the cell before the beam exits through window W2.  Therefore, the total
pathlength in the sample section is approximately 12 meters.

     After the beam  exits through window W2, a CaF2 lens directs the beam
through a narrow-bandpass filter and forms  an image of mirror M7 on the PbSe
element of the detector. A 2-stage thermoelectric cooler maintains the sensi-
tive element  of the  detector at approximately -25°C in order to improve its
detectivity.  The filter passes a narrow bandpass that has been selected to
provide good  sensitivity to CO while minimizing  the sensitivity to C02 and
H 0, both of which occur in the atmosphere  and absorb in the same spectral
region.
                                     11

-------
                                   M6


                        26 22 18 14 10  6  2
                                 Ox
                                (5)(7)
It.
it:
                            4   8  12 16 20 24
 Figure 5.   View of the images formed on Mirrors  M4  and M6.  The view is
 observed when  facing mirror M6 from inside the sample cell.  An image of
 the slit at the entrance window Wl is formed by  mirror M3 on mirror M4 at
 the position indicated by a 0.  (See Figure 2 for other components.)
 Energy directed from M4 to M5 is imaged by M5 on M6 at the position ad-
 jacent to  the  number 2 on M6, which indicates the beam has made 2 passes
 of the multiple-pass optical system.   The energy continued from image #2
 to mirror  M7, which forms image #4 on mirror M6, etc.  Images in the top
 row are formed  by mirror M5; those on the bottom row are formed by M7
 After  28 passes, the main beam goes through window  W2 to detector C (s^e
 Figure 2).   Lens LI forms an image of mirror M7  on  the CO detector   The
 centers  of  curvature of mirrors M5 and M7 are near  the center of the
 «?Yrf  m ml£°r M6K8t **? position8 indicated by corresponding numbers
 (5)  and  (7).  The number of passes can be changed by re-adjusting the
 position of the center of curvature of mirror M5.

 Energy from the portion of the beam that  goes directly from mirror M4 to
mirror M7 is imaged after 2  passes near the  upper window at the position
marked with  2  .  This  portion of  the  beam passes on to the assembly of
filters, lenses and detectors  that measures  H20 concentration (see Figure 6)
                                 12

-------
     Apertures are placed  on window Wl, mirror M4,  and window W2  to  reduce
stray  light and  to improve the  stability  of  the  instrument.   The  aperture on
window Wl  is approximately 1.3  mm wide  and  5.5 mm high.   This aperture  is
focussed by mirrors M2  and M3 onto  mirror M4.  The  3.5 mm wide  aperture on
mirror M4  is more than  wide enough  to include  the enlarged  image  of  the aper-
ture on window Wl.  Thus,  the aperture  on mirror M4 does  not  restrict any of
the main beam; however, it reduces  stray  energy  that might  reach  the detector
by some path other than that of the main  part of the beam.  The aperture on
window W2  limits the height of  the  beam that can pass through the window but
does not restrict the width. At window W2  the beam is higher than it is wide,
and the full width can  pass through the window and  lens LI  to the detector.
The masks  on the top and bottom of  window W2 that form the  aperture  block rays
that might otherwise strike the edge of the  lens LI.


HJ) MONITOR
"™i

     The H20 monitor makes use  of some  of the radiant energy  that enters the
multiple-pass sample cell  but does  not  follow the path described  above.  The
beam reflected from mirror M4 is somewhat oversize  so that not all of it is
intercepted by mirror M5.   Some of  the  "extra" energy goes directly  to mirror
M7 which directs it out through window  W3 after  only two  passes of the multiple-
pass optical system.  The  concentration of H20 is relatively  high, and the ab-
sorption by the  1^0 is  sufficiently strong that  only two  passes of the sample
section provide  adequate sensitivity to the  H20.  After leaving the  sample
section through  window  W3  the beam  of radiant energy passes to the assembly of
mirrors, filters, and detectors shown in  Figure   6.  A mask  on window W3
serves the same  purpose for the energy  reaching  the H20 monitor as does the
mask on window W2 for the  CO channel.   The requirements for the stability of the
optical beam reaching the  H20 monitor are much less  stringent than those for
the beam in the  CO channel.

     When  the sample section is filled with  argon, or some other non-absorbing
gas, the outputs of detectors A and B are made equal by adjusting the relative
bias voltage across the two detectors.  When H20  is  added to  the  sample section,
the amount of radiant energy chopped at 360  Hz that  reaches detector A is dimin-
ished because of the H20 absorption in  the spectral  region passed by filter A.
The presence of  the H20 in the  sample section makes only  a very small, if any,
difference in the amount of chopped energy incident  on detector B.  A tuned
amplifier  and synchronous  demodulator measures the difference between the two
signals from the two detectors.   This difference  can be related to the concen-
tration of H20 in the sample section.

     The output  VB from detector B  is used as a  reference signal  for the syn-
chronous demodulator that  measures  (VB  -  VA).  An AGC circuit maintains the
amplified  signal from detector  B at a constant level to account for  the changes
in source  brightness, detector  temperatures, and  other factors  that may affect
the detector output.  The  gain  of the first  stage of the  amplification of the
                                     13

-------
      Det A  -
  Figure   6.  Optical diagram of the  H2
-------
                                    SECTION VI

                             SAMPLING AND CALIBRATION
 GAS HANDLING

      The air to be monitored can b'e circulated through the sampling section
 by either of two different methods.  In the first method,  both the  top  and
 side covers for the sample section are removed so that the ambient  air  is free
 to circulate through the monitoring beam.  This method has the advantages that
 no fans or pumps are required; however, it has the disadvantage that the covers
 for the sample section must be re-installed in order to flush  the sample section
 with a non-absorbing gas to adjust to zero.

      The second method of sampling is usually more satisfactory and takes place
 with the covers on the sample section.  The air being measured enters the
 sample section through a hole in the small top plate at the end near mirror M6.
 It then traverses the sample section lengthwise and exits  through holes in the
 end plate behind mirrors M5 and M7.  A small exhaust fan mounted on the end be-
 hind mirrors M5 and M7 draws the air through the sample section at  a steady rate.
 The flow rate can be adjusted by either varying the fan speed  or by varying the "
 size of the opening where the air emerges from the sample  cell.  An adjustable
 shutter on the end of the instrument is used to adjust the size of  the opening.
 The opening is completely blocked when the shutter is pushed farthest in.  The
 opening and the shutter are shaped so that the size of the opening  can be con-
 trolled with good precision when it is first being opened.   When the shutter
 is moved more than approximately 0.4" from the closed position,  the size of
 the opening increases more rapidly as the shutter is moved toward the wide-open
 position.   If it is desirable to leave the cell completely closed with no cir-
 culation for an extended period of time,  it is advisable to turn off the exhaust
 fan to decrease leaks around the shutter.   A bolt with a knurled head is threaded
 through the supporting block for the sample exhaust fan.   When this  bolt is
 tightened it is forced against the shutter to form a tighter seal.

     All of the holes in the wall of the  sample section through which screws
 extend have been potted with RTV to prevent leakage.   It is unlikely that any
 adjustments of the multiple-pass optics will be needed.  If a  minor adjustment
 is  required,  the RTV can be removed from  the screws that will  be moved.  Windows
Wl  W2  and  W3  are also bonded in place with RTV.   The gasket material on the
 covers  for  the top and backside of the sample section provides adequate seal
 for most  applications.   However, if a gas  circulation system is  used that would
cause  the pressure in the sample section  to drop  significantly below atmos-
pheric  pressure,  it  would be advisable to  provide a better seal around  the
edges  of  the cover plates.


                                     15

-------
      The  exhaust  fan  is designed for relatively high flow rates with  little
 pressure  drop  in  the  airstream.  Because  the air enters the sample section very
 near the  monitoring beam,  the rise- time of the instrument can be made very
 short.  For  example,  a small amount of gas entering the sample section near
 M6 can  cause an observable change in the  instrument output within 1 or 2 seconds,
 depending upon the rate of flow.  In some applications, it is desirable to place
 the instrument inside of a building or possibly in an enclosed vehicle, and to
 draw air  in  through a probe from the outside.  In this case, the time lag will
 depend  on the  flow rate and upon the volume of the line in the probe.  In this
 type of application,  it is also advisable to not draw the air in through the
 sample  section by the exhaust fan located on the exit end of the sample section.
 A small pump or fan should be located in  the intake line just ahead of the
 sample  section.   The  exhaust fan at the exit end of the sample section would
 be turned off  or  removed,  and the shutter on the exhaust end would be left
 slightly  open  to  allow the air to exit while maintaining the pressure in the
 sample  section very near,  or slightly above, the ambient atmospheric pressure.
 This type of circulating system would decrease the amount of air from within
 the enclosed building or vehicle that might leak into the sample section.

      The  zero-reading of the instrument is adjusted electronically when the
 sample  section is being flushed with argon, or some other non-absorbing gas.
 Nitrogen  should ordinarily not be used for this purpose because commercially
 available nitrogen usually contains a few tenths  of a  ppm of CO.   This amount  of CO
 naturally interferes  with  the zero setting.  The gas line is connected to the
 -gas inlet near mirror M6,  and the shutter on the gas outlet should be opened
 between approximately 0.2" and 0.5".  The sample exhaust fan should be turned
 off.  By  observing the signal output, it  is possible to determine when the sample
 'section has  been  completely flushed.  Standard samples of known CO concentra-
 tion are  flushed  through the sample section in the same manner in order to
 check the span calibration.

      When checking the span calibration for the H20 monitor, or when  adjusting
 the H20 correction factor, the non-absorbing gas can be bubbled through liquid
 water at  about room temperature.  With a  flow rate between  1 and 3  liters/minute
 for about 5-10 minutes, the FLjO concentration in the cell will probably increase
 to within 10%  of  the  concentration corresponding to saturation at the  temper-
 ature of  the liquid water.  Some time is  required for the initially rapid rate
 of adsorption  on  the  walls to decrease.   Less time is usually required to flush
 the H£0 mixture from  the sample cell with dry argon.
CALIBRATION CURVES

     Figure   7  shows the calibration curves for  the  four  ranges  A,  B,  C,  and
D for the instrument at the time it was delivered to  EPA.   Each  of the  four
individual span calibration potentiometers on the top panel changes  the full-
scale concentration for the corresponding channel.  The master  span calibration
potentiometer on the top panel changes the gain for each of the  four ranges
by the same factor.


                                      16

-------
100
   0.02
.   0.1
                                                  CONCENTRATION (ppm)




      Figure 7.  Calibration curves relating  the  output signal to the concentration of CO.
200

-------
                                   SECTION VII

                       ELECTRICAL CIRCUITS AND CONTROLS
     Wiring diagrams of the electrical power circuits are shown in Figures  8,
 9  and  10.   All of the circuits contained in the main instrument box are illus-
 trated in Figure  8.    The male pin connector is mounted on the front panel of
 the main instrument box.  The motor transformer supplies 60 V ac power for the
 motor  that drives the rotating gas-filter cell when operating in both the ac
 mode and the dc mode.  The 5 V dc power supply, the + 15 V dc supply, and the
 source transformer are all used in the ac mode only.  Power for the sample ex-
 haust  fan is supplied to the fan motor through a plug on the end of the instru-
 ment near the fan.  The cooling fan is powered through a terminal strip located
 near the fan.  Both the exhaust fan and the cooling fan are interchanged when
 converting from ac to dc, or vice-versa.  The ac power switch on the front
 panel  is in the active circuit only during ac operation.

     The power cables for both ac and dc operation are illustrated in Figure  9-
 The appropriate cable plugs into the male pin connector on the front panel of
•the main instrument box.  For ac operation, the end of the power cable opposite
 the main instrument plugs directly into a 115 V ac power line.  The corresponding
 end of the dc power cable plugs into the front of the power supply assembly.

     The power supply assembly and the battery cable that connects it to the
 battery pack are illustrated in Figure  10.  These components are used only
 during dc operation.  Power for the 60 Hz synchronous motor that rotates the
 gas-filter cell is supplied by the dc-to-ac converter.  The 24 V dc-to- + 15 V dc
 converter supplies power for the electronics during dc operation.  During cool
 weather, the cooling fan in the power supply assembly may not be required.  No
 tests  have been performed with the cooling fan disconnected.

     A commercial battery charger can be connected to the batteries through a
 3-pin  plug mounted on the front of the power supply assembly.  In order to charge
 the batteries, the rotary switch on the front panel of the power supply assembly
 must be in the charge position.  A small plate bolted on the front of the battery
 charger prevents the charger from being switched to the 12 V position.  For normal
 dc operation,  the rotary switch must be in the run position, and the 24 V dc  and
 6 V dc switches must be turned on.

     The wires on the battery cable are bundled in pairs.  One wire in each
 bundle has a red band near the end to indicate that it connects  to the positive
battery terminal.  The other wire of the same pair has a black band and connects
                                      18

-------
VO
          Figure  8.   Electrical wiring diagram of main instrument box.

-------
<_r
   H
   M
   T
Key *t A

      e/vt>
              BWE
             RED /i/««K
             ORftK&E /h/*cx
                BJ-WCK
                dc power  cable
                                                    CewutcToK
                                                     SHfU.
                                                      H
                                                      T
                                                       Pin
flC to«(7is
NeUTRlH.
KC $ 0i»CK


,r
j *

C

BuiLDma- GKEeu
/I
_8_
C
D
£
P
S
H
J
f<
L
M
N
P
R
5
T
r^
                                                            ac  power cable
                                                                            Key/»r ft
  Figure   9.   Wiring  diagrams  for dc and  ac  power cables.

-------
Bl

62

33

84

85
+ 8L(JE/iuJ>X*
T -t GR.EGN
+0 Ol?fll\IG6
I -, BLUE
+ If/// re/i/.,^
T~-, RED/A1***
+_ GREEN/^ck

+ F Bl~V£/i,,'*cK

SLACK
FEMALE PIN CONNECTOR
in?v RRTUKP.N K and L
^^^•H
fl
8
C
D
E
F
H
,1
K
L
M
N
P
R
S
T
                                                                      TO CHARGER
FEMALE PIN
RECEPTACLE
KEY AT A
             Figure   10.  Wiring diagram of  the battery cable and the power  converter assembly.

-------
 to the negative terminal of the same battery.   Battery No.  1  powers  the  detector
 cooler and the radiant energy source.  By using two batteries connected  in  parallel
 in the place of battery No. 1, the maximum operating time  from fully charged
 batteries can be increased from approximately  4 hours to 8 hours.  Batteries  2,
 3, 4 and 5 provide power for the other components  for approximately  8 hours.   The
 total power consumed during dc operation is approximately  42  watts.


     Most  of  the electronic circuitry is contained on the four circuit cards in-
 dicated  in the  lower panel  of Figure  1.  The small preamplifier card is located
 near the detector and the bias battery, and the  larger, main electronics card is
 mounted  on the  underneath side of the top panel.  All of the controls on the main
 card are easily accessible when the top cover is inverted as shown in the lower
 panel  of Figure 1.  All of  the electronic components used to process the signals
 from the two  detectors in the 1^0 channel are contained on the l^O electronics
 card,  which is  mounted to the baseplate beneath  the main electronics card.  The
 reference  card  processes the signal from the reference pick-up and supplies a
 30 Hz  reference signal for  the synchronous demodulator in the main electronics.

     The primary purpose of the main electronic  circuits is to process the signal
 from the detector in the CO channel and to produce a dc output signal that is pro-
 portional  to  Va/Vc, the ratio of the detector signal components at the correspond-
 ing freouencies f a  (30 Hz) anf fc  (360 Hz).  The 30 Hz component is processed by
 a  synchronous demodulator as a modulation of the 360 Hz carrier signal.  An auto-
 matic  gain control  circuit maintains Vc, the amplified component of the carrier
 signal at  a constant level.  The instrument output is proportional to Vg, the
 amplified  signal resulting from the 30 Hz modulation.  Because Vc is constant,
 this output is  therefore proportional to the ratio Vg/Vc, which is directly re-
 lated  to the  concentration of CO in the sample cell.

     The output signal can be read from the meter on the top panel or from an
 output jack designed for use with a strip-chart  recorder.  A switch and potentio-
 meter mounted near  the output jack make it possible to vary the signal at the
 output jack corresponding to full-scale of the panel meter from approximately
 0.01 volts to 10 volts.  The electronic time constant can be switched to either
 0.3, 1, 3  or  10 seconds.

     The span calibration is checked or adjusted with standard samples of CO  in
 argon  or CO in clean air according to the following:  Four different ranges of
 sensitivity are available; range A is the most sensitive and  is used for the  lowest
CO concentrations; and range D is the least sensitive.  The range switch indicates
 the range being used.  At the time the instrument was delivered to EPA,  full  scale
meter deflection (FS) corresponded approximately to the following CO concentrations:
Range A,  2 ppm; B, 5 ppm; C, 20 ppm.  Because of excessive non-linearity between
the signal output and the CO concentration for higher concentrations,  it was  not
                                     22

-------
practical to adjust range  D  to match a fixed concentration.  Instead, half-scale
was adjusted to  correspond to 50 ppm.  Within certain restrictions, the span
potentiometers for each  range can be re-adjusted for other sensitivities.  Cali-
bration curves for the instrument at the time it was delivered to EPA are shown
in Figure 7 „

     The electronics  for the H20 monitor produce a dc signal proportional to
(VB ~ VA)/VB> where the  voltages Vg and V^ correspond to the amplified signals
from detectors B and  A,  respectively  (see Figure  6).  This voltage ratio is pro-
portional to the H20  concentration.  An automatic gain control circuit maintains
VB at a constant level and produces the same fractional change\in the amplifica-
tion of the signals from both detectors.  Thus, the ratio  (Vg - VA)/VB is propor-
tional to the difference between the  two amplified detector signals and is, to a
good approximation, independent of  source brightness, dirt on windows, etc.

     The output  of the 1^0 monitor  is  displayed on a panel meter and on an output
jack similar to  the ones in  the main CO channel.  By adjusting the H20 span poten-
tiometer, full-scale  reading of the 1^0 panel meter can be varied from less than
1% H20 to more than 10%  H20. Another  potentiometer associated with the 1^0 moni-
tor is used to vary the  "correction factor", which is the  fraction of output of
the H20 monitor  that  is  fed  into the CO channel.  This correction signal is ad-
justed to automatically  account for a  small amount of interference by 1^0 in the
CO channel. By  carefully  making this  adjustment, any remaining error in the CO
channel due to the interference by  atmospheric H20 can be  made less than 0.01 ppm
of CO.
                                       23

-------
TECHNICAL REPORT DATA
(I'lcasc read fnaructions on the reverse before completing}
1 REPORT NO. 2.
4 TITLE AND SUBTITLE
AMBIENT CO MONITOR
7. AUTHOR(S)
Darrell E. Burch, Francis J. Gates, John D. Pembrook
9 PERFORMING ORGANIZATION NAME AND ADDRESS
Aeronutronic Ford Corporation
Aeronutronic Division
Ford Road
Newport Beach, California 92663
1? SPONSORING AGENCY NAME AND ADDRESS
Environmental Sciences Research Laboratory
Office of Research and Development
U. S. Environmental Protection Agency
Research Triangle Park, N.C., 20011
3. RECIPIENT'S. ACCESSION-NO.
5. REPORT DATE
July 1976
6. PERFORMING ORGANIZATION 1 ODL
8. PERFORMING ORGANISATION fiEPORT N.
U-6210
10. PRCiGRAM LLEMENT NO.
11. CONTRACT/GRANT NO.
68-02-2219
13. TYPE OF REPORT AND I'ERIOD COVERCI
Final
14. SPONSORING AGENCY CODE
EPA-ORD
15 SUPPLEMENTARY NOTES
 16. ABSTRACT
        A portable instrument has been designed and two units have been built to monitor
   the concentration of CO in ambient air.   The air flows through a sampling section
   that is approximately 43 cm long with a  28-pass optical system that produces a total
   path of 12 meters.  Gas-filter correlation methods are employed for the detection
   and discrimination against other gas species in the air.  An H^O monitor built into
   the .main instrument measures the concentration of I^O and automatically accounts
   for a small amount of interference by this gas.  Interference by all other atmos-
   pheric gases is negligible.  The minimum detectable concentration of CO is less
   than 0.02 ppm.   The instrument is powered either by batteries or by a battery pack.
17. KEY WORDS AND DOCUMENT ANALYSIS
.1 DESCRIPTORS
Air Pollution
Carbon Monoxide
; < ',i ' iMH' . ! IQN ;. f A 1 L-MI NT
Release to public
h.lDENTIFIF.RS/OPEN ENDED TERMS
Gas-Filter Correlation
19. SECURI 1 Y CLASS (This Hi-port)
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COSATI l-'ich. /Group

21. NO. OF HAG' S
28
22. PRICE
fPA Corm 3220-1 (9-73)
                                         24

-------