|
11. Calibrations for Energy vs Channel, NaI(Tl) Detectors
We are now ready to perform a sodium iodide energy calibration. The Energy Calibration program is reached by selecting the “Calibration” option from the Main Menu, see Figure 11-1, below, which shows the main menu. Please remember that this menu is the same for all “HELGE” software.
There are six steps which must be performed to assure that the calibration data are obtained properly:
The menu which you receive is a function of the type of hardware being used. Figures 11-2 through Figure 11-4 show the menus from:
the “Do-It-Yourself Whole Body Counter,” The screens for the Quicky I, Quicky III, and Quicky VI are identical because they do not have any moving parts, i.e., there is no "Speed" calibration. These three screens are shown below.
The "Energy vs. Channel" program may be protected by demanding a password before allowing the technician to use the program as shown in Figure 11-5 below.
11.1. Energy Calibration, Part 1, Verification of Dial Settings Upon entering the sodium iodide energy calibration routine, you are presented with the screen shown below. Examine it. It shows that we are looking for the 241-americium photopeak of 59.6 keV in channel 14.282. Likewise, we are looking for a high energy photopeak from 60-cobalt at 1173.2 keV, which should be found in channel 230.370. Remember that we are accounting for the non-linear response of NaI(Tl) detectors in determining the center channel values. Please refer to Chapter 10 for a complete discussion of the non-linear features of NaI(Tl). Each of the preamplifiers has a dial position recorded in the parameters file. The computer prints these values and asks if the values are correct. You should inspect each of the preamplifiers to see whether or not the dial is actually set at these numbers. You have three choices. If the dials are all set correctly you can answer “YES” and the program will proceed. If the dials are not set at these settings you have the other two choices:
Before you respond “Yes” to this screen, be sure that the energy versus channel calibration
sources are in front of the detector(s) at the reference position(s). Please
refer to Chapter 10 for further information.
11.2.
Energy Calibration,
Part 2, Calibration of Detectors 1 Through “n.”
The more recent versions of the “HELGE PC” software (starting with Version 2.0, we are now on Version 5.0 for DOS and the w\Windows version is almost finished) allow multiple detectors to be calibrated simultaneously. The actual calibration is accomplished by placing check sources in front of each of the detectors, which you should have done before you reached this screen.
Figure 11-7, below, shows the spectrum as it is being accumulated. In this particular
example the total counting time was 60 seconds. We see that we still have
21 seconds to complete the data acquisition. The total sampling time comes
from the parameters. See Figure 9-17.
Note the red markers. These are located in the channels specified in the “Examine/Change
Parameters” menu, Option 8, as shown in Figure 9-17.
Typically, the sources used are 241-americium and 60-cobalt. Therefore,
the program will read the parameters value for the correct center channels
and place markers in the integer value (rounded off to the nearest whole
channel). Thus, in Figure 11-7 we
see red markers in channels 14 and 230. The marker in channel 230 will
give us an idea immediately of where the lower energy peak of 60-cobalt
should fall. It is very important that we look at this graph. In the past
there have been a number of instances where the operator has not paid attention
to the position of the 60-cobalt photopeaks and, because of carelessness,
has allowed the 1332.5 keV photopeak to fall in the channel where the 1173.2
keV photopeak should fall. Obviously, this would give poor results. By
using the “Up Arrow” or “Down Arrow” you may look at each detector
in a multi-detector system. Don’t forget to check each detector for the
proper location of the photopeak with reference to the red marker.
Now look at Figure 11-8, below. At the end of the counting time the computer prints the results. If this figure could be shown in color on the printed page, we would see that the lines for detectors 1 and 3 would be high-lighted in yellow against a red background. This is because both of these detector dial settings need adjustment. The 241-americium photopeak for detector 1 was found in channel 14.23 and the 60-cobalt photopeak was found in channel 233.64. This is 3.27 channels too high, as the data for detector 1 shows. The limits of deviation from the correct 241-americium and 60-cobalt photopeak center are found in the “Examine/Change Parameters” menu, Option 8, Figure 9-17, lines 8 and 9. The new dial setting has been calculated to be 4.84 compared to its original setting of 4.94 (see Figure 11-6). The calculation is very simple: divide the number of channels by which the photopeak center has deviated from the correct channel, -3.27 channels for 60-cobalt, by its dial factor, which was 32.4 in this example. Thus:
This is the change in dial setting, so if we add the original dial setting of 4.94 we obtain
The same calculation is done for detector 3. The two lines at the bottom of the screen tell
us that we must change the dial settings (for detectors 1 and 3, the ones
highlighted in yellow) and press “F2” to calibrate again.
Stated in outline form the computer does the following: a. subtract the center channel from the correct center channel The computer found the center channel at 233.64 but the correct center channel is 230.37. Therefore, for this first step we have: b. divide the channel difference by the dial factor The dial factor is 32.4 channels per 1.0 revolutions of the dial. Therefore, the difference in dial setting is: c. add this quotient to the old dial setting to obtain the new dial setting: Thus,
Figure 11-9,
below, shows the next screen, the one obtained after we made the adjustments
shown on the previous page. This is not a screen which you would
expect to see frequently. The significant part of this screen is the fact
that we have a different note at the bottom of the red portion:
If you could see
this screen in color, you would note that the only portion in yellow was
the 241-americium center channel for detector 1 which shows a value of
13.74. This is lower by 0.54 channels than the desired center channel of
14.28 - the difference is 0.54 channels. If you look at Figure 9-17, line 8 (at least for this particular set of parameters) the acceptable
error for the 241-americium center channel is 0.5. Therefore, this result
is 0.04 channels too low. For this example we have deliberately made very
careful adjustments to the ADC “Zero” control. This is something which
an inexperienced technician should not do. The limit of deviation
from the correct 241-americium photopeak center can be as much as 1.0 channels
without seriously affecting the gain and zero shift.
After we made
the careful adjustment, we pressed “F2” and continued.
Figure 11-10,
above, shows the next screen, the one obtained after we made the adjustments
shown on the previous page. This time detectors 1 and 4 were highlighted
in yellow, signifying that adjustments of the dial settings were necessary.
After making them, we obtained the screen shown in Figure 11-11,
above. Note that the background screen for the results is blue. This signifies
that all detectors are within their stated parameters. The “Energy versus
Channel” calibration is finished.
11.3.
Calibration, Part 4, Recording the Data
As the calibrations were being performed, all important information was being transmitted to the printer. When the calibrations are complete, as shown at the bottom of Figure 11-11, above, and you must press the “ESC” key. This will instruct the program to save the new dial settings. The printer will complete the printing. This information is valuable for your Quality Assurance records. It is also valuable for Helgeson Maintenance Personnel in monitoring the performance of the system. An example of this printed record is shown in Table 11-1, immediately below this paragraph, and in Tables 11-2 and 11-3, below.
Table 11-1 corresponds to the problem which we saw at the bottom of Figure 11-9. As we stated, this is an unusual condition but one which should be recorded in the Quality Assurance records.
The data in Table 11-2, below, is much more typical
of the calibration record. Note that only two iterations were necessary
for the four detectors to be returned to their properly calibrated conditions.
11.4.
Energy Calibration, Part 5, Obtaining a New Background
Now that the counter has been calibrated, you must obtain a new background. Remove the
source from the counter and store it safely away from the counter. Select
the “Quick Count” option of the Data Acquisition Program (Option 1 from
the “Main Menu”), and obtain a normal background.
11.5.
Statistical
Errors in Channel Center Measurements
We have already
seen in Chapter 3 that statistical errors are an expected part of all nuclear
counting. This is also true of the determination of the center channel
of a photopeak. Each channel used for the determination of the center of
the photopeak will have errors which are proportional to the square root
of the number of counts observed in that channel. Please refer to Chapter
6 for a discussion of the method for finding the center of a photopeak.
Without going further into a difficult statistical explanation of the errors
to be found in the determinations of the center of a photopeak, it is obvious
that the errors will be a function of the number of counts obtained in
the measurement. Thus, if you are using a weak source and are counting
for a short period of time, the determination of the center will contain
relatively large errors. Therefore, you should count for a reasonably long
time. As a rough rule of thumb you should obtain at least 1000 counts in
the center channel because then the counting error in that channel will
be
Table 11-4,
below, shows data obtained from a series of 45-second counts of an 850
nCi 60-cobalt source located in front of Detector #1 of a “Quicky III.”
Figure 11-12, below, shows a graph of the same data. Note that the fluctuation in the center channel is small but real. The reader is invited to make similar measurements of his own using different counting times to see the influence of time on the determination of the center channel. This may be done in “background” mode or you may obtain the “DRIFT” program from Helgeson Scientific Services which allows you to select the counting time, the delay between counts, and the number of tests to be made. The output is a table of data points for the various photopeaks found as well as the apparent resolution of these photopeaks.
Figure 11-13, below, shows the long-term drift of an NaI(Tl) detector if it is continually subjected
to a strong source of radioactivity. It will recover if the source is removed.
11.6. Energy Calibration,Part 6, Determining the “Dial Factors”
The “dial factors” are defined as the number by which the difference in the observed photopeak channel from the standard photopeak channel is divided to obtain the dial setting change. This value, the dial setting change, is added to the old dial setting to obtain the new dial setting. The general method is performed in several steps:
11.6.1. Gain Setting
The gain is set at a nominal 5 keV/channel for all Helgeson counting systems. This does
not mean that other gain settings could not be used, however, for most
work with mixed fission, activation, and corrosion products, this is a
convenient gain. If you are working mainly with natural thorium and its
daughters you may wish to re-calibrate your equipment to use a gain setting
of about 6 keV/channel since this will allow the 2.615 MeV photopeak
of 208-thallium to be visualized completely.
11.6.2. Sources
Most of the time you will use the standard 60-cobalt sources, although you may wish to use
40-potassium in some form since no license from the Nuclear Regulatory
Commission or any state regulatory body is required. The typical 0.5 to
1.0 microcurie 60-cobalt source will generally give a sufficient number
of gamma rays at 1.1732 MeV to allow a calibration to be made in 30 to
60 seconds. The trade-marked
products “Lite Salt” or “No Salt,” contain a mixture of ordinary table
salt (NaCl) and potassium chloride (KCl). An 11 ounce (311 gram) package
of “Lite Salt” contains “550 mg sodium and 73 mg potassium...in each one-half
teaspoon.” These containers have typical dimensions of 2.5 inches diameter
by 5.375 inches high (approximately 65 by 135 mm). Thus, about six packages
of these containers stacked in a postal mailing tube provides a column
of 40-potassium which may be placed in from of the detectors, allowing
simultaneous calibration of all detectors. This type of source will give
a sufficient number of gamma rays at 1.4607 MeV to allow a calibration
to be made in one to two minutes.
Thorium gas mantles may also be used since they provide a multitude of gamma rays. As in the case of the 40-potassium, they may be mounted on a column device so each detector views a source. One envelope, which contains two mantles, mounted in front of each detector will generally give a sufficient number of gamma rays at 2.615 MeV to allow a calibration to be made in one to two minutes. Regardless of which sources you use the establishment of the gain factors is done in exactly the same manner.
11.6.3. Data Acquisition
There is no programmed method for determining the dial calibration factors since they are determined
so infrequently. The best way to accumulate the data is to do it in Background
Mode. Let us assume that the detector, amplifier, and high voltage settings are approximately correct. This means that the photopeaks should be found near the following channels:
Approximate
241-Americium 14 59.537 14.269 60-Cobalt 230 1173.200 230.408 60-Cobalt 261 1332.500 260.510 40-Potassium 285 1460.800 284.670 11.6.4. Data Accumulation a. Place the source(s) in front of the detector(s). b. Start data accumulation by typing “ALT-R.” c. Find the Center of the Photopeak After about 30 seconds, or when there is a good photopeak, move the variable cursor to the apparent center of the photopeak, press “Page Up” to mark the photopeak, then press “P” (for peak information) and “1” (for marker 1) and read the center channel in the center of the screen, assume it is channel 234.5 for this example. Record the dial setting and the center channel.
d. Repeat the previous four steps at about 5 or 6 different dial settings, where the first dial setting
is that in which the center channel is closest to the ideal channel and
the other five or six consist of two dial settings which are approximately 0.5
and 1.0 dial turns below the ideal channel and two dial settings which
are approximately 0.5 and 1.0 dial turns above the ideal channel.
Table 11-5, above, and Figure 11-14, below, show the degree of non-linearity which you might expect from one of the older amplifiers. Note that this shows the change in the 60-cobalt 1.17
MeV center channel over 40 percent of the entire dial capacity, a condition
which is not observed frequently, since most of the dial changes occur
within 1 to 1.5 turns of the dial.
Now let us look at the linear amplifiers used in the newer Helgeson equipment.
Table 11-6, above, and Figure 11-15, below, show
the excellent linearity which you will observe from one of our newer amplifiers.
Note that we cannot easily distinguish the difference between the plot
of the original data and the plot of the calculated data. A correlation
coefficient of 0.999863 explains why these values are so good.
11.6.5. Calculate the New Dial Factor The calculations for determining the new dial factor may be made by several simple methods:
11.6.5.1.1 Example from a “Reciprocal” Gain vs Channel Amplifier
Let us use the data of Table 11-5 for
our first example. We can see from the graph that the curve is not linear,
which means that we must choose the range within which we will make our
calculations. If we use dial settings from 4.5 through 5.5, we will have
five data sets. This should be sufficient to determine the correct dial
factor. Since the method is the important part to learn, we shall use only
the data from Detector 1 shown below in Table 11-7.
The third and
fourth columns contain the differences between the dial settings and the
center channels, respectively. The fifth (far right) column is obtained
by dividing the channel differences by their corresponding dial differences.
The bottom number in the fifth column is the arithmetic average of the
four estimates of the dial factor. Therefore, you may now go to the “Examine/Change
Parameters” selection from the Main Menu, select option 8, “Calibration,”
and enter the correct dial factor for the particular detector you have
just calibrated.
11.6.5.1.2. Example from
a “Linear” Gain vs Channel Amplifier
Let us use the data of Table 11-6 and Figure 11-15 for our second example. We can see from the graph that the curve is very linear, which means that we may use all of the data for making our calculations. These calculations are shown in Table 11-8, below.
The third and fourth columns contain the differences between the dial settings and the
center channels, respectively. The fifth (far right) column is obtained
by dividing the channel differences by their corresponding dial differences.
The bottom number in the fifth column is the arithmetic average of the
four estimates of the dial factor. Therefore, you may now go to the “Examine/Change
Parameters” selection from the Main Menu, select option 8, “Calibration,”
and enter the correct dial factor for the particular detector you have
just calibrated.
11.6.5.2. Dial Factors by Least Squares Calculations The Theory of Least Squares is presented fully in Chapter 36. Please refer to this chapter for a complete discussion of the operations presented here.
Last Update: 1-May-2003 |
.
.
