NONIUS CAD4/MACH3 User manual

Introduction
Setting the detection unit
    Setting a reflection
    Fine adjustment of the high voltage
    Analyses of pulse height distribution and window size
    Analysis of the pulse height distribution with closed shutter
    Setting dead time correction
Dead time and filter factor determination

Introduction

The counting chain of the CAD4 consists of a number of parts which are depicted in the scheme given below.
 

Photomultiplier

Pulse shaper

Pre-amplifier

Amplifier

Analyser

Ratemeter

The detector consists of a sodium-iodide scintillation crystal and a photo multiplier. When an X-ray quantum is absorbed in the crystal, a number of light photons are emitted. Most of these photons are converted into electrical pulses by the photo multiplier and magnified. The photo multiplier current pulses are shaped by a shaping network, which is connected to the input of the pre-amplifier. In the analyser (which is an integral part of the CAD4 interface), the pulses of the amplifier are processed depending on the lower level and window settings. The X-ray quanta, however, appear at random times, whereas the detection equipment needs a finite time to process the detected quanta. This introduces a dead-time, since the quanta appearing during the processing time cannot be detected.
The result is a loss of counts, generally called dead-time loss or coincidence loss. lt increases at increasing count rates. To compensate for this loss the software emulates a dead-time correction circuit, which continuously adds a number of counts depending on the count rate detected.

With a correctly set counting chain the CAD4 will operate with less than a 1% dead-time loss up to 50,000 cps. In the following sections some recommendations for proper setting and using of the radiation detection chain are given. It should be emphasized that the radiation detection unit might be damaged, if it is exposed to too high intensities of X-rays for too long periods.

Setting the detection unit

It is assumed that all cable connections are made correctly. The shaping mode of the detector is factory pre-adjusted.

Setting a reflection

Setting the radiation detection unit is achieved by use of a reflection of the testcrystal or any other crystal of good quality. The orientation matrix of the testcrystal is known, thus in that case it is relatively easy to find a strong reflection. First the lower level (LL) should be set to 150 (using the pocket terminal) and the window (WD) is set to 1125.

Then all kinds of quanta are detected. The high voltage (HV) is set to 300 V. The reflection chosen, generally the 002 or 004 of the test crystal (see section F. of Chapter XII) should be able to produce an intensity of about 50,000 c/s with this LL and WD setting. With a HV of 300 V, however hardly any counts will be registered.

Fine adjustment of the high voltage

The LL should be set to 200 (a value well above the noise level) and the window to 1000 through the pocket terminal. Gradually increase the high voltage (starting at 300 V) using the pocket terminal and record the count rate displayed by pressing the IA key. Make a plot of count rate vs. the HY. With increasing HV the count rate detected in the window sequentially increases, reaches a plateau, decreases a little bit and finally increases. The HV setting for the middle of the plateau is the required setting. Usual values fall in the range 300-800 V.

Analyses of the pulse-height distribution and window size

This is done in a number of steps, viz.:

• Determination of the intensity-distribution as a function of pulse height (called the pulse-height distribution). A small window (WD) preferably 10 or 20 is scanned through the entire amplitude range (using the LL command on the pocket terminal) and the count rates measured are noted. A plot of the results with a window of 50 is given in Fig. XIII. 1.

Fig. XIII.1 Plot of count rate vs. window (WD)
 

• There should be a dip in the curve (Fig. XIII.1) between level 0 and 450. This dip should be as low as possible. If it is not low, this may have various reasons:

• The HV setting is too high, causing the photo multiplier to be over excited.
• The optical coupling between the scintillation crystal and the surface of the photo multiplier is not good. The space between the scintillation crystal and photo multiplier surface may be inhomogeneously filled with optical grease. Such a problem may be solved by pressing the crystal onto the photo multiplier surface (by carefully tightening the front ring.)
• The pre- and main-amplifier produce noise pulses in the same amplitude range as the pulses of the X-ray quanta. In this case, service from your local service organization is required.
• The scintillation crystal has blind spots. If the detector is moved across the reflection by a 2-theta scan, the non-uniformity of the detector surface will show up.
• The quality of the pulse-height distribution can be expressed in a quality factor, being

(A₁-A₂)/[0.5*( A₁+A₂)]. The magnitude of the quality factor should be leas than 0.5.
• Now, the optimal LL and WD setting must be determined from the pulse-height distribution curve. The window should contain the whole peak including a bit of the peak-bases. Care should be taken not to choose too small a window, because there might be some drift.

• Repeat the procedure described in section B.3 (I) with the shutter closed. Make a plot of the results and compare it with your version of Fig. XIII.1.

• Measure the background activity in the window chosen in section B.3.iv. with a closed shutter. This background activity should be only a few counts or none.

At high count rates two pulses might overlap, this causes a doubling of the pulse-height. Such an amplitude, however, falls outside the window. This effect is called dead-time loss. The system contains software to compensate for this loss of counts. The attenuation factor of the X-ray attenuator should be independent of the intensity lever; if this can be confirmed, the dead-time compensation is correctly set. The 004 reflection of the test crystal (c-axis 11.07 A) can be used to check the dead-time. When the test crystal is too small, another crystal should be used, since it is essential to do a dead-time adjustment using a reflection which produces enough intensity.

The results of the adjustment procedure should be well documented and saved, together with the crystal used, for comparison with results of later adjustments.

Dead-time and filter factor determination

Using combinations of stationary scans with a large (9mm) aperture or small aperture (horizontal alit), both with or without attenuator it is possible to calculate the values of dead-time and filter factor.

Theory and description

When M is the measured value of the true intensity I and p is the dead-time, it is assumed that: I=MI(l-p*M).

Since this formula also holds for measurements with a filter (M’) the following formulas apply:

I’=M’/(1-p*M’) and I=F*I' where F equals the attenuation factor.

Thus, the following equation remains:

M’*F/(1-p*M’)=M/(1-p*M)

From this a linear equation in 1/M and 1/M’ can be derived. All measurements are used to do a least ;squares calculation using this algorithm improving the accuracy with each cycle. From the slope and interest of this line the actual values for dead-time and attenuator factor can be calculated.

In the first cycle 4 measurements are done with the fixed time ratio:

1 : 4 for hole : slit

1 : 20 for without : with filter

In subsequent cycles these ratios are changed to obtain an equal amount of counts in each measurement, however, the total measuring time for one cycle will be kept constant.

With the intensities registered during the four measurements the program updates a number of values:

Operation of the DEADT program

Normal dead-time values are about 1 microseconds

Meaningful dead-time measurements can only be done with a sufficiently strong reflection (40.000 - 60.000 counts/sec for 9mm aperture) so that events of two or more pulses reaching the counter within the dead-time can be expected to occur frequently enough to be measured.

Note: Since every measurement will be corrected for dead-time, it is necessary to reset the dead-time to zero before starting.

Procedure:

• Set the dead-time to zero, using the CASPAR procedure.
• Set SR=7600.
• Center and position a strong reflection.
• Start the DEADT program and let it run for several cycles.
• Stop the measurement,;usi-ng,SR=2001, when you are satisfied with the standard deviations printed after the last cycle.
• Fill in the calculated dead-time using CASPAR.
• Insert the filter factor using WI.

The program may be interrupted by setting SR=XXXI or SR=2XX1. Answering IN' to 'GO AGAIN?’ will cause a printout of the results from the last cycle, if no output had been selected.

The message ‘STRONG' will be printed if the buffer counter capacity is exceeded (i.e. the intensity ¾ 254 counts/2.5 msec). The measurements are repeated until I(hole) can be determined without overflow Terminal output is controlled by the switch register: SR=2XXX (e.g. 2000)-cycle output.

Example:

With SR=2000 act there will be output on the terminal.

CDO>
DEADT<CR>
Time may be entered if crystal in reflecting position 200<CR>
 

Cycle	Filter	dfilt	Dead-t	dDead	1(HS,At)	I(HS)	I(Hole)	I(Hole,At)
1	19.337	0.000	2.251	0.000	179.4	3444 4	14741.8	787.1
2	19.069	0.211	1.282	1.633	179.3	3368.4	14614.2	780.3
3	19.018	0.128	1.119	1.001	178.1	3359.9	14600.0	780.4
20	19.047	0.039	1.445	0.302	179.7	3370.3	14743.6	795.3

SWITCH 1 1
/<CR>
GO AGAIN?
N<CR>
CDO>
DEADT<CR> ;specifies maximum time ;SR=OOOO, selects no output.
Time may be entered if crystal in reflecting position
<CR>
;SR=2001, selects output and halt

SWITCH 11
/<CR>
GO AGAIN?
N<CR>
 

Cycle	Filter	dFilt	Dead-t	dDead	I(HS, At)	I (HS)	I(Hole)	I(Hole,At)
35	19.013	0.007	1.190	0.052	181.7	3446.	115019.	7804.9

CDO>

Previous:  Alignment of the Diffractometer
Next: Automatic Measurement Procedure

---

(C) Nonius BV, 1999, 2000. Last change 21 December 2000