3.3.10 Finding the beginning of the transient

The first few signal samples derive from the recovery time constant of the signal source (capacitance) meter overload due to the filling pulse applied to the sample. These data have no physical significance and so before the transient is sent for the calculation they have to be removed, otherwise the numerical methods may give erroneous results. It is difficult to automate this process reliably because the recovery time depends on many factors which are, at this stage, impossible to quantify in general. These factors include the sample electrical or optical characteristics, the pulse parameters (and the lead length between the capacitance meter and the sample for the DLTS measurements). In the version 3.2 there is a testing routine for automated transient beginning removal implemented. However, their results are only for user information. If it is concluded that this routine works reliably it will be implemented permanently. Consequently, it is recommended to select the beginning of the transient manually. This can be done in two ways dependent on selections made in the Customise form (accessed from the Configuration menu). However a facility exists within the software to select an option which, when undertaking a measurement (not retrieving a stored file), displays and processes only the section of the transient occurring after the pre-defined delay. These three options are described below detailing the two manual methods first.

 

Manual Selection of Start Point (Default Mode)

NB Default mode is defined as the following settings in the Customise form (Configuration menu). From the tab ‘Other’ only the top box on the right of the form is checked (Right button up action enabled).

 

Press the Beginning of the transient button. The program will zoom into the initial part of the transient where the overload behaviour can be expected. In this mode the plot shows individual data points instead of a continuous curve. Use the right mouse button to drag the Beginning

marker along the transient to the signal sample which seems to be the first sample of the genuine transient (after the overload) and drop it there. The marker line can be moved repeatedly in the same way if the choices are incorrect. It is also possible to look at the entire transient again by unzooming (toolbar button or left mouse click). As an aid to visualising the region of the transient displayed on the graph in relation to the entire transient a marker bar is provided just below the main graph. The displayed region appears in blue.

 

If the Keep the beginning button (i.e. keep the chosen beginning point) is pressed the signal sample indicated during the above procedure as the starting point is remembered. In this case the Unzoom transient action does not reveal this unwanted initial part of the transient.

 

From this point onwards all calculation and storing events will normally be undertaken on the transient with the initial (overload) section removed. However, it is possible to choose to ALWAYS store the entire transient. This is done from the Customise form (Configuration menu) on the tab ‘Other’ by checking the box ‘store and hide’. When such a stored transient is reloaded the start of transient point previously selected is remembered and only the selected portion is displayed and processed. However this initial part of the transient can be seen on the Main graph by unpressing the Keep the beginning button and redrawing the graph.

 

Manual Selection of Start Point (Option Mode)

NB Option mode is defined as the following settings in the Customise form (Configuration menu). From the tab ‘Other’ none of the boxes on the right of the form is checked.

 

In this mode the cursor defining the start of the transient can be moved either by dragging and dropping with the right mouse button as described above or by clicking the appropriate point on the Main Graph with the right mouse button. In this latter case the Beginning button on the toolbar has to be set to the unpressed position to complete the procedure of defining the start point for calculation. During these procedures it is possible to use the left mouse button for zooming and unzooming the graph (see the Mouse events paragraph for more details).

 

It is possible to choose to store the entire transient within this mode as described above.

 

Fixed choice of Start Point

NB Fixed mode is defined as the following settings in the Customise form (Configuration menu). From the tab ‘Other’ the bottom box on the right of the form is checked (Fixed starting point choice)

 

This mode should only be used if the value of the overload transient can be quantified and is likely to remain constant during the measurement series. (e.g. same hardware, similar samples and excitation conditions). In this case after the measurement is completed the initial part of the transient is removed from the Main graph (not from the memory). In order to make this function reliable, one has to experimentally define a delay between the end of the filling pulse and the beginning of the transient. This delay is set by the scroll bar in the Customise form (Configuration menu) from the tab ‘Other’. The values displayed are in signal samples and in microseconds. This initial part of the transient is removed from the graph, and will not be used in the calculations so it is possible to go directly from measurement to calculation. However if the user wishes the fixed removal of the start can be undone by an unzoom action with the left mouse button (unzooming from the menu or toolbar button does NOT undo the fixed start point removal).

 

In this mode all the manual actions and options are operative i.e. it is possible to select a new start point for the analysis manually … to store the entire transient as an option etc.