Set measurement channel attributes – the toolbar button

This window (invoked by selecting the Set Channel Attributes item of the Measurement menu or by pressing the above toolbar button) is used to set the basic attributes of the measurement channels, see Figure 5-5.

Figure 5-5: The Channel Attributes window

The controls in this window are as follows:

Main buttons

select one of the channels as main input channel. The measured value on the main channel will be displayed while recording the thermal response.

Driving point buttons

set to ON if the sensor is near the power drive point on the chip (only one of these buttons can be ON at a time). If switched off, transfer impedances will be calculated.

Diode buttons

select if the sensor is of diode type

Thermocouple buttons

select if the sensor is thermocouple or infrared sensor. J, K or T types can be selected. These buttons have no effect without the thermocouple preamplifiers installed.

Sensitivity (input field)

Enter the absolute value of sensitivity in [V/K]. (Approximately 2.0 mV/K for silicon diodes at room temperature). This value can be automatically measured by MicReD's thermostat unit attcahed to T3Ster - see details in section 6.4.2.

The group of controls named 1-1000 us correction can be used for measuring slow diodes. Quite frequently the measured transient curves are resulting from parasitic electrical transients (such as the effect of pull-off currents in large diodes) and the thermal transients, superimposed.

Basically there are two cases when electrical parasitic transients (inherent in the measurement of e.g. large diodes with a huge amount of charge to be pulled out from the junction when switching off):

A. During the measurement, the measured (thermal) transient curve distorted by parasitic electrical transient remains within the measurement range. In this case the distorted curve possesses a definite minimum, see Figure 5-6. The curve on the left of that minimum is mainly due to the distorting electrical parasitic, while the curve on the right is definitely a thermal transient.

Figure 5-6: Measured thermal transient distorted by parasitic electrical transients possessing a minimum

The strategy of separating the thermal transient from the electrical parasitic is to chop off the part of the curve that is on the left of the minimum point and replace it with a constant value (taken as the measured value at the minimum point).

B. In the second case, the electrical transient has a large under-shoot, the actual transient curve runs below the lower limit of the A/D converter. That is why the measured curve is distorted, see Figure 5-7.

Figure 5-7: Measured thermal transient distorted by parasitic electrical transients possessing an inflection point

In this case, since the curvature of one transient is concave whereas the curvature of the other one is convex, there is surely an inflection point in the distorted thermal transient. The part of the measured curve left from the inflection point belongs to the parasitic electrical transient, the part on the right definitely corresponds to the thermal transient.

The strategy of separating the thermal transient from the electrical parasitic in this case is to chop off the part of the curve that is on the left of the inflection point and replace it with a constant value (taken as the measured value at the inflection point).

Figure 5-8: Raw data of a trial measurement to identify parameters for the correction algorithm

The controls in the 1-1000 us correction group can be used to set the parameters of the correction strategy that is to be used to get rid off the distorting parasitic electrical transients.

To find out what settings are required for a particular measurement setup, it is advised to carry out some trial measurements of a few seconds in order to have a look at the raw, measured transient responses. In this way it can be decided whether the distortion is of A. or B. type and the range where the minimum or inflection point is to be searched, can also be identified. Figure 5-8 shows raw data of a trial measurement where parasitic electrical transient of type A can be observed. The minimum is between 0.1 sec and 1 msec. Figure 5-9 shows a B type distortion (a) and the results of the correction (b). In this case the inflection point was somewhere between 3 m sec an 5m sec.

The controls corresponding to the correction algorithm are as follows:

No button

Press, if you want to carry out "raw", trial measurements to see, what type of parasitic transients – if any – take place.

Min button

Press, if the distortion due to parasitic electrical transients is of type A (search for a minimum point).

Infl button

Press, if the distortion due to parasitic electrical transients is of type B (search for an inflection point).

Min, Max input fields

Enter numbers that specify the lower and upper limits of the time interval where the minimum/inflection point is to be searched.

	a)
	b)

Figure 5-9: Raw and corrected measurement results in case of a distortion with an inflection point: a) raw data, b) corrected curve

Note, that initial electrical transient correction is also possible in the T3Ster-Master program as a post-processing option.