Automatic Big Data Acquisition of Electrical Parameters in Wide Bandgap Devices

This paper presents an automatic big data acquisition experiment platform dedicated for online monitoring of
$V_{th}$
and
$R_{ds, on}$
of GaN HEMTs. A new circuit topology combining power cycling test and double pulse test (DPT) together is proposed for measuring
$R_{ds, on}$
and
$V_{th}$
. Gate driver signals are used to trigger two oscilloscopes to continuously and automatically store data. A quasi
$V_{th}$
online monitoring circuit has been implemented using fast comparator to monitor turn on transient of GaN devices. A hot plate has been used to preheat device under test (DUT), so the initial temperature in each cycle can be controlled. At the beginning of each cycle, DSP gives a double pulse signal to DUT and the platform is working in double test mode. So
$V_{tk}$
can be acquired at the same junction temperature. After that, the platform changes to power cycling test mode through controlling super junction MOSFET to turn on/off and also sending turn on/off signals to power supply. In order to make sure
$R_{ds, on}$
is recorded at the same junction temperature in each cycle,
$R_{d\mathrm{s},on}$
should be calculated at the beginning moment of the first cycle power cycling test. After 10 cycles of power cycling test, there should be a long cooling period before starting the next DPT. In this way,
$V_{th}$
can be acquired at the same junction temperature. The whole experiment process is controlled by DSP, so it can operate automatically and acquire big data efficiently.