Sparks Measuring Vds on Oscilloscope

I am using the LNK 302 for a buck converter intending to achieve a 12 V output. I have designed the PCB and built the board, but when I measured the voltage at the output it was only 4.86 V average instead of 12V. Because of this low output voltage I measured the input to the drain and found that the voltage was about 176V with about 12V of ripple with respect to neutral. Eventually it was suggested that I measure the voltage from source to drain. Because reference clip of the probe was large I ended up measuring the voltage from drain to source instead. The clip was attached to pin 5, the drain and the probe was placed at either pin 7 or pin 8 which are both source pins. For a little bit, maybe a second or two, the oscilloscope was able to show a switching waveform, but then there was a spark, a lot of smoke, and I dropped the probe. I turned off the power  by switching off that variac and tried to remove the reference clip. It seemed to be stuck, but I was able to remove it. When I removed it, there was an additional spark. It appears that the clip had been welded to the drain of the chip. At the end a hole was burned into the packaging of the LNK 302 chip and a capacitor was blown. Pin 5 was severed most likely from me removing the clip and having pin 5 welded. The design did not use the suggested input stages or the suggested PCB layout the design has been proven too work with the selected input stage and a different PCB layout.

I was wondering if you might have any idea why sparks might have been made when simply measuring the voltage from drain to source with an oscilloscope probe. I don't really think it was the chip, but I want to make sure that I've covered all of the possibilities since I still have not discovered the reason. Thanks.