Yesterday, I fried the +15 V power supply rail on one of the Eurocrate extender cards while I was checking out the binary switching in the 1X5 rack. I will describe what I did it in the hopes that everyone else will be less stupid than me.
I wanted to monitor the voltage across a resistor on the suspension OSEM whitening board. Since I knew that both sides of the resistor would be at non-zero voltage (including possibly at the power-supply rail), I used a battery-operated scope with floating inputs, so that the scope would not try to pull the probe shield to ground. That should be OK, although not recommended, as you'll see, because you must be very careful to make sure that the scopes inputs are indeed floating.
Let's call the original signal 'A'. The trouble came when I then connected another signal (B), whose shield was connected to the ground on the whitening board, to the scope. Apparently the grounds on the scope inputs are connected, or were in the configuration I was using. When I connected the signal B, B's ground shorted A's shield to ground, which had been sitting at the +15V rail. That short circuit then fried the +15V supply line on the extender card I was using (escaping magic smoke was detected). Thankfully this only blew the extender card, and not the Eurocrate or the Eurocrate power supply or the whitening board or the scope etc, all of which would have been much worse.
The moral of the story is to be very careful when connecting power supply voltages to the shield or ground of a scope. In short, don't do it. I didn't ultimately need to, since I could have found other ways to measure the same signal.