I tried it on a few Agilent bench and system supplies at my hobby lab:
E3615A, there is no output off, so I have to set the output to zero and turn the unit on:
Unit 1: 0 spike at power on, I had to turn up the voltage just to make sure everything is connected properly.
Unit 2: Same as unit 1
Unit 3: It got a 20mV spike that smoothly ramp down to 0 after 300ms
6632A System Supply, it default to output on, but also default at 0V:
0 spike at power on.
6643A System Supply, it defaults to output off at power on:
Unit 1: 0 spike at power on.
Unit 2: 0 spike at power on.
E3631A bench supply, it defaults to output off at power on at all 3 outputs:
6V output: There is a -280mV voltage that lasted for 550ms
+25V output: 0 spike
-25V output: There is a 330mV voltage that lasted for 580ms
I expected the system supplies to be at a much higher standard than the bench supply, so their results is what I expected. The E3615A are my go to supplies for bench work, those are rock solid and the results bear that belief. I did learned something for the E3631A, I definitely will not leave it attached to sensitive DUTs while powering it up.
Back to the Rigol, I have to think the turn-on spike is excessive, you have to be careful not to leave delicate DUTs attached to it during power up.