Do not measure grounded circuits unless you know what you’re doing. That usually means that any AC signal from the wall is a no-no. Our DC supplies and cellphone chargers should be okay. The scope is earth-grounded, which means that if you clip the ground lead of a probe to a portion of AC that is not grounded, then you will cause significant and irreversible damage to the scope or probe or both.
There are cases where you can find earth-grounded HV DC or low voltage grounded AC, or something else that could be damaging, but those are edge cases. Probably not something a beginner will encounter. So the guideline could further be simplified by saying:
If you can reasonably expect 120VAC to be near where you’re probing and can’t list “best practices” off the top of your head, stop. Ask for help.
Oscilloscopes are deceptively simple-looking devices, and easy to break. There’s a reason students spend years learning all the ways they can be used
A mandatory training program for them would be awesome, if I ever thought we could get it off the ground and keep it going. I don’t, though, it’s not one of the fun tools.
I’m currently trying to fix a welder - this is a perfect example of what to be super-careful with and what should never be tested unless you know precisely what you’re doing (or are ballsy enough and own your own equipment… so no hackspace equipment use for me on this). Please note that being ballsy is not a replacement for knowing the safety required with something that can kill you, though.
The welder has a 3-phase bridge rectifier and there is no connection to earth-ground on the main PCB except through it’s shield or screw posts - this board set runs at a floating voltage somewhere around +/-60V dependant on load and what the input voltage is (+/-60V for single-phase 115V). I can not reference my scope (or connect the ground probe) to anywhere on the board except earth-ground as I know mine also isn’t isolated. This has made troubleshooting difficult as all the signals are riding on this +/-60V signal.
The workarounds to this are also very… error-prone… the isolation transformer I have, for instance, only has the power phases isolated and the case is still tied to case-ground on both input and output sockets - something to be very aware of. They also do not make any of this safe - on the contrary, the workaround makes the entire oscilloscope unsafe and knowledge of what voltage is present where is critical.
On the topic of computers, power supplies, wall warts, etc; Almost all of them are isolated internally. It’s only things such as VFDs, Welders and some specialized equipment that isn’t. If you’re working on something with AC in and a low-voltage line on it, it’s probably isolated. If you’re unsure, be careful, use a multimeter to check - there should be no AC voltage on the case or if there is any it should be very little (devices with no ground prong may have a little due to coupling).
is there any legitimate reason to ever test the mains?
YES - I developed a whole range of power line communication devices at one stage and so yes I regularly had a scope and spectrum analyzer directly on the 120V line
BUT … I always had the instruments powered by a fully isolated 120/120 transformer
Maybe a transformer is a good idea - but some users already know how and what is possible/safe and appropriate
My apologies, I probably phrased that wrong. I meant it in the context of a hackspace, and I was also under the impression that main was primarily intended as live mains, not so much “emulated” mains.