"Tough Guy no. 1265" wrote in message
news
I remember at work when our lab was rewired, every bench was fed from a
different phase and there were warnings not to connect equipment on one
bench (eg by RS-232, Ethernet, USB or other low-voltage data cable) to
equipment on another bench, which caused us horrendous problems.

That's a weird idea, you'd think they'd at least have each room on its own
phase.

Although I can't see the problem. USB for example would have its ground
the same as the appliances earth, which remains constant for all the
phases.

Yes I thought the restrictions seemed draconian considering that all signal
cables have a signal ground which may be connected to a common mains earth
ground through the equipment or else will be floating wrt mains earth and so
will be no worse for two pieces of equipment on the same phase compared with
equipment on two different phases. It meant that if we took any portable
equipment into the lab, we had to be careful to plug it into the same bench
as the equipment, even if that meant unplugging something else to free up a
mains socket. (*) They even supplied little opto-isolators if we ever
needed temporarily to connect a server on one bench by Ethernet to a router
on another bench when setting up a private LAN.

Maybe the worry is that if mains-live gets onto a signal wire separately on
both pieces of equipment (**) then there's phase-to-phase voltage between
the two whereas there's no voltage between equipment on the same phase, so
you wouldn't get a shock if you touched signals on both devices (providing
you weren't earthed) in the latter case.

I remember our department manager called site services when he became aware
of the operational problems it was causing us, and got them to cost out
rewiring the lab onto a single phase, with SS covering the cost, given that
the change was something that site services had taken it upon themselves to
introduce when we'd not asked for it. Sadly the whole department was made
redundant before the matter could be taken any further... So all the
building/rewiring work was a complete waste of everyone's time and money,
given that no other department with similar need for a lab moved onto that
floor - but such is life!

(*) What they didn't consider was that some of us had old VT220 terminals on
our desks connected by RS232 to servers in the lab, and no attempt was made
to ensure that the desk was on the same phase as the bench in the lab. It
didn't take some of long to see that unwitting loophole in the regulations.

(**) Unlikely for one piece of equipment, let alone two, to fail in this
way, but you've got to cover the worst-case scenario.