Isn't the answer to this thread - a proving unit?
Or....
a shrubbery!!!!
I have a proving unit. If you mean use it to prove the voltage indicator, no it's not the answer as far as I can see (but see below). Please feel free to correct me on this, maybe I've misunderstood you. But the issue is not in proving that the voltage indicator is working.
I get the abiding sense people are not getting what I'm saying. Let's try again :-
1. If you have a board where the supply tails are
unenergised, and you switch off lock off and 'prove' dead on the load side,
you haven't proved you've isolated the board from the supply. A knackered switch with a load-to-supply short on the live side would give you the same 'dead' result. So would leaving the switch closed for that matter.
2.
An isolation test which cannot tell the difference between isolation and non-isolation is worthless. This is just going through the motions, as if the mere ritual of safe isolation will save you. (Ooh, came over all lyrical there. Never mind. Welsh blood. Sorry.)
3. If the means of energising the tails is not under your control (for example Economy 7 teleswitch) and you have not really proved isolation,
you are working at unreasonable risk. You might die or, even worse, breach the Elec @ Work Regs 1989.
Best solution to date in my personal (inexperienced, incompetent, clearly unsafe to practice and probably with questionable personal hygiene to boot) opinion is safe isolate and 'prove' dead, then check if the supply is itself dead. If it is but could become live (e.g. Economy 7), then IR test at 250V-then-500V across both sides of the open switch* to prove isolation. Neither side should be live (tails are currently unenergised) so IR test instrument should not complain. (Testing on continuity range only shows that at 24Vdc the open contacts provide >2000 ohms resistance, on my MFT anyway - my lust for life much prefers to know it is >1000 Mohms at 500Vdc.) I AM MAKING THIS UP FROM FIRST PRINCIPLES. IT IS NOT LAID-DOWN BEST PRACTICE. DO NOT TAKE THIS AS GOOD PRACTICE JUST BECAUSE YOU READ IT HERE.
The only way I can think that a proving unit helps is (NOTE TO ANYONE READING - HERE FOLLOWS REALLY STUPID IDEA - DO NOT DO) if you rig it up to apply the proving voltage across the supply side of the switch and then check the load side is dead. On any proving units I've ever seen, the arrangement would have to be seriously Heath Robinson - I wouldn't go there. Might be a good way to blow up your proving unit if the supply gets energised while doing it.
Any of you experienced guys out there care to shout out if you have ever come across a main switch which does not actually isolate on one or both poles, just to show me I'm not in cloud cuckoo land here?
(*Worth bearing in mind that, assuming the supply neutral tail is connected to earth potential, the IR test on the neutral side of the switch will have the effect of applying the test voltage E-N within the board, and so potentially E-L as well through any loads not disconnected.)