Accuracy of test instruments for NIC

[ban-all-sheds]

You have the "casting vote" in the form of the check box.

You start with 2 meters, calibrated, which one assumes start giving the same results from the check box. If one later diverges and one does not, then the divergent one must be rogue, barring a coincidence of a fault and a change in Zs less likely than Osborne renationalising the railways and utilities with no compensation.

 

[JohnW2]

You have the "casting vote" in the form of the check box. ... You start with 2 meters, calibrated, which one assumes start giving the same results from the check box. If one later diverges and one does not, then the divergent one must be rogue, barring a coincidence of a fault and a change in Zs less likely than ...

Since the check box knows nothing about absolute values, it is irrelevant to knowing whether one of the meters is a 'rogue' in terms of absolute values. Are you simply saying that if you test on the same installation/circuit as in the past (and assume that Zs has not changed), then if one meter shows the same Zs as before and the other doesn't, that the latter must be the 'rogue'? If so, that does not require a chek box, but does require that all checking is done on the same circuit whose Zs is assumed never to change.

If one cannot always check using the same circuit and/or is not happy to assume that the Zs never changes, then that approach cannot be used. A check box will then tell you whether the slope of Δindicated/Δtrue remains correct for the meter under test (which doesn't require a second meter). However, if you have two meters, both of which still show the correct Δindicated/Δtrue with the check box, but which display absolute values which are somewhat** different, without a third meter you cannot 'tell' (at least, get an indication of) which is the 'rogue'.

(** if they were 'vastly different', it would usually be fairly obvious which was probably the 'rogue'!).

Kind Regards, John