Resistivity nikel coatings on copper

Are the contacts rated in the voltage / current range you want ? They might require some flashing to get electrons flowing. The meter might be using to low of a voltage to break the skin effect that sits on the contact. Use different scales and compare. Higher resistance scales use higher voltages. Might have to get a 20k/ohm meter instead of a high impedance circuit. e.g. analog meter not digital. Martin
Kar wrote:

Great !
I will try this asap, I want to go inside this problem. We are measuring using a digital tester, we let a fixed ac current flows inside the interrupter (some amps, the interrupter nominally works at 2500A!) then we measure the voltage fall between terminals. Yesterday I spoke with my colleague and she said that this strange behaviour has been found also in other apparatus but they never understood why.
What do you exactly intend with "might require some flashing", could you explain better ?
thanks a lot

is this in vacuum? air? other gas?
what is your application? why are microohms important?
each time you make and break an electrical contact in air, you'll get some degree of arcing and oxidation. therefore, no two contact resistances will be identical. if you need consistency, but operate at a reasonable voltage range and can tolerate [considerably more than microohm] resistance, you should look at semiconductor switching. if it has to be mechanical, you can look at reed switches since they have their own atmosphere and therefore much more consistent make/break outcomes.

Some high currents need plasma arc for a continuous current flow - as the contacts are not as one but are touching.
Some switches are mercury baths and the bath is vaporized allowing conduction and once it is, it remains until an anti-polarity voltage is applied.
e.g. consider the contacts might be porous and might touch point to point and not flat to flat as expected or when in full conduction.
Good luck in the search of truth. At extremes, there are unique parameters often ignored in mathematics.
I remember an RF spot welder that had wire that was soft copper stand and twice the thickness of my thumb. The unit was a pulse welder. (University R&D lab)
Someone was servicing the point set and had untied the hold-downs. That allowed the wires to flail about in wild motions when pulsed. We got the hold-downs affixed in due time!
High current might effect meters (change positions and take a comparison measurement. Rotate in all quadrants in 3 axis. Hall devices and meter movements are all modified.
Martin
Kar wrote:

Dear all
I thank you very much for your ideas ! I'll go deeper into this application. Interrupters work in low voltage -300V- and high current, 2500A, contacts are "in air" (no vacuum or gas).
Problem arised because we have to do some customer acceptance tests and they asked us to demonstrate that overall resistance of a single interrupter is be inside a limit defined by their specifications (max 15% higher than the value measured during normal thermal test). This means that if we measured 50 microohm during thermal tests we can jump till max 57,5 microohm.
In the past we did similar tests (we had silvered contacts) but never experienced problems like this in our devices. Our problem is of course the high resistance value but more, the way to make them accept the unstable value 50-400 micorohms changing just after every open - close !
I'll keep you informed !
Greetings