Logo PTB

Key comparison CCEM-K4.2017 of 10 pF and 100 pF capacitance standards


A key comparison of capacitance and capacitance ratio, piloted by the BIPM, was successfully completed. Comparisons are a key instrument of metrology to test the principal techniques and methods in the particular field and to demonstrate and maintain equivalence of the realised physical units at the lowest possible level of uncertainty, as stipulated in the CIPM MRA.



In the frame of this key comparison, seven participants carried out simultaneous bilateral inter-comparisons with the BIPM. Each participant measured a set of its own capacitance standards for six weeks each before and after the standards were transported for an 8-week measuring period to the BIPM. The measurand was the capacitance of thermostated fused-silica standards at nominal values of 10 pF and optionally 100 pF to inter-compare the 10:1 ratios. Compared to a ring circulation, this scheme shortens the comparison time and, thus, minimises the influence of long-term instabilities of the standards. The comparison was restricted to participants being able to realise and to maintain a representation of the farad at the best level of uncertainty. Three participants and the BIPM traced the farad to the quantum Hall resistance measured either at ac or dc, and four others to a calculable cross-capacitor. PTB is the first, and still the only, national metrology institute which realises the capacitance unit from the quantum Hall resistance directly measured at ac. This approach requires specially shielded quantum Hall devices but has unbeatable advantages and uncertainties. Therefore, a couple of other national metrology institutes started activities in this field of work.

For the mandatory 10 pF measurements, the results of the comparison show a good agreement within a relative uncertainty as low as 2·108 (k = 1). In fact, some participants have an even lower uncertainty for a single capacitance measurement (PTB with 6·109 has the smallest uncertainty), but the short- and long-term instabilities of the standards and an imperfect travelling behaviour often limit the total uncertainty of a comparison. Furthermore, a good agreement has also been found for the optional 100 pF measurements and the 10:1 ratio.

Beside the actual comparison, the results allowed an evaluation of the von-Klitzing constant which was found to agree with the last CODATA adjustment within a relative uncertainty of about 4·108 (k = 2). The final report presenting the details of the measurements and the analysis of the data has been accepted and will soon become available at the Key Comparison Data Base of the BIPM (KCDB).

The limiting properties of the commercial capacitance standards seem to be due to an imperfect temperature stability. Therefore, we started to develop improved thermostats to allow lower uncertainties in shorter measurement periods - for the benefit of future international comparisons as well as for customer calibrations.