Capacitance and Impedance Bridges

The effectiveness of these instruments for dielectric cure studies depends on sensitivity and accuracy. The sensitivity is related to the minimum resolvable phase angle, which for general cure studies, should ideally be less than about 0.10. Unfortunately, actual sensitivity in use depends strongly on the measurement frequency, on the admittance of the sample, on the details of the cabling and shielding, and on the electrical noise level of the environment. Therefore, analysis of published sensitivity specifications is difficult. It is easier to evaluate intrinsic instrument accuracy, which can be expressed in terms of either the tan8x accuracy or the conductivity accuracy. An example is useful. 

A pair of 1 cm2 area plates spaced apart by 0.25 mm and filled with a resin having a permittivity of 10 (a typical value early in cure) has a capacitance of about 35 pF. The HP4192A has a tan 5X accuracy of 0.002 when measuring 35 pF at 1000 Hz 18), which is satisfactory for most resin studies at that frequency. However, the tan 5X accuracy of the HP4192A degrades to about 0.05 at 5 Hz, which limits the smallest conductivity that can be measured. In the final stages of typical cures, e approaches a value of 4-5, while e approaches a value that depends on frequency. At low frequencies, the e value is usually dominated by ionic conductivity, denoted by ct (see Sect. 3.1.1). In this case, the resistance Rx is L/ctA, which when combined with Eq. (2-16) yields 

Additional problems are introduced by the overall admittance level and by cabling and shielding issues. In Fig. 9, the parallel-plate sample is connected to a meter 

---