The measurement of samples or test specimens with particularly high resistance requires a different procedure compared to the measurement in the usual resistance range. Resistances greater than 10 MΩ are referred to as high resistances. The greater the resistance, the more the points discussed below come into play.

The suggestion

From an instrumental point of view, it is usual to use a current source and measure the voltage applied. With high resistances, however, only a small current flows and the capacitance of the lines becomes noticeable. It is therefore better to use a voltage source and measure the current. This point leads us directly to the next modification; instead of the usual 4-point contacting, 2-point contacting is preferable. Since the resistances of the cables are very low compared to the sample, the influence can be neglected. A DC voltage is preferable because of the time effects that can occur with an AC voltage. It should be noted that there is often a shift in the zero line. An additional measurement with a different sign is therefore necessary. The resulting resistance is the mean value. This is shown schematically in Fig. 1. 

Cable & earthing

The cables deserve special attention. They exhibit parasitic capacitance and/or leakage currents. Both effects are undesirable and should be minimized. Triax cables are a sensible choice as they reduce these effects by providing a medium protective shield at the same potential. The cable lengths should be limited to what is actually required. If various components are involved in the measurement setup, e.g. a cryostat in addition to the measuring device, you should think about earthing. Many independent earths lead to ground loops, which increase the noise. Ideally, there should be a point at which different masses converge ("star point"). 

Calculation example 

Our test object has a resistance of 10 GΩ. We use the M81-SSM with the new SMU module, which is also shown, as the measuring equipment. The SMU module can be used in the voltage source and ammeter configuration, among other things. We start with 10 V DC voltage. Using I = U / R results in a current of 1E-9 A or 1 nA, which can still be measured very well; the accuracy of the ammeter function is specified as ±(0.5% + 300 fA) or in this example this is approximately ±6 pA. 

More about electrical measurements