We are using an Agilent 4284A LCR meter to measure capacitance and dissipation factor of a dielectric material, which is a thin film with two sides coated with gold by sputtering. The capacitance is about 180 pF (~1 Mohm impedance at 1 kHz, -89 deg phase angle). The dissipation factor is about 0.1% at 1 kHz to 1 MHz.

We need to measure it at 200 deg C, so we made a fixture by ourselves which can be put inside an shielded oven. The fixture is connected to the 4284A an agilent 16048A 4-terminal pair type BNC cable. It is connected to our sample holder using Figure 3-13. Shielded 2T extension as discussed in Keysight Low Impedance measurement handbook.

As the D of our 180 pF capacitor is low (0.1%). we decided to perform the short and open correction ("on"), and we also selected the "1 m" cable correction for the 16048A cable.

However, after these corrections, we measured a D of 0.03%, which is lower than our expected value of ~0.1% -- it seems to us that there is "over" correction. In the "Agilent Impedance Measurement Handbook", it mentioned that "negative D" may be caused by improper "short" correction, however we still cannot figure out what is wrong with our short correction.

Any comments and suggestions will be greatly appreciated.

We need to measure it at 200 deg C, so we made a fixture by ourselves which can be put inside an shielded oven. The fixture is connected to the 4284A an agilent 16048A 4-terminal pair type BNC cable. It is connected to our sample holder using Figure 3-13. Shielded 2T extension as discussed in Keysight Low Impedance measurement handbook.

As the D of our 180 pF capacitor is low (0.1%). we decided to perform the short and open correction ("on"), and we also selected the "1 m" cable correction for the 16048A cable.

However, after these corrections, we measured a D of 0.03%, which is lower than our expected value of ~0.1% -- it seems to us that there is "over" correction. In the "Agilent Impedance Measurement Handbook", it mentioned that "negative D" may be caused by improper "short" correction, however we still cannot figure out what is wrong with our short correction.

Any comments and suggestions will be greatly appreciated.