Error in bandgap measurements of n-type Ge

[joel.martens]

In conducting an experiment to measure the bandgap of n-type Ge using the 4-probe method, heating the sample and measuring the change in voltage across the probes as it cooled, it was necessary to change regularly between the milliammeter and the millivoltmeter setting on the instrument to ensure the current remained constant while voltages were recorded at temperature intervals. It was found that the act of switching to the ammeter and back caused large anomalies in the recorded voltages, specifically it caused the voltage readings to drop considerably lower than the ones taken previously. The voltages should steadily increase as the temperature drops gradually from around 170 dergrees to room temperature, levelling off to almost constant when extrinsic conduction begins to dominate.
Can anyone explain this fault, i was told that it may be that the millivoltmeters impedance doesn't match the milliammeters impedance but i am not familiar enough with impedance and how it relates to the operation of the millivoltmeter / milliameter to understand how this causes the error.

 

[Bob S]

Hi Joel-
Use a constant-current regulator for your power source, and two meters, one for current and one for voltage. Make sure the voltmeter is not measuring the voltage drop across the miiliammeter. Correct the milli-amp reading for the current drawn by the voltmeter. Lastly, understand the instruments you are using.
Bob S