= Diode characteristic measurement
The New Jersey Institute of Technology laboratory uses a <silicon> 1N4148 <diode>. For its forward <current-voltage characteristic>, the supply operates in current-source mode; a benchtop <voltmeter> independently reads the <diode> <electric potential difference>. Small <electric current> steps resolve the steep branch. For reverse measurements, a sensitive <ammeter> sits in series; the handout suggests a $10$–$100\,\mathrm{k\Omega}$ series <resistor>. Reversing the applied polarity supplies the comparison. Meter placement must avoid counting the <voltmeter>'s own <electric current>, and the zero reading may have an offset. The same <diode> is then measured with an Agilent U2722A source measure unit for graphical comparison.<Part B Semiconductor Devices: Diodes 1 by New Jersey Institute of Technology>[Part B, Experiments with Semiconductor Diodes, items 1 and 2(a): preparation, current control, independent voltage readout, reverse series resistor and same-device comparison]{ref}
The prescribed analysis uses a <logarithmic scale> for forward <electric current> to infer $I_S$ and $n$ from the straight region. Large-<electric current> departures can indicate series <electrical resistance>. The handout cautions that the accessible range may not determine it precisely; meter loading also matters. This is a measurement procedure, without a published set of device readings.<Part B Semiconductor Devices: Diodes 1 by New Jersey Institute of Technology>[Experiments with Semiconductor Diodes, item 1, parameter extraction and high-current trend-line limitations; Report and Analysis]{ref}
The reason for the <logarithmic scale> follows directly from the <Shockley diode equation>. If $I_D\gg I_S$ and series <electrical resistance> is negligible,
$$
\ln\left(\frac{I_D}{I_0}\right)\approx\ln\left(\frac{I_S}{I_0}\right)+\frac{V_D}{nV_T},
$$
where $I_0>0$ is a fixed reference <electric current> making the <logarithm> dimensionless. The slope is $1/(nV_T)$, so extracting $n$ needs the junction <temperature>. Fitting only an appropriate <interval> matters: a straight-looking subset cannot establish the model at all <electric currents>.
Back to article page
