4.2.2  Comparison: Simulation / Measurement Data

Figure 4.13 is the simulation data for the circuit in Figure 4.12, Figure 4.14 is the corresponding measurement data. The diagrams show the charging of the smoothing capacitors at a load of R1 = 10MW (V_power_1 = 60V). In the simulation, 4ms after turn-on, a voltage-level of 1120V is reached, in real 1056V. In the simulation, the 1120V is the highest voltage, in real, the voltage increases 20ms after turn-on to 1112V. These differences may be caused by a voltage drop at the dc-supply at fast load variations (to be verified).

Figure 4.13: Simulated voltage: output-signal @ R1

Figure 4.14: Measured voltage: output-signal @ R1

Figure 4.15 and Figure 4.16 show the voltages at the output of the IGBTs, (R1 = 4.9kW, V_power_1 = 60V, circuit: Figure 4.12. (a): rise at IGBT1+4, (b): rise at IGBT2+3.

Figure 4.15: Simulated voltage: IGBT-output

Figure 4.16: Measured voltage according to Figure 4.15

Figure 4.17 and Figure 4.18 show the simulated and measured gate signals at IGBT2, Figure 4.19 and Figure 4.20 show a magnification of the noise spikes (R1 = 4,9kW, V_power_1 = 60V).

Figure 4.17:Simulation data: gate signal at IGBT2

Figure 4.18: Measurement data according to Figure 4.17

The noise spikes (see arrows in Figure 4.17/Figure 4.18) occur at the switching of the other channel. The oscillation on the measured diagram is caused by parasitic inductances in the tracks and pins. The voltage level of the spike is below 4V, a fault turn-on does not occur. But it is likely that at a higher dc-voltage a higher noise spike is induced. At a level of over 5V such a spike turns on the IGBT and causes a short circuiting of the dc-supply. Such a short circuit destroys the IGBTs.

Figure 4.19: simulated data: noise spike @ Gate IGBT2

Figure 4.20: Measurement data according to Figure 4.19

This page is part of a Frameset: Electrodynamic Sculpture: A Thesis by Rafael Bräg.