Experimental verification of the original design of the laminated busbar and the current design
In order to make an accurate horizontal comparison, we will absorb the power at both ends of the IGBT in the original design
After taking off, get the following:
1) Original design scheme The turn-off current and voltage waveforms at both ends of the IGBT are
Since the parasitic inductance of the DC-side capacitor is about 100nH, the inductance at the bolt pin is about 120nH,
The parasitic inductance of IGBT is about 20nH, and the total inductance of the busbar in the original design is about 760nH. Therefore,
The total line distribution inductance is about 1000nH.
In this case, according to formula (1), the induced voltage is:
ΔU=Ls*di/dt=1000x10-3x125/0.28≈446V
Where dt is the IGBT turn-off fall time (0.28μs), so that the total voltage across the IGBT is:
Ut=Ud+ΔU=800+446=1246V
This is basically consistent with the actual measured value, therefore, the original design must add an absorption capacitor, and the IGBT
The rated voltage must be 1700V.
2) The current design scheme The turn-off current and voltage at both ends of the IGBT
Consistent with the above method, first calculate the total line distribution inductance Ls:
Ls=0.3+100+120+20=240.3nH
The stray inductance of the bus is 0.3nH.
The induced voltage across the IGBT is:
ΔU=Ls*di/dt=240.3x10-3x125/0.28≈107V
Where dt is the IGBT turn-off fall time (0.28μs), so that the total voltage across the IGBT is:
Ut=Ud+ΔU=800+107=907V
This is basically consistent with the actual measured value, therefore, the current design does not need to add the absorption capacitor, the amount of IGBT
The constant voltage can choose 1200V level, which greatly reduces the probability of IGBT breakdown.