IGBTs cannot switch OFF quickly due to dreaded "Tail Current"?

Hello,

We have designed the attached Mains Transient imitator circuit, which exposes our LED Power supply to a synthesized “Mains Transient” voltage spike so as to see if it survives. Schem and LTspice simulation attached.

We must use IGBTs (not MOSFETs) for the switchs, since they have far better overcurrent rating, and we may see overcurrents in the event that the overvoltage spike ‘kills’ the LED power supply.

We need to be able to switch the IGBTs ON/OFF very quickly (within 1us). However, years ago, I learned that IGBTs have a large “tail current”, which means that they simply cannot be switched off quickly. We are using the STGW30NC120 IGBT (for all 4 IGBTs) , and its datasheet does not refer to this “tail current”…do you know why?...and do you believe there will be a significant “tail current”?

The modus operandi of the circuit is that Switch 1 comes on first and gets the LED power supply up and running. ..Then, Switch 1 is suddenly turned OFF….and then 5us later Switch 2 is switched ON for just 100us and then suddenly switched OFF again…and then 5us later Switch 1 turns back ON again………………………………………….this is repeated continuously.

-During the short period that Switch 2 is ON, the overvoltage spike is applied to the LED power supply under test. The Overvoltage is caused by the sudden voltage step input to the LC circuit.

So what do you think about the “tail current “ problem?

STGW30NC120 IGBT Datasheet:

https://www.st.com/content/ccc/reso...df/jcr:content/translations/en.CD00077681.pdf

Flyback said:

So what do you think about the “tail current “ problem?

Click to expand...

It looks like the turn off is slow; from Gate=0V to Collector current = 10% is 600nS.

First of all, this looks nothing like the combination waveform generator and the coupling and decoupling network that is in the standart. Second, why do you need to turn the switch off quickly? The normal circuit uses most likely an SCR that stays on until the current dies out - remeber these requirements are pretty old so there would be no IGBTs or high voltage FETs then so they had to do it somehow with not too exotic parts. See the attachemet for a CWG circuit with correct values that match the open circuit voltage and short circuit current requiements of IEC 61000-4-5.

Your chosen IGBT may be particularly slow.
A sim using a IXDH20N120 model shows a turn-off tail of only 10nS.