• Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Mains overvoltage transient protector using linearly controlled FETs.......or IGBTs?

Not open for further replies.


Well-Known Member
We have an offline, 30W, non-isolated, LED driver circuit that we wish to protect against mains transients. The LED driver IC is in linear mode and connects directly to the post rectifier DC Bus. Unfortunately it has a max Vdd voltage of 450V.

We therefore are using the regular, time old transient protection method of having an NFET connected across the DC Bus, and controlled as a shunt regulator, which linearly regulates the DC Bus to 413V in the event of an overvoltage transient. When activated, this shunt NFET shunts current away to another NFET which is series connected in the “ground” return of the DC Bus…..and this ground connected (900V rated) NFET is connected to act as a current clamp (by use of the time old ‘BJT method’ of having a BJT pull the current clamp NFET’s gate down depending on the amount of current flowing through a base-emitter connected resistor.).
Anyway, the current clamp clamps the current to 700mA.

Incidentally, there is also a SMCJ400A TVS in parallel to the shunt NFET.

Anyway, a summary of the way this protection circuit works, I am sure you know already , is that basically the 'Shunt NFET' effectively “transfers” the transient overvoltage to the 'Current clamp NFET', so that the load (LED driver) can survive and avoid the overvoltage.

*** The thing is, should we use IGBTs instead of NFETs? ***

The datasheets dont allure to the SOA, so its hard to say

We just wonder if the passing of 700mA through these DPAK NFETs with several 100 Volts across them (even though only for some 200us maximum), is going to damage them? –We have certainly had failed returns. Unfortunately, the below NFET datasheets don’t give any details of Safe-Operating-Area.
The IGBT below is made of sterner stuff than the FETs. Can you think of any reason that we shouldn’t use it for the Shunt? , and also use a 900V IGBT for the "current clamp" bit too?

AOD5B65N1 IGBT datasheet: (prospective "shunt IGBT")

FQD1N60 NFET datasheet: (‘shunt NFET’)

STD3NK90Z NFET datasheet (‘current clamp NFET’)


Active Member
There are SOA curves in both datasheets. Looking at them for the DPAK package, you look to be well inside the SOA curve if the transient is only 200us. At 700mA, it looks like the STD part can survive for at least 1ms with 450V across it. For the FQD part, just under 1ms with 450V/700mA... this part is a bit closer to the wire in terms of SOA. You should not need to resort to an IGBT
Not open for further replies.

EE World Online Articles