Some caveats about the circuit attached below:

1. The input ranges from 100 to 600VAC. (Mains are isolated)
2. The coil L1 has a DC resistance of 300 Ohms.
3. It's oversimplified to show the concept. SW1 is a MOSFET. When SW1 is open, the load should not draw more than 10 mA at worst. Normally about 5-6 mA. When SW1 loses, the load gets bypassed, and a higher current will flow through the coil L1 until the supply is cut. The coil should trip the power supply as soon as possible to avoid FET overheating. Once the supply is cut off, the only way to turn it on is manually. I mean, the coil has to be reset manually.
4. The FET is configured as a current limit. This mean it's always On, not switching. So I have to make sure I don't go outside the minimum and maximum current limits.

How do I do the right power calculations to find a suitable transistor? This is what I think:

- I would need a high VDS breakdown. At least 1200V for the case when the input is 600VAC. It gives about 848VDC. I can add TVS diodes across VDS to protect the FET.
- When SW1 is open, VDS is high. Supposing the load draws 5 mA, P = 5 mA x 800V = 4W
- When SW1 closes, VDS gets a bit smaller, about 400V. This yields P = 1.2A x 400V = 480W! But this is a very short pulse. The trip coil cuts the supply in less than 50 ms.

What would you suggest?

 

4. The FET is configured as a current limit. This mean it's always On, not switching. So I have to make sure I don't go outside the minimum and maximum current limits.

Before starting I need to say "if you need to ask, you probably shouldn't be doing it!"

And one big reason I sat that is your statement #4. Your mosfet is configured as a high side device. To turn it on the gate needs to be at ground or below since it is a Pmos. So how is it going to get there the way you show it?

 

Before starting I need to say "if you need to ask, you probably shouldn't be doing it!"

And one big reason I sat that is your statement #4. Your mosfet is configured as a high side device. To turn it on the gate needs to be at ground or below since it is a Pmos. So how is it going to get there the way you show it?

It's a NMOS device as shown in the schematic.In fact, that's how it's been built physically. It works for 480 VAC. I'm just trying to calculate if I need to change the FET for a 600V system.

 

It's a NMOS device as shown in the schematic.In fact, that's how it's been built physically. It works for 480 VAC. I'm just trying to calculate if I need to change the FET for a 600V system.

Looked like a Pmos in the drawing. Still it is a high side switch so how is it going to conduct when the gate voltage needs to be around 10volts higher than the drain? So that still means your contradicting your statement #4.