Cheaper overvoltage protection

[snx]

I'm currently using a IC solution (TPS2400) to protect a circuit from more than 18V of Supply power. A single Diode protects the circuit from being powered in reverse polarity.
My goal is to improvise the circuit to switch off at around 14-15V's so, that its possible to use (cheaper) 16V Tantalum Capacitors.

It is desired that in the new version the reverse polarity attempt will not cause a fuse blow, so this will be solved with a N-MOS in the ground path in combination with a zener diode.

I also would like to ged rit of the TPS2400 since its priced around 2$ which is a lot, compared to other components.  On the web, i saw the "crowbar" overvoltage protection, but again, that would result in a fuse blow, that is very undesired.

Is there a economic possibility to remove the TSP2400 and use standart components to get the desired overvoltage protection (Vmax= 14...15V)?
Any help and ideas is appreciated!

I've attached the current schematic for reference.

Thanks

[Kleinstein]

If it is just for the tantalum capacitor, I would consider using MLCCs instead. They need to be specified a little high anyway to avoid the capacitance dropping too much.  A possible solution could be using a PTC fuse and shunt type protection against overvoltage and reverse. The resistance of the PTC has some downsides but also some advantages - so it depends.

[Huluvu]

Tantalum Caps should be designed in with no more than 60% of the Voltage than rated for .
Additionally you should avoid high peak currents.
I would try to avoid them.....

[David Hess]

It is desired that in the new version the reverse polarity attempt will not cause a fuse blow, so this will be solved with a N-MOS in the ground path in combination with a zener diode.

Be careful about low side switching if ground can be provided by some other input or output.

Low side series protection also means that any exposed circuit become elevated to the input voltage which may be unacceptable.

[AngraMelo]

How about using a comparator switching a relay off in case of over voltage?
would the relay cost be too much?

[TimNJ]

It is desired that in the new version the reverse polarity attempt will not cause a fuse blow, so this will be solved with a N-MOS in the ground path in combination with a zener diode.

Be careful about low side switching if ground can be provided by some other input or output.

Low side series protection also means that any exposed circuit become elevated to the input voltage which may be unacceptable.

I've been looking at reverse polarity/input protection circuits recently...and I've seen this said a lot. Can you give a tangible example of what the problem would be? And by "ground", do you mean "earth"?To me this only really makes sense in a system with an earth connection elsewhere?

[TimNJ]

How about using a comparator switching a relay off in case of over voltage?
would the relay cost be too much?

I would say relay response time is much too slow for catching transient events.

But speaking of comparators, you can probably do it with a TL431 and a low-side N-channel MOSFET switch. TL431 has an internal 2.5V reference. Just set up a voltage divider that will produce 2.5V when the rail reaches the overvoltage threshold. Will probably be a few mA of quiescent current. Not sure what your power requirements are.

[David Hess]

It is desired that in the new version the reverse polarity attempt will not cause a fuse blow, so this will be solved with a N-MOS in the ground path in combination with a zener diode.

Be careful about low side switching if ground can be provided by some other input or output.

Low side series protection also means that any exposed circuit become elevated to the input voltage which may be unacceptable.

I've been looking at reverse polarity/input protection circuits recently...and I've seen this said a lot. Can you give a tangible example of what the problem would be? And by "ground", do you mean "earth"?To me this only really makes sense in a system with an earth connection elsewhere?

Ground in this case just means common or the return current return path.  Really it should be called common but almost everybody uses the term ground instead.

If multiple devices share common and the same power supply, then they can happily supply common to each other even if one disconnects it like through an interface cable.  And if that is not a problem, then having one device disconnect its common return internally risks raising all of its output signals to the positive supply unless precautions are taken in the design to prevent this.

I have had this happen a couple times with RS-232 and more recently with USB with the result that a "ground" trace got violently removed.