Current sensor and relay as resetable fuse, feaseable?

[PeterFW]

Hello!
So... lets say i take INA226 current shunt monitor, it has a (latching) open drain alert pin.
This pin goes low between 1ms and 5ms after a fault has been detected, page 29 in the datasheet, can be longer depending on the configuration.

Then i take a relay and fet with a pullup resistor at the gate and turn a relay on with it.

Now i connect to open drain alert to the gate and have it disable the relay in case of a fault, the next best datasheet if found quotes a release time of 5ms.

Lets say, for good measure, after 20ms the output disconnects after the powersupply is shorted to ground or too much current is drawn.

I know it depends on the power supply, for the sake of the argument assume the power supply in question is a standard "cheap" ATX powersupply with a 20A output on the 12V or 5V line.

Before someone asks, yes, i have everything at hand could just wire it up and test it to see what happens but i wanted to ask here before i proceed.

Alternatively i could use a beefy FET instead of the relay, that would cut the response time a bit shorter.

Greetings,
Peter

[Kleinstein]

With DC current I would prefer using a MOSFET, as relays can have trouble breaking a high DC current due to arcing. There are relays for cars than work reasonably well.

[PeterFW]

Hello and thanks for the reply!

With DC current I would prefer using a MOSFET, as relays can have trouble breaking a high DC current due to arcing. There are relays for cars than work reasonably well.

I have a H-Bridge motor driver on the output as well as a relay, i could put some discrete logic in and have the sensor control the H-Bridge instead of the relay.
But that still leaves the question open if this is a decent approach to hold back the magic blue smoke for a bit :-)

[Neilm]

You haven't said what you are trying to do with this. A relay can weld, and a FET can fail short circuit. So if there is any safety function for this, then I would not recommend it. The only safety standards I know do not allow resettable fuses to protect against fire. If all you are trying to do is stop something releasing the magic smoke then it would be OK. Personally, I would prefer the FET to relay - I have seen relays weld after a couple of operations when switching high DC currents.

[PeterFW]

If all you are trying to do is stop something releasing the magic smoke then it would be OK.

Ah, sorry, yes!
I just want to keep the blue smoke from escaping if the supply accidentally gets shorted a few times during some tests i want to run.
A proper power supply in the current range i need is a bit pricey and i do not want to blindly buy one without running some tests first.

[timb]

An SCR crowbar + fuse would be preferable to a mechanical relay solution. Basically wire up the INA output to fire the SCR, which shorts the output of the supply. That would fully load it down causing the fuse to blow. You could also use a resettable polyfuse; these tend to take a bit longer than a glass fuse to trigger, so make sure the SCR can take the power drain.

Ideally you need something that can react in <10mS or the uS range.

Personally, this is what I'd do: LT has a line of "Surge Stopper" chips that will actually do exactly what you want. They monitor a small current shunt and control a pass FET to prevent the current or voltage from going over a set limit. They contain all the gate drive and monitoring circuitry needed to do the job. Very nice little chips.