Looking for some help on creating a circuit to be able to monitor 4 power supplies (could be any combination of +12V, -12V, +24V, -24V power supplies that all share a common ground) and provide a single relay contact closure in the event that any one of those power supplies turns off.. I know I can do it with 5 or so relays and some diodes but hoping to get a better (read..slightly lower cost transistor or similar) solution

I started going down this road and found out it only works for specific situations.. All neg polarity if supply ground leg is disconnected or all pos polarity if other leg is disconnected..
Just looking for other suggestions..

 

Looking for some help on creating a circuit to be able to monitor 4 power supplies (could be any combination of +12V, -12V, +24V, -24V power supplies that all share a common ground) and provide a single relay contact closure in the event that any one of those power supplies turns off.. I know I can do it with 5 or so relays and some diodes but hoping to get a better (read..slightly lower cost transistor or similar) solution

I started going down this road and found out it only works for specific situations.. All neg polarity if supply ground leg is disconnected or all pos polarity if other leg is disconnected..
Just looking for other suggestions..
View attachment 115096

Quite a lot of ATX PSUs have a common off the shelf quad comparator driving the fault shutdown - its easy to see if the chip is present, and not all that difficult to hand trace the circuit.

A couple of the thresholds might be just right - a few minutes with a calculator should sort the others out.

 

I can't access/modify the power supplies at all..

 

I can't access/modify the power supplies at all..

So how are you going to monitor the voltage?!

I would've thought someone who calls themselves; "mcgyver" should be able to figure out opening an old ATX PSU and copying the voltage monitor comparators......................

As for how you connect those comparators to a PSU can't access at all - well that's a tricky one.

 

Is there at least one supply voltage that's constant to use to power the circuit?

 

Is there at least one supply voltage that's constant to use to power the circuit?

The supply to a voltage comparator doesn't have to be constant - it just has to remain within the chip's Vcc requirements.

A typical Vref might be 1.25V fed by any rail that's always more than that (you can wired OR all the rails with diodes) and any rails that vary are dropped down to the same as Vref with divider resistors.

 

How about this...

A supply failure causes the relay to de-energize. The source supply for the mosfet circuit could be any one of the monitored supplies or a separate supply.

 

@eetech00:
There are two errors in the circuit, the polarity of input led of pc817 in the -12V and -24V.

Don't you think the current flows through the input led of pc817 could be too less?

Collector current:
If = 5 mA, Vce = 5V, Ic = 2.5 - 30 mA

Collector-emitter saturation voltage:
Vce(sat) = 0.1v, 0.2 v, If = 20 mA, Ic = 1mA.

 

U2, U4 opto leds are drawn the wrong way round, otherwise it should work.

Another solution would be a quad nand gate,if any of the inputs has a voltage removed it would give a high output to give an alarm,.

 

heh...yes....there are errors. Here's a redrawn version.
Anyway....you get the idea