Comparator help.

[Oldcodger]

"Advice needed. I have a problem which needs a circuit to switch a relay depending on the voltage. At supply V1, the output should be HIGH, but when voltage on supply line rises to V2, the output should go low. My ideas are centred around a 741 comparator circuit. One input is tied via a zener to a fixed voltage, with the other input fed from supply via a pot. Any suggestions welcome, as it's over 30 years since I dabbled in this. "

[Biggles]

There are better chips to use than the 741, as I seem to remember the output doesn't quite fall to 0V. Use one of the dedicated comparator op amp chips instead. I can't remember a number off the top of my head but could find out. You will probably also need a switching transistor to increase the current output enough to switch the relay. The configuration will depend on whether you need the switch to happen on an increasing voltage or decreasing voltage. Peg one of the inputs to a mid point voltage with a zener and use a potential divider across the supply on the other to adjust the set point. A couple of capacitors across the inputs will reduce false triggering by slugging the response. I use one on my motorbike to provide a simple low battery volts alarm, such as if the alternator fails.
Alan.

[TrevorG3VLF]

If you are using the power supply to drive the comparator which is measuring it, then some carefull design is necessary.
The comparator has two inputs these should be connected the way round to get the correct action. A small amount of feedback to the + input will give fast action and some hysteresis.
Comparators usually have an open collector output so will only pull a voltage down. A PNP transistor may be the easiest way to interface.

Some numbers are required.

[Oldcodger]

I'd decided on a 741 as I have a few in my box+ it's not so bothered by supply ripple problems. O/P- I recognise the current output limitations of the 741, and my design at moment includes a BFY50 to drive a relay.
The circuit will be set to switch on as voltage drops below a set voltage,and off as voltage rises, with a differential of perhaps half a volt.
At present I'm looking at tieing the non inv input(+) to a zener diode ,with the inv input floating with a divider on the voltage input. And with the suply volts, well within the 741 operating limits.

[Argus25]

It will work if you keep the input voltage ranges within recommended limits and as noted that the 741's output voltage won't swing rail to rail, but you can choose a suitable series base and base to emitter resistor value for your output transistor to ensure the output transistor switches on and off properly from the 741's output.

One advantage of the 741, is that it is very sluggish with a low slew rate, so your switching circuit will be intrinsically RF immune. With high speed comparators they always need hysteresis (+FB added) and they sometimes can go unstable, at least with the 741 you won't have any of these issues. And obviously speed is not an issue as you are switching a relay, so the "better" and faster devices have no advantage in the application, and just add risks of instability. I keep a stock of mil spec 741's just for these sorts of applications. Also, because you are switching a relay, it has magnetic hysteresis, meaning its drop out voltage is lower than its cut in voltage, so you don't need to bother with any + feedback around the 741.

[TrevorG3VLF]

Also, because you are switching a relay, it has magnetic hysteresis, meaning its drop out voltage is lower than its cut in voltage, so you don't need to bother with any + feedback around the 741.

I would disagree with this. Slow change of relay current will give high dissipation in the BFY50. Better to define the dead band with the 741 and switch firmly.

[Argus25]

I would disagree with this. Slow change of relay current will give high dissipation in the BFY50. Better to define the dead band with the 741 and switch firmly.

I doubt it, the open loop gain of the 741 is about 100,000, it will switch firmly.