Using RC time constant as a timer

[strantor]

I was reading this thread: https://www.electro-tech-online.com/threads/need-help-with-a-time-delay-cirsuit.121399/

and I had an idea. I did not want to post on his thread, as I have no idea if this works, so I'm asking here in my own thread.

macgyver states:

what i am needing is a time delay that requires you to hold down the switch for say 30 seconds to start another timing circuit that goes for say 10 mins...

My idea is to use a resistor, comparator, and capacitor.
Could you wire the button to a resistor, with a capacitor to ground and calculate the time it takes to charge the capacitor for given resistor (choose resistor based on time requirement) and wire the positive side of the capacitor to a comparator. have the comparator with a reference voltage same as supply voltage so that when the capacitor reaches full charge (ref voltage) the comparator will change state and activate a latching relay.

could that work? would it be reliable?

 

[crutschow]

Yes, it's a very common technique. Variations of such a scheme are used for many types of analog delays and timing (including in a 555 timer). For long delays the resistor and capacitor values tend to get large.

 

[simonbramble]

This will work, although the 555 timer is based on exactly this. The error in the system would be determined by the tolerance of your resistor (and more specifically) the tolerance of your capacitor (which will be much larger than that of the resistor) and the accuracy threshold of the comparator.

You dont want to set the supply rail as the trip point. Practically it takes about 5 time constants for the RC to charge to its full value, but theoretically using (1-e^(t/CR)) it will never get there.

Look at a 555 timer circuit. This will be a more practical way of doing it and indeed is the way many engineers do it in reality