4060 Timer

[mike_mike]

Hi! I need a timer for a application which will light a LED light bulb for a limited amount of time. For example, I will set the time to 1 h and then I will switch on the circuit, and the count starts. When the time is about 1 h, then the timer turns off the LED light bulb, and it never turn it on again, until I reset the circuit or until I cut off the power. For switching the mains (230Vac) voltage I will use a 12Vdc coil relay.
I found on the internet the attached schematics, which of them is better and if they are not good, what schematic should I use ?

Circuit 1
http://www.zen22142.zen.co.uk/Circuits/Timing/24hour.htm


Circuit 2
https://ethcircuits.com/simple-timer-circuit-diagram/

[Benta]

It will be incredibly imprecise. The oscillator is running at a few Hz, and tolerances and leakage currents at the oscillator (especially the electrolytic cap) will be immense.
I'd suggest you simply take one of the cheap 8-pin MCUs on the market, clocked from a crystal or resonator.
Yes, you'll need to do a bit of programming, but that can also be fun.

[Kleinstein]

I have used a similar circuit for  an old charger. It was working reasonable OK, though with a film cap (AFAIR some 2.2 µF) and larger resistor for the requency. The relay control comes from the last counter stage to get the most divider.

Today a small µC would be the better choice.

[Cliff Matthews]

   I've got a fancy dual drivers-cam on my car that draws about 5 watts continuous. I also want to use the 4060, so so I'll have the accessory circuit latch-on a relay to battery power the cam and timer. After I turn off the car, I want to run the cam for 3 more hours of security recording, so the 4060 cct. can un-latch the relay for true zero consumption. I hate living in Canada at -40 C with dead batteries..

[Benta]

A better choice for this application is the MC14541. It's made for the job. But still not for hour-long delays. Minutes, yes.
If you absolutely want to avoid the MCU, then clock the 4060/4541 from another counter with a higher clock frequency.
A 4060 -> 4541 solution might be the way to go.

[mike_mike]

I also though about the microcontroller idea, but being a very basic application, I do not care if the relay is in ON state 1 h 0 minutes or 1 h and 10 minutes.

[Benta]

I also though about the microcontroller idea, but being a very basic application, I do not care if the relay is in ON state 1 h 0 minutes or 1 h and 10 minutes.

We're talking more like 30 min. to 1 hr. 30 min. in the best case. If that's OK, go for it. And the MC14541 (CD4541) is still the better choice.

[Zero999]

I have used a similar circuit for  an old charger. It was working reasonable OK, though with a film cap (AFAIR some 2.2 µF) and larger resistor for the requency. The relay control comes from the last counter stage to get the most divider.

Today a small µC would be the better choice.

Here's a PIC program I wrote a long time ago, which could easily be changed to do what the OP wants.
https://www.eevblog.com/forum/projects/my-first-pic-program/msg16630/#msg16630

[jpanhalt]

If you want to avoid a microcontroller and want the accuracy of a crystal, here are some circuits I built many years ago.  The Schmitt circuit is quite common (see the circuits used for MCU's).  The 555 circuits are a little less common, but also work.

[Kleinstein]

With a few µF of film capacitance and a resistors of a few MOhms one can rund the 4060 clock with a little below 1 Hz and get times to the 24 hours range. For the charger I had it adjustable up to about 20 hours and the LED with the full clock as an easy way to check in reasonable time. OK for some 10-20% for battery charging. It would take 2 x 4060 to get to a similar range when starting from a 32 kHz crystal and than one is limited to fixed binary steps.