Build any working circuit, that you know is going to give you the results your looking for, rather than trying to design it yourself, then tweek it to your needs.

 

Ok, thanks. again, I am really sorry for been so slow

 

I know that you wanted a discrete logic solution but something like is begs to be done via a mcu.

 

I think I'll stick to the solution already posted, it took a while for me to get the grasp of what to do I dont want to have to start the learning process all over again

 

I know that you wanted a discrete logic solution but something like is begs to be done via a mcu.

Pish posh... It took a while but I have the timing and decoding down to an elegant solution, one that leads to semi-independent adjustability of the on and off times, with power-on reset, a single 4060, no other gate packages, no uC, and most important of all -- ***no 555***.

The oscillator frequency is chosen such that the half-period of the Q13 output is 6 hours. After Reset, Q13 sits low for 6 hours, then goes high and drives the output transistor through D4. At this time, oscillator pin 10 is low. 5.3 seconds later it goes high. Now both D2 and D3 are reverse biased, and R3 can discharge C2 in a few milliseconds and reset the 4060. This turns off Q13. Since 5.3 seconds is less than the 6 seconds requested, C4 and R4 stretch the output transistor ON time about 2 seconds.

R3 and C2 form a power-on reset timer, and filter any meta-output noise from the diode-AND connection.

To adjust the off time, R1 can be replaced by a fixed and adjustable resistor in series. This will adjust both the on and off times proportionally. To adjust the off time, R4 can be replaced by a fixed and adjustable resistor in series.

ak

 

I'll stick to the solution already posted

I think you will likely find the posted solutions quite "fragile": it may work well in an airconditioned environment for a few days but once exposed to the elements, or over a long period of time, it may not be able to keep accurate timing - all depends on your needs obviously.

If you can, try to stick to a crystal oscillator if you think it needs to work reliable over a wide range of temperature.

 

It should activate for 6 seconds every 6 hours.

In discrete land:

solution #1: find a 1-hr counter, and feed it a 1/6Hz signal;
solution #2: find a 6-hr counter, and feed it a 1Hz signal;
solution #3: find a 24-hr watch, and feed it a 6Hz signal;
solution #4: find a smart phone and set the timer to 6hr / repeat.
...

 

In discrete land:

solution #1: find a 1-hr counter, and feed it a 1/6Hz signal;
solution #2: find a 6-hr counter, and feed it a 1Hz signal;
solution #3: find a 24-hr watch, and feed it a 6Hz signal;
solution #4: find a smart phone and set the timer to 6hr / repeat.
...

How do you go from extremely helpful in some posts to captain chaos in another post? Have you even read the OPs posts? Or is this your project? If so, yes, I am happy to report that all of your Solution options will work (depending on what kind of output you need). Danny, If you need verification for additional options for your projects, please open a new thread.

 

In discrete land:
solution #1: find a 1-hr counter, and feed it a 1/6Hz signal;
solution #2: find a 6-hr counter, and feed it a 1Hz signal;
solution #3: find a 24-hr watch, and feed it a 6Hz signal;
solution #4: find a smart phone and set the timer to 6hr / repeat....

None of your "solutions" address the only tricky part of the OP request - the 6 second activation time. Four of your "solutions" still require a very long period timebase clock. In fact, #1 is almost identical to what I posted. I notice that you offered no help in how to "find a 1-hr counter" or "find a 6-hr counter".

but once exposed to the elements, or over a long period of time, it may not be able to keep accurate timing - all depends on your needs obviously.

Actually, it all depends on the oscillator timing capacitor. +/-10% parts are readily available and cheap, and more than stable enough over the 40C temperature range of a typical home project.

If you can, try to stick to a crystal oscillator if you think it needs to work reliable over a wide range of temperature.

Without getting into bizzaro parts, the lowest frequency crystal is a 32.768 kHz watch crystal, and that isn't even a real crystal. Still, it would need a 30-bit counter to divide all the way down to 6 hours, and the cycle gating becomes much more complex because it locks you into multiples of a 1 Hz period, and neither 6 nor 21,600 are binary increments.

 

I feel the question I asked originally has been answered. I have to put the circuit together to test to confirm. Sine the timer is to power a propergator it will not be exposed to the elements. It will be nice and snug in a 12 degree room. So I don't fully understand stand how the elements come into play?

 

I feel the question I asked originally has been answered. I have to put the circuit together to test to confirm. Sine the timer is to power a propergator it will not be exposed to the elements. It will be nice and snug in a 12 degree room. So I don't fully understand stand how the elements come into play?

An RC circuit time constant can change with temperature.

 

I think as stated Earlier the window of free play would make this acceptable? Also. It is indoors so 20 on the thermostat all day? I could use heat sinks to cool more efficiently on the hotter months?

 

All of what you are trying to do can be done via a 8pin mcu, with more flexibility.

You may also look into the 4000 family of counters as well.

 

I could use heat sinks to cool more efficiently on the hotter months?

No need for heatsinks. The parts in this circuit are not intended to be attached to any kind of thermal management, and there is absolutely no need. Once you stated that the circuit will be operating indoors, all thermal issues with the circuit's accuracy and stability went away. I can go into more detail about why the thermal issue was brought up, and how changes in the air temperature *might* affect the circuit's timing accuracy, but since you have stated some very broad tolerances, none of that matters. I think you should concentrate on learning how the circuit operates.

Also, have you ever programed a microcontroller for a project, or done any other kind of computer programming?

ak

 

I do a lot of computer programming never done micro chips though

 

To program a uC to do this task, you need to know either the assembly language of the uC you select, or C++ (each uC has a variation), or maybe a strange variation of BASIC. You also need a programing adapter. And after all of that you still have to wire up a circuit to drive the motor. It will have 16 fewer total device pins compared to a discrete version, but still... Again depending on which uC you select, there might be a small development board available that has a place for the uC, a prototyping area for the output drive transistor, and a programming interface that talks to a serial or USB interface. Separate from all of that, each uC has its own weird control register structure that defines how each pin functions (input, output, analog, digital, etc.), something I've found to be the biggest PITA in a small project.

Once you are up to speed on a uC and its environment, whipping out small timer projects can be pretty quickly done. But the schematic in post #46 has 37 device pins if you add in two 2-pin connectors for power in and motor out. There is no way that wiring it up will take longer or debugging it will be harder than your first (or even your 20th) microcontroller project.

IMnsHO

ak

 

Well, I have experience in Java, C# and .net... of all the crap lmao. but I am sure when I do start playing with MC's i will be ok with c++ but for now I just want to build the circuit above

 

.net - sucks to be you.

 

lol, I just like the simplicity with basic apps. Don't really use it that much these days

 

This would be how I would implement it in a lowly 8pdip mcu.

#define SW_DURATION        (6sec)                //sw on for 6 seconds
#define SW_PERIOD        (/*1hr+30min+*/20sec)    //every /*1hr+30min+*/20 seconds

It is a timer that turns on for 6 seconds every 20 seconds - if you uncomment the code it goes to 1:30:20 just for example.

from a hardware perspective, it is considerably simpler than a discrete solution. Obviously, you will have to learn to code and program a mcu - it took me 5 minutes to get it going.