Hi,
Really don't want to seem unpleasant, to you or anyone. Here is a circuit that works in simulations (and therefore, from my real-world experience of this exact kind of circuit, with fine tuning of timing components only, will work). Do expect a little unpredictable drift of the 3mins on/3 mins off "clock timer" as 66uF is not a precise method of creating timing. For greater repeatable precision, the way around that would be to make that first astable timer far less time (half the capacitance or much less, preferably - I don't like going over 10uF with 555s), and connecting it to another CD4017 stage to get another 10 (really 9) time stages out of the whole circuit - roughly 180 second (3 minutes) long on and off clock pulses rather than 360 seconds (6 minutes) long. You could use a CD4060 (divide by x ripple counter) instead of the 4017 I recommend, and use an astable 555 as the clock rather than RC network or XTAL for the 4060 - but, this can become quite annoying as working out the clock timing based on repeatedly multiplying or dividing an initial clock pulse length or the desired time by 2 can result in repeatedly getting results that give frankly unrealistic clock times to create, especially with passive components, like e.g. 6.25 seconds and so on.
I hope you can use this, it is a complete schematic and exactly what you have asked for.
- - - Updated - - -
The schematic says TLC555. This is not a common 555 and I do not recommend it. I only use it as it is the least annoying version to use in the simulator due to pinout and similarity to SE555 and LMC555, a simulation with the TLC555 should always work in the real world with the TTL (NE, SA, etc.) and CMOS (LMC) versions I've used a lot. While most often used and seen in schematics are the NE555, SE555, SA555 and LM555, these are not as good for timing circuits as the LMC555. LM555 is the same as the SE et al 555s, it's not at all the same as the LMC555. LMC555 is CMOS and frankly, more precise; only "downside" is its output power is only 100mA, not 200mA as in SE/SA/LM/NE - but it's not common/recommendable to make circuits feeding the timer output to high power-draw devices as it lowers the device's output voltage a lot, it's better to use the 555 as a power source driver (e.g. for a BJT or MOSFET driving whatever is used) rather than as the power source itself.
Good luck with the endeavour!
- - - Updated - - -
Hi,
Whoops, fuzzy division in last post, to add a second CD4017 would be a(-nother) divide-by-ten, not a divide-by-two result. This is a two-stage version using two CD4017s - the astable timer capacitance is divided by 10, the gain in accuracy and repeatability is really worth the few extra parts and PCB space it requires. To add a third CD4017 would mean the astable capacitor were around 660nF - even better for repeatable timing events that don't constantly go over and under the estimated 1 hour required. Given that most ICs come in packs of 2 or 5 or 10, it's not a bad idea.
My objection to the CD4060:
1hr = 3600 seconds. Start dividing down to get a sensible clock for that with an RC combination that won't be wildly inaccurate each hour, and forget using a crystal for such long times with one 4060, even a 555 needs an unrealistic RC clock to get 1 hour...
3600s/2 = 1800s
1800s/2 = 900s
900s/2 = 450s
450s/2 = 225s
225s/2 = 112.5s
112.5s/2 = 56.25s
56.25s/2 = 28.125s
28.125s/2 = 14.0625s
14.0625s/2 = 7.03125s
7.03125s/2 = 3.515625s
- I've tried this and frankly, getting anything usefully near 3.515 seconds from a resistor and a capacitor is nonsense on more than one level, for that it's best to use a PIC or whatever similar device than can do precise time. That's why I'd use the 4017.