Timer Circuit with 555 and 4060b

[chriswms]

Hello. I am new to the forum and to the actual building of electronics itself. I am trying to put together a timer which will be off for 8 hours and on for 29 seconds. Odd i know. I have pretty much figured out the design i think and the calculations for the resistors and capacitors to give me my desired time. My question at this time before I buy my materials is that I read that electrolytic capacitors values are not entirely accurate and in an 8 hour span it could increase the time by as much as 1 or so hours. Is there another type of capacitor that could be used to more accurately reach my goal? Thank you for your help.

[kam]

Hello chriswms.

If you get my advice, i would not use a 555 timer for such a long delay. There is no equivalent to electrolytic capacitors, except the new 'solid state' capacitors that i do not have a clue if they are good or not, and how stable they are. 8 hours is way too much. I suggest you use a crystal oscillator for example 32,768KHz, and with 256 and 128 bits division you get a pretty accurate 1-sec timebase. Furter divisions will get you the 28800 seconds (8 hours) and your 29 seconds. This makes the circuit rather complicated, but is very accurate. Do you need accuracy and how much?

[chriswms]

I need accuracy definitely on the 29 seconds. The 8 hour pause could be plus or minus 5 or 10 minutes. I've done a lot more research and found a schematic to implement into my design to give my 4060 a more accurate timing but, if my calculations are correct it will give a signal every 1/2 second and I don't understand how to get it to 8 hours. Using just the 4060 I understand how to get that to four hours with resistor and capacitor which will give me a 4 hour on 4 hour off time which will equal my 8 hours. Can you please offer any information or where to get it on changing the timing of the crystal to achieve my 8 hours? Any help would be MUCH appreciated. Thank you.

[chriswms]

I meant to add that I am using the 555 timer for my low time of 29 seconds while I was using the 4060 to time my 8 hours; then it sends the signal to the 555 timer.

[kam]

With one 4060 you may get some accuracy for the 29 secs, yet never too accurate. For the 8 hours with an RC oscillator i think is a little bit difficult. I cannot think of a way for example to test it and adjust it. You need to wait 8 hours every time to see if you need to increase or decrease the frequency. Do not NOT NOT rely on the theoretical calculations of an RC circuit. Remember that a resistor and a capacitor have a rather big tolerance. You will find out that the calculations have usually a big difference from the reality.

What i suggest is something like this:

http://en.wikibooks.org/wiki/Practical_Electronics/Plugins/1Hz_Oscillator

A crystal oscillator is very accurate and needs of no test/adjustment. You can rely on it, i do. This circuit will produce you a 1Hz pulse that you can use to count 29 seconds extremely accurate. Then, you can use the other flip flop that there is on the 4027 chip and divide the pulses once more, and get 0.5 Hz frequency. Use then the 4060 as frequency divider, and from the 14th output you get the 8 hours pulse very accurate.

[chriswms]

I appreciate all your time and help. I am going to start my project planning over for the third time I just started learning/researching about the divider chips. Mostly 47ls90, 92, 93. Which I was using in series I'm assuming its called to get 60x60x8 using a 2hz signal from a 4060 Q14 then to my 555 timer for my 29secs. But those 5 chips cost almost more than half of all the other parts at almost $2 a pc at the only site I can find them. So I'm restarting Basicaly my knowledge of electronics is just beginning as it only consists of the parts I have been studying (4060, 555, 47ls90 92 93 and a basic diagram of a 4060 using a crystal). What you are saying and the way I understand it seem to be different. If you take a 32768 signal into a 4020/4060 you get 2 hz which is .5secs (.25 sec on .25 sec off) so if you put .5hz in you get and even more extremely faster time on q14? Which means my power circuit will not be off long enough? I'm going to continue researching to see if I can get a better understanding of this.

[chriswms]

Ok, sorry for my lack of knowledge. After further studying how the crystals hz work with the counter I see what you are trying to say. But, by putting .5hz into the 4060/4020, the output on Q14 will be a little over 9 hours instead of my goal of 8. To make sure I am understanding, I got this by dividing the .5 by 16384 (output of Q4) which gives me 1/.000031 approx. (32760 secs exactly). So, how would I go about getting my goal of 8 hours?

[chriswms]

I figured out that if I use a 38400hz crystal and put it into a 4060, from Q14 to a 4040, from Q2 (/4), to a 4060 and the output at Q14 will give me 7.85hrs. Now what would you advise me to put that 7.85hr time in to get a separate 29sec signal that will turn on my 120vac device for only that amount of time? Because it seems you don't care to much for the 555 timer And again THANK YOU for all your time and help.

Also, do I use the same value of cap and res as the 32785 would use to get a 2hz signal from the 4060.

[kam]

It seems that you learn fast

So, You have a pulse (we call it long pulse) that is HIGH for about 8 hours, and LOW for another 8 hours correct? What you really need now, is to get 29 seconds a relay armed, whenever the long pulse is going from LOW to HIGH. This happens 1.5 times a day (correct me if i have understand wrong).

You need a Monostable multivibrator circuit to be driven by this long pulse. Check this page

http://pcbheaven.com/wikipages/Transistor_Circuits/

In section 6 i have the circuit that you need. Watch the video. Select proper RC values for 29 second. The GREEN LED is turned on whenever the pushbutton is pressed. The pushbutton may be kept pressed for 8 hours (from the long pulse), but the LED will only be turned on for 29 seconds! It then shall be turned off. The LED will stay off until the control pulse goes to LOW and back to HIGH again (after 16 hours-29 seconds)

[chriswms]

The way I was understanding it is that a 2hz output originally (using 32785hz/4060) gives you a .5 second output which is .25 on, .25 off (per internet resource) So I thought that after my division chips I am getting a freq. of 0.0003576hz (7.85hrs)(approx) which I assumed to be 3.92hrs on, 3.92hrs off. This way every 8 hours (or thereabout) I get a high signal for four hours then off for four hours. And when it goes high again at the 8 hour mark my other timer kicks in for the 29 seconds of power I require. I got this by using frequency = 1/time.
2hz=1/x which equals to x=1/2 divide by to and you get .25 on .25 off

Is this correct?

[kam]

Yes of course is correct. I thought you wanted longer time. You can divide one less and get your desired 4hours HIG-4hours LOW. Then, you will have 3 times a day the 29 seconds running.

In the same page that i sent you before with the , you can find the rely driver that you need to connect. What are you preparing really? Plant watering? Some kind of pump? 

[chriswms]

Actually its quite odd, but my dad raises chickens. He has an incubator and has to rotate the trays every 8 hours. It takes approximately 29 secs to rotate the tray.

I'm going to have a 5vdc supply powering my board. I think this relay will work without a transistor? http://mouser.com/Search/ProductDetail.aspx?R=T7CS5D-05virtualkey65500000virtualkey655-T7CS5D-05

I am having a hard time figuring out how to connect it even looking at the data sheet. It has 5 pins. one i am guessing is ground,  another for 5v power supply, two more for my 120vac leads that will connect and give me power, the 5th?

Also, do I use the same value of cap and res as the 32785 would use to get a 2hz signal from the 4060 that it shows on the site you sent of the oscillator? Of course I will be using a 38khz instead.

After I get those two problems figured out I should be able to finish drawing up my schematic and order my parts. I may try and find a way to fun it from 120vac with a transformer I'm guessing to convert to 5v if its not to expensive as I'm worried about battery life.

[kam]

the relay will be the most easy part, believe me. Two pins are for the 5V coil indeed. Does not matter which one is positive and negative. Looking from the pin side, you see on the left hand 3 pins placed like < and on the other side (right) 2 pins placed like |. from the left side, the top and bottom pin are the coi. There you connect the 5 volts to actuate the relay.
The middle pin (on the left side) is the common. This pin is related to the other 2 pins on the right side. When the relay is relaxing (no power on the coil) then the middle pin makes contact with one of the two right pins. When power is on the coil, it breaks this contact and makes contact to the other pin (on the right side). You need to use this contact, the one that is made when power is put to the coil of the relay. The technical terms of these contacts is NC (Normal Closed) for the contact that has contact when the coil has no power, and NO (Normal Open) for the contact that makes contact when the coil has power. You will use the NO contact.

Also, do I use the same value of cap and res as the 32785 would use to get a 2hz signal from the 4060 that it shows on the site you sent of the oscillator? Of course I will be using a 38khz instead.

I really do not understand your question about the cap and the resistor.

The power supply is to be taken from mains (110/220 to 5 volts) no question about it.

[chriswms]

Ok, on this site: http://en.wikibooks.org/wiki/Practical_Electronics/Plugins/1Hz_Oscillator ,, they used R1 (10M) C1 (10pf) C2 (100pf) to regulate crystal I assume.
On this site: http://www.reuk.co.uk/Making-Very-Accurate-Timers.htm ,, they use R1 (15M) R2 (330K) C1 (10pf) C2 (10pf)
And on this site: http://electroschematics.com/225/1hz-generator-and-2hertz-oscillator/ ,, they use R1 (10M) R2 (220K) C1 (39pf) C2 (4-40pf) and all used 32786 crystals.
I, however, am using a 38khz crystal.

So would I use the same resistors and capacitors as if I were to use a 32.7khz or do I use different parts?

And if I were to use the same parts which one is the better one to follow as all three are different?

Thank you so much for your help thus far.

[kam]

Use the same parts. The result will be the same. use for example the circuit as in the second page. (http://www.reuk.co.uk/Making-Very-Accurate-Timers)