I am trying to build a circuit that turns on an LED about every hour for 1 minute, and started by building a 555 timer circuit (according to http://www.csgnetwork.com/ne555timer2calc.html) with calculated R1 at 540000 ohms and R2 at 9000 ohms. I have the capacitance at 9400 uF by two 4700 uF capacitors in parallel. When I tried using this however, the led never turned on. I then tried changing the values to make high time two minutes and low time one minute, which worked fine. I also tried 10 minutes and one minute, which caused the led to turn on first after fifteen minutes and on for seven seconds and then continue turning on every four minutes. Is the 555 timer inaccurate or could I be doing something wrong? Should I be using other methods to construct this timer? Any advice would be greatly appreciated.

 

And you will be lucky if the LED turns on at 60min +- 15min, or at all. This delay is too long for a 555. Look into using a 4060 chip instead.

 

I am trying to build a circuit that turns on an LED every hour for 1 minute, and started by building a 555

Start over. They are very sensitive (read "unstable") over time as temperature impacts the resistor and capacitor. Especially on the second to minute scale.

Look up CD4060. It uses a crystal oscillator like a watch. Easy and reliable to get half second pulses. Then use a flip-flop and decade counters to do the rest.

Damn, mike beat me!

 

Thank you I'll look up the 4060 timer.

 

...
Look up CD4060. It uses a crystal oscillator like a watch. ...!

Not only a crystal. You can use an RC oscillator, too. Right off the TI data sheet:

 

Is there any timer that would be ideal for this project?

Other than a purchased on/off interval timer, start with the CD4060 and a 32768Hz watch crystal.

If you are ok with an off-the-shelf solution, then there are various lab timers available.
http://www.omega.com/subsection/timers-counters.html

 

Not only a crystal. You can use an RC oscillator, too. Right off the TI data sheet:

View attachment 77538

With RC you don't have the accuracy of the crystal.

 

Other than a purchased on/off interval timer, start with the CD4060 and a 32768Hz watch crystal.

If you are ok with an off-the-shelf solution, then there are various lab timers available.
http://www.omega.com/subsection/timers-counters.html

The circuit as you said that I show in my blog -- 1Hz timebase - using CD4060,CD4013,32768hz crystal.

 

To me it seems that the 555 for long time periods is much maligned. In the past have used 10,000 uF & 1 M, just tried a 22,000 uF & 1 M for a time of 6 hours. Checked leakage of a 3300 uF = 10 M. Just have to select low leakage caps.
My Heathkit VOM X 10k ohm scale uses a 9 V battery, so I pre charge C under test to 9 V then discharge with 10k ohm back to full scale. In about ten min. reading will be leakage resistance. It is best for leakage R to be at least 3 X timing resistor. If timing & leakage are = C can only charge to 1/2 V & will never reset 555.