1
\$\begingroup\$

I am researching how to design a timing circuit that will activate a relay with a push button and deactivate the relay after two hours. My understanding is that I can feed the output of a 555 timer to the input of a 14-stage binary divider. Then feed the dividers' output to a relay driver. Then, after two hours of the divider receiving pulses, its output will change thus changing the state of the relay drivers output.

My question is, can a 555 timer reliably produce a ~1Hz pulse? And if so, will this be too low of a frequency for a binary divider to reliably handle?

Thank you!

\$\endgroup\$
1
  • 1
    \$\begingroup\$ You should use something based on a crystal, not a 555 timer. \$\endgroup\$
    – Justin
    Jan 22 '20 at 18:55
1
\$\begingroup\$

Yes, 1Hz is a reasonable frequency for the 555, especially for the CMOS version. But it may not be optimal.

The temperature stability as an astable multivibrator is typically +/-150ppm/°C for the bipolar version, provided the Ra is between 1K and 100K. To stay in that range, implies a capacitor of the order of 10uF, which is a large and expensive film capacitor or a tantalum capacitor. It will operate with higher resitances, but the stability will generally be worse.

Various CMOS versions such as the LMC555 have improved temperature stability (+/-75ppm/°C typical) and you could us a resistor more in the 1M range and therefore a 1uF film capacitor or even 10x 100nF NP0 ceramic capacitors in parallel.

If you're using the bipolar version it might make more sense to add divider stages and operate at a higher frequency such as 50-200Hz. It also makes it easier to trim the oscillator if you have a reciprocal-counting frequency meter.


As to whether the dividers will work- yes, they're virtually all static and will work down to DC provided the clock edges meet specifications. The only exceptions that come to mind are GHz-range prescalers that won't work at lower frequencies (like < 50 or even 500 MHz). If the divider you pick does not have a schmitt-trigger input then the rise and fall times may have to be fairly fast, but the 555 will produce adequate edges generally.

\$\endgroup\$
0
\$\begingroup\$

555 runs between 100kHz to 0.1Hz in a-stable mode. And depending of your divider (check datasheets some have only a inpulse input so they are not dependant of a frequency CD4020 I think) ; so this will just work as low as 1hz or lower

a good practice when you are not sure about precision is to use in circuit trimmers & variable-capacitors to get high precision around limits until it fix any bugs.

\$\endgroup\$
0
\$\begingroup\$

+1 for at least not trying to do it without the divider. Pick a number that will give you a possibility of calibrating the 555. If you divide by 213 then a 1.138 s clock will give you two hours and it would be reasonably easy to calibrate and adjust using a stopwatch over ten cycles or so.

From what I can see the 555 has a temperature stability of 50 parts per million (ppm) per degree Celsius change in temperature which is equivalent to 0.005 %/°C. Build this into your calculations along with drift of the timing capacitors and resistors and you can check if it is good enough for you.

My question is, can a 555 timer reliably produce a ~1Hz pulse? And if so, will this be too low of a frequency for a binary divider to reliably handle?

No problem for both questions.

\$\endgroup\$
0
\$\begingroup\$

A better option is a 4060 which has 14 stages and two inverters you can use as an oscillator, making the 555 unnecessary.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.