0
\$\begingroup\$


I've been doing some research on rewiring clock circuits for a project, but I'm new to these types of pulse-counting circuits. My question at this point is:
If I replace the existing quartz crystal oscillator in a clock's timebase generator circuit with an oscillator of a different frequency, will the timebase change?
My goal is to make clock circuit with a shorter or longer space between pulses - making the clock's "second" shorter or longer. Is there a smart way to do this without use of arduino or additional stepper motors? Quartz clocks are fascinating so far- I look forward to learning more. Thank you for any input!

\$\endgroup\$
  • 2
    \$\begingroup\$ This sounds like it might be an XY problem. What range of frequencies are you trying to generate? \$\endgroup\$ – uint128_t Jun 20 '18 at 19:56
  • \$\begingroup\$ If you have a simple controller with a single crystal, then changing the crystal will probably change all timing on the controller the same way. BUT, there is probably a range of frequencies that will work. It could even be a narrow range. And there could be side effects. For example, Baud rates for any serial port will no longer be correct if you change by too much. USB may not work due to frequency error. Video outputs may not sync correctly. Basically the whole machine will run faster or slower. \$\endgroup\$ – mkeith Jun 20 '18 at 19:59
  • \$\begingroup\$ You need to tell us what change in frequency you want to acheive? How accurate the new frequency needs to be? Is it just one fixed frequency, or do you have multiple fixed frequencies you want? Do they need to change during operation? \$\endgroup\$ – Henry Crun Jun 20 '18 at 22:39
2
\$\begingroup\$

In principle, yes. If you have a device that uses a crystal oscillator to generate its timebase then you can change the timebase simply by changing the crystal oscillator frequency.

In practice things get more complicated. A given device will only operate over a specific frequency range and you can't violate that limit. Furthermore, if the oscillator circuitry is built in to the device and a simple crystal is connected to it then you have to consider the electrical characteristics of the oscillator circuit. The manufacturer of the device should provide information on selecting a suitable crystal, as well as the associated capacitors.

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

In addition to the "In principle, yes" answer from Elliot Alderson I'd like to add something:

Almost all clocks use a 32768 Hz crystal. That seems an odd number but it makes sense if you realize that \$2^{15} = 32768\$. If that 32768 Hz signal is divided (in frequency) by a chain of 15 divide-by-two circuits then you'd end up with a 1 Hz signal. So 1 pulse per second!

If you want to change that 32.768 kHz crystal for a different one I bet that such a crystal will be extremely hard to find. That's because there is almost no application where such a low crystal frequency is useful. Also with a higher frequency crystal and some (simple) chip or microcontroller, almost any frequency can be generated in that low (32 kHz) frequency range.

There are many different crystals (and crystal oscillators) in the 1 MHz to 50 MHz range but that's a much higher frequency. Very likely too high for a clock circuit to work with.

So yes your idea could work but will probably be impractical with any clock using a 32.768 kHz crystal and since those crystals are also very cheap, almost every clock uses one of those.

\$\endgroup\$
  • \$\begingroup\$ There are lots of crystals right around 32768 Hz. I have substituted different crystals as a way of evaluating timing margin before. You can go up or down by several percent from there and still find crystals. Just don't expect your serial port to work. \$\endgroup\$ – mkeith Jun 20 '18 at 20:58

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.