My info about the 16-bit counter is just because "that's the way it's done", you take a 32.768kHz (often referred to in shorthand as 32kHz, but in the case of a RTC oscillator, it's going to be 32.768kHz, not 32.000kHz), and you feed that signal into a binary counter gives you a 1/2 frequency divider per (n-1) bits, so that the MSB of a 16-bit counter clocked at 32.768kHz switches from 1 to 0 and back again at a rate of exactly (omitting tolerances) 1s, which is used as the RTC clock source. I'm pretty sure that using a 1Hz oscillator directly would have pretty horrible drift in terms of tolerance vs doing it this way.nocash wrote:The http://www.snescentral.com/pcbboards.ph ... VC-LJ3R-01 photo doesn't show a part number. Or are those tiny cylindrical oscillators always having 32.768kHz?qwertymodo wrote:Also, from the SNESCentral pics of the S-RTC board, that's just a run-of-the-mill 32.768KHz tuning fork. It gets fed into a 16-bit counter in order to get out millisecond clock ticks. So, mystery solved, you now have all of them.
I think I have also came across circuits that used "32kHz" in the past... but maybe they did mean kilo=1024 in that circuits. And aside from the 32.xxxkHz range, theoretically it should be even possible to put something like a 20MHz quartz into that package.
Where do you have the info about the "16-bit counter" from? Do you have a matching Sharp datasheet?
As to whether or not you can get MHz frequencies in the same package, yes you can. However, the fact that it is being used as a RTC basically means 99% certainty it's a 32.768kHz. If you REALLY want to confirm it, then you'll have to get better photos, or get your hands on one. But it's a pretty safe bet that I'm right
Edit: I forgot how to math... 32,768Hz * 2 ^ -15 = 1s, not 1ms
