> Keep the DS3231; it's very accurate. One of my clocks uses it, and it has > drifted about 10 seconds since daylight savings started more than 6 months > ago. My wristwatch demo board has been running 1.5 years and it's also within > a few seconds. If you want anything better, you'll probably need GPS or a > rubiduim time standard.
There are 10MHz rubidium standards available fairly cheaply in the surplus market (a glut of them appeared in a big round of upgrading cellular phone transmitters). > I got a few laughs from a $1 Chinese RTC board; wickedly inaccurate. Probably the "DS3231" in that was a cheap microcontroller emulating the protocol and running from an on-board RC oscillator. > I just swapped the DS3231 with a genuine Maxim device and use that in my > clock. Yeah, the rest of the components are likely acceptable, as they're not important for timing. > I recall most crystals are accurate to around 10ppm, which is about 1 second > per day. My interpretation of that is the manufacturing tolerances result in > multiple batches of crystals that will fall within 10ppm of their rated > frequency; I suppose you can pay more for cherry-picked units that are more > accurate. I doubt individual crystals drift back-n-forth, so you could add > some offset to your software and fine-tune it to each crystal if you dont > want to use RTC/GPS/Internet time. Crystals have both drift and temperature sensitivity. The temperature sensitivity can be dealt with by operating them at a fixed temperature in an oven (bulky and power hungry) or adjusting the frequency (or compensating for it in software) depending on temperature (this is what the DS3231 does). The temperature sensitivity also depends on the angle from the quartz crystallographic axes the crystal is cut (the usual "AT" cut has a cubic S-curve of temperature sensitivity with a flattish part typically around 25°C). Drift can be reduced by preprocessing the crystals in various ways (including "swept" quartz, where the crystal is subjected to a strong electric field at high temperature to remove imperfections and impurities). As you might imagine, in the pursuit of ever-higher accuracy, a bewildering amount of science and engineering is involved. - John -- You received this message because you are subscribed to the Google Groups "neonixie-l" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send an email to [email protected]. To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/FE3E84BF-B712-43AD-B94E-7FB0220E23FF%40mac.com. For more options, visit https://groups.google.com/d/optout.
