Hello all I been Puzzled by this for a long time it's been in the back of my mind and i always wondered why is the RTC real time clock 1307 not so Accurate Why is that? can someone please explain to me i know it loses time and seconds each month or or year but each one is different.
Probably comes down to how good the crystal being used is.
Hey cross roads yeah that's what i always think. Me i think that the chip is fine but the crystal is always a problem but i don't know enough about it to come to answer in my head.
No clock is perfectly accurate (although atomic clocks are pretty close to perfect).
Is the RTC module any worse than your wristwatch or the clock on your wall? The typical quartz crystal is rated for accuracy of about 50 - 100 ppm (parts per million). I think 100ppm is about 1 minute out of one week, but you should probably double-check that. I think a lot of wristwatches & clocks are better than that.
AC powered wall clocks are often synchronized to the 50/60Hz power line frequency and the power utility usually speeds-up or slows-down the frequency very-slightly as required to keep the long-term average essentially perfect. So, it's possible to have a wall clock that's nearly perfect over a year or more as long as there is no power outage.
The clock in your phone (and probably in your computer) is synchronized to an atomic clock. But, if you are not connected to the phone network or Internet for some period of time the time will drift-off.
I notice the RTC 3231 or is it the RTC 3230 real time clock ones are much more better but cost wise is higher i guess they are better also i found out there only lost about 8 seconds per year from what I'm reading not sure if that's correct or not but why are they better ? what makes them better?
It's all about the accuracy of the crystal. [u]Here's a datasheet[/u] for a crystal.
The watch/clock manufacturer may even select the most accurate crystals for their better products. (If it's rated for 50ppm that's the worst case and some will be better). It's also possible to measure the actual frequency and put a correction factor in software, but I don't know how often that's actually done.
Sometimes the crystal is placed in an "oven" to keep it at constant temperature and that keeps the frequency from drifting-around with temperature. (Of course you can't put an oven in a wristwatch or an RTC module, and it's not practical for a consumer clock either, but it might be done for a time-standard or frequency-standard.
This application note addresses the accuracy issues
http://www.maximintegrated.com/en/app-notes/index.mvp/id/58
Essentially, the crystal, the capacitive load match, and the temperature.
Just download the datasheet for the DS3231, it explains how it works in great detail. Basically, it uses a precision temperature sensor which is then used to compensate for the temperature coefficient of the crystal frequency. Quartz has a very predictable coefficient of expansion which then in turn relates to the frequenceny of oscillation.
As for the error of the DS1307, the 100 ppm accuracy assumes the worst case. To calculate the worst possible error for a week, calculate the total number of seconds: 24 x7 x 60 x 60 = 604,800. Now multiply that times 100 ppm = 0.001 x 604,800 = 604 seconds which is approximately ten minutes. Pretty bad but this assumes the crystal is at its maximum or minimum ambient temperature spec. In real life, the ambient temperature changes but it tends to average out over time in most applications.
Thank you all for the information i have learned a lot and yes Downloading the datasheets on both the 1307 and 3231 thank you. I finally have what I been trying to understand for some time 