-1

How do I calibrate the RTC of energy meter IC 71M6541F?

Here are the details:

I am working on the above mentioned SoC on smart meter project. Using 32.768 kHz frequency I am getting an accurate time without any mismatch with global time. But in the crystal there is a small deviation so that I am undergoing 2 to 4 second difference in real time in 24 hours of time. How do I rectify this problem?

Peter Mortensen
  • 30,738
  • 21
  • 105
  • 131
Lokesh V Gowda
  • 79
  • 1
  • 9
  • This question is kind of off-topic, as you are asking about the behavior about a specific IC, rather than about a programming problem. Such questions are more suited for http://electronics.stackexchange.com/ or since it is such a niche question, it is best to ask Maxim tech support directly. – Lundin Aug 25 '15 at 06:56
  • 1
    I'm voting to close this question as off-topic because asking about the physical properties of a hardware device (crystal). – too honest for this site Aug 25 '15 at 12:38
  • @Lundin: That is not about the IC, but crystals and their tolerances in general. OP might ask how to add/subtract a value from an integer. – too honest for this site Aug 25 '15 at 12:40
  • Use an atomic clock. But that still has some deviation. You should learn about the properties of physical devices you use and their implications. – too honest for this site Aug 25 '15 at 12:43
  • @Olaf According to the datasheet, this particular device comes with numerous intricate on-chip calibrations. There's lots of things that could cause the clock to drift, for example incorrect temperature compensation. – Lundin Aug 25 '15 at 12:54
  • @Lundin: Just had a short peek. You are right. So OP should read the DS and any application note, if available. Now it looks more like a "please do my job" question. – too honest for this site Aug 25 '15 at 13:12
  • Energy meter for what? If it is mains you most likely have a very stable long-term clock source in the mains frequency, 50 Hz. I think in most places the utility companies are required to provide a long-term very high accuracy in the frequency (although the short-term accuracy can be off by on the order of 10 seconds (as accumulated error)). – Peter Mortensen Sep 06 '15 at 08:38

1 Answers1

0

The problem is that affordable crystals have always a deviation. Typical values are 30 ppm.

And as each crystal has its own deviation you need to calibrate each system or you need an external sync mechanism.

Peter Mortensen
  • 30,738
  • 21
  • 105
  • 131
jeb
  • 78,592
  • 17
  • 171
  • 225