The right RTC , it's the jungle to choose it !

Thread Starter

LE-BU

Joined Nov 3, 2017
18
Hello, I have a project and I need to choose a RTC . My criterias are Year/month/day...sec / 3v supply / up to 20 min drift accuracy by year is acceptable / I2C communication / Alarm output to awake the µc / Temp -20 to 50°C , and of course the best price !!!! Someone knows an appropriate tool or selector ? or the best way , if you are the "king" of the RTC ... Tell me what ref you will choose for that ..
 

MrChips

Joined Oct 2, 2009
30,806
RTC stability is dependent on how well you can set up, trim and stabilize the RTC clock crystal. Accuracy will depend on the reference you choose to calibrate the clock.

Many high performance microcontrollers already have the low-power RTC circuitry built-in using a 32768Hz watch crystal. Even if it doesn't you can use the MCU clock for time keeping with 10ppm stability.

20 min/year is about 40ppm which is easily achievable with any quartz crystal oscillator.
 

Parth786

Joined Jun 19, 2017
642
Hello, I have a project and I need to choose a RTC . My criterias are Year/month/day...sec / 3v supply / up to 20 min drift accuracy by year is acceptable / I2C communication / Alarm output to awake the µc / Temp -20 to 50°C , and of course the best price !!!! Someone knows an appropriate tool or selector ? or the best way , if you are the "king" of the RTC ... Tell me what ref you will choose for that ..
see ds1307 module https://www.ebay.in/itm/302456012266?aff_source=Sok-Goog
it may be useful for you
 

Thread Starter

LE-BU

Joined Nov 3, 2017
18
RTC stability is dependent on how well you can set up, trim and stabilize the RTC clock crystal. Accuracy will depend on the reference you choose to calibrate the clock.

Many high performance microcontrollers already have the low-power RTC circuitry built-in using a 32768Hz watch crystal. Even if it doesn't you can use the MCU clock for time keeping with 10ppm stability.

20 min/year is about 40ppm which is easily achievable with any quartz crystal oscillator.
Really , you think I can use the MCU and expect to have less 20 min drift ? I have read a publication of atmel in this way but they not specify the accuracy --> http://www.atmel.com/images/atmel-1...ing-the-asynchronous-timer_ap-note_avr134.pdf
 

Raymond Genovese

Joined Mar 5, 2016
1,653
Hello, I have a project and I need to choose a RTC . My criterias are Year/month/day...sec / 3v supply / up to 20 min drift accuracy by year is acceptable / I2C communication / Alarm output to awake the µc / Temp -20 to 50°C , and of course the best price !!!! Someone knows an appropriate tool or selector ? or the best way , if you are the "king" of the RTC ... Tell me what ref you will choose for that ..
I would recommend the DS3231.

It fits well all of your criteria and is simple to implement and there is lot of existing software.

Your 20 min/year should be feasible, the data sheet advertises:
• Accuracy ±2ppm from 0°C to +40°C
• Accuracy ±3.5ppm from -40°C to +85°C

But, there is also an "aging trim" register.
 
Last edited:

Picbuster

Joined Dec 2, 2013
1,047
I used the mcp79511 lost about 1 sec a month however; the accuracy is mainly defined by the quality of the xtal.
it's available in SPI and i2c.

Picbuster
 

MrChips

Joined Oct 2, 2009
30,806
There is no one size fits all. You really have to tell us what you want to do.
Not everyone needs 20m/year accuracy, battery operation, -20° to 50°C.

There are many ways to maintain 1m/year accuracy:
  • constant temperature crystal oven
  • NIST internet time server
  • WWV and WWVH time signals
  • GPS
  • power line time base
If you need stand-alone RTC clocks, constant temperature crystal oven is one option assuming you have enough power to maintain the oven. If you are powered by AC mains, power line frequency is accurate to 1 second.

If you can relax the -20° to 50°C limitation, then all options are open.

I have built data loggers that are battery operated that run for a whole year in the field. The time base uses the MCU clock crystal for all time-of-day record keeping. We needed to know the time-of-day ± 5 minutes.
 

Thread Starter

LE-BU

Joined Nov 3, 2017
18
There is no one size fits all. You really have to tell us what you want to do.
Not everyone needs 20m/year accuracy, battery operation, -20° to 50°C.

There are many ways to maintain 1m/year accuracy:
  • constant temperature crystal oven
  • NIST internet time server
  • WWV and WWVH time signals
  • GPS
  • power line time base
If you need stand-alone RTC clocks, constant temperature crystal oven is one option assuming you have enough power to maintain the oven. If you are powered by AC mains, power line frequency is accurate to 1 second.

If you can relax the -20° to 50°C limitation, then all options are open.

I have built data loggers that are battery operated that run for a whole year in the field. The time base uses the MCU clock crystal for all time-of-day record keeping. We needed to know the time-of-day ± 5 minutes.
It's for an outdoor system with battery , and every day I need to activate an output during 20 seconds at a hour defined. But if there is a drift of 20 Minutes for one year , it's not a problem for the system
 

philba

Joined Aug 17, 2017
959
Is there some caution about the choise of the quality crystal ?
Yes, get a low ppm crystal and high quality load caps. Good layout is needed as well. The data sheet of what ever RTC you pick should give you those details and more. Page 11 on the Microchip one linked above.
 
Top