xmega A3U TOSC pins

Go To Last Post
8 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi,

I am going to use ATxmega128A3U in my design. I need to use both 16MHz external crystal for system clock and 32.768kHz external crystal for RTC. Moreover, I need to support both frequencies to the on-board peripherals via CLKper (PC7) and RTCOUT (PC6) pins.

Though the datasheet shows connection of the RTC crystal to TOSC pins on PORTE, only the grey color in "pinout diagram" leads to derivation that the TOSC1/TOSC2 signals might be connected to pins PE6/PE7. However, I could not find such information explicitly written anywhere else in the datasheet. Even the "Pinout and Pin Functions" misses the TOSC pin assignment.

Could you, please, confirm that the following configuration is correct?

16MHz - XTAL1/XTAL2 (PR1/PR0)
32.768kHz - TOSC1/TOSC2 (PE6/PE7)

Thank you for all your help
Ota

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

According do A3 datasheet, PE6-PE7 are TOSC pins in A3 series and this should be the same for A3U series.

Ozhan KD
Knowledge is POWER

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Okay, I will try it. Thanks

Ota

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hello
I'm going to do similar project with 2 external cristals.
Could You please write here effects of Your try @feifao1 ?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I use two xtals on a couple of products and there are no issues.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:

I use two xtals on a couple of products and there are no issues.

I'm curious--as [most?} Xmega models have "10.3.2 32.768kHz Calibrated Internal Oscillator", is an external 32kHz crystal that much more accurate as to justify adding another part?

From the datasheet, at a constant supply voltage there is about 0.5% swing over the entire temperature range. If an external watch crystal is 50ppm per degree C, that would be 0.4% over an 80 degree temperature swing--about the same. Are you then using 20ppm or 10ppm watch crystals to minimize this effect?

I can see that the calibration register has quite coarse steps and the curve looks like a sawtooth. So you can't always get real close?

For my industrial apps with AVR8 + RTC (often DS1305), I consider "a minute a month" good enough. A batch of samples will look like a classic bell curve around the ideal, and few outliers. Indeed, at 2 seconds per day that is less than 0.01% typical/uncalibrated. I guess that is much better than the internal 32.768 clock source.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

We use 10ppm xtals. They are about 7p each. 50ppm is useless for any kind of medium or long term RTC, seconds of drift per day.

We temperature compensate ours down to less than 1 minute a year.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Could you, please, confirm that the following configuration is correct?

16MHz - XTAL1/XTAL2 (PR1/PR0)
32.768kHz - TOSC1/TOSC2 (PE6/PE7)

Atmel technical support confirmed to me that the atxmega128a3u and atxmega64a3u:
TOSC1 = PE7
TOSC2 = PE6