NANO with DCF77-receiver and too much drift (DCF77_Scope)...?

[Jo-Achim]

Hi, I have a little problem with the Libraries v.3.0.0 and before on my Arduino Nanos (IDE 1.6.11). ;-)

E.g. "Simple_Clock" will not synchronize; so I see "Initializing..." and a lot of '.' only.

Because my Nanos with Conrad-DCF77-receiver are working (https://arduino-hannover.de/2012/06/14/dcf77-empfanger-mit-arduino-betreiben/) with synchronization normally within 2 - 3 minutes), ...

Looking at the output from "DCF77_Scope", I think, I found the reason: the output drifts to the right side within 1.000 seconds round about 340 ms.

    1, +---------+---------+---------+---------+---------+---------+---------+---------+---------+---------
    2, +---------+---------+---------+---------+----9XXXXXXXXXXX---+---------+---------+---------+---------
    3, +---------+---------+---------+---------+----9XXXXXXXXXXXXXXXXXXXXX9--+---------+---------+---------
    4, +---------+---------+---------+---------+----7XXXXXXXXXXXXXXXXXXXXX3--+---------+---------+---------
    5, +---------+---------+---------+---------+----9XXXXXXXXXXXX--+---------+---------+---------+---------

  995, +---------+---------+---------+---------+---------+---------+---------+--------7XXXXXXXXXXXXXXXXXXXX
  996, XX--------+---------+---------+---------+---------+---------+---------+--------3XXXXXXXXXXXX5-------
  997, +---------+---------+---------+---------+---------+---------+---------+--------2XXXXXXXXXXXX6-------
  998, +---------+---------+---------+---------+---------+---------+---------+--------3XXXXXXXXXXXXXXXXXXXX
  999, X8--------+---------+---------+---------+---------+---------+---------+--------2XXXXXXXXXXXXXXXXXXXX
 1000, XX3-------+---------+---------+---------+---------+---------+---------+--------2XXXXXXXXXXX8-------- 

This result I get with my best Nano (with the lowest drift)! (I have 3 Nanos from 2 different manufacturers tested.)

Is my suggestion correct that the 'no-synchronize'-problem is based on this drift?

Can you say, how high is, based on your filter-algorithm, the maximum drift per unit of time / or the max. tolerable divergences of the 16-MHz oscillator in Hz or in ppm? Are they any variables available to solve the problem in software?

Thank you for help.

Best regards, Joachim.

[udoklein]

340 ms / 1000 s = 340 ppm drift. You are correct that this causes the issue. This is an indicator that you are running with a resonator instead of a crystal. The issues is documented here: https://blog.blinkenlight.net/experiments/dcf77/dcf77-library/. Search for "Hardware incompatibilities". Solution: get a board with a crystal or replace the resonator by a crystal + proper load capacitors.

[Jo-Achim]

Hello Udo,

thank you for information's.

First I will see to get an original (Gravitech) Nano - hopeful with crystal oscillator and proper load capacitors - for reference.

Best regards, Joachim.

PS: Great work!

[udoklein]

On 23.09.2016 11:14, Jo-Achim wrote:

Hello Udo,

thank you for information's.

First I will see to get an original (Gravitech) Nano - hopeful with crystal oscillator and proper load capacitors - for reference.

Best regards, Joachim.

PS: Great work!

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/udoklein/dcf77/issues/17#issuecomment-249142713, or mute the thread https://github.com/notifications/unsubscribe-auth/AFG9hNkGybDHbYa8N1ik2OxsebLB5eyAks5qs5hugaJpZM4KEI_k.

According to the Eagle files the Gravitech has a resonator. You may be luckier at *bay:

http://www.ebay.de/itm/New-design-Pro-Mini-atmega328-5V-16M-Replace-ATmega128-Arduino-Compatible-Nano-/172271346955?hash=item281c2c250b:g:0S8AAOSwo6lWL0gM

has a crystal and is very cheap. If you like my project you can donate here: https://blog.blinkenlight.net/good-karma/

Best regards, Udo