[time-nuts] iPhone keeping better time?

Miguel Gonçalves mail at miguelgoncalves.com
Wed Nov 16 16:38:01 UTC 2011


Hi David!

On 16 November 2011 16:18, David J Taylor <david-taylor at blueyonder.co.uk>wrote:

>  Yes, Miguel, someone did mention an NTP synched clock some time back, and
>>> I thought it might be a fun project.  Based on an Arduino board IIRC?
>>> Although I don't think it had Wi-Fi by default....
>>>
>>
>>
>> I have it running at the moment. Have to build a case tough. See attached
>> picture.
>>
>
> It looks very smart!


I have a small plate of frosted plexiglass in front of the display to
diffuse the strong light of the LED. I am about 4 meters away from it and
it looks great. The digits are 29 mm tall while with 5 mm LED they would be
50 mm tall. Too big for my office but perhaps some might want larger digits.


>
>  It syncs at the 9th, 19th, 29th, ... second mark from a local GPS based
>> NTP
>> server.
>>
>> On every sync, the timestamp returned from the NTP server is on the 6 ms
>> mark this means that the local clock of the Arduino drifts a lot.
>>
>
> 6 ms per 10 s.  Outside the 500 ppm allowed for NTP!  <G>


A bit too much... The RTC will help a lot. Actually, the RTC performance is
great. 2 ppm means 0.02 ms every 10 seconds (0.12 ms every minute). Not
bad! With 3 local GPS stratum 1 servers peered together I can pool them
every 10 seconds and maintain sub ms accuracy. Inova clocks talk about 200
ms accuracy. Even with Internet NTP servers and pooling every 5 minutes the
accuracy of this RTC would be 0.6 ms. 200 times better than the commercial
product. I don't understand why a company like Meinberg sells these clocks!


>
>  I am
>> installing a realtime clock (Chronodot) this weekend that has an accuracy
>> of +/- 3.5 ppm from -40C to 85C (I read somewhere that between 0 and 40C
>> it
>> is 2 ppm). This RTC can output a square wave signal at 1 Hz and Arduino
>> can
>> read that and use it to update the display at the exact second mark.
>>
>> With the RTC and synching every 10 minutes (9th, 19th, 29th, 39th, 49th
>> and
>> 59th of every hour) I expect a maximum error of 1.2 ms (based on 2 ppm).
>> My
>> eyes can't read that :-)
>>
>
> Nor mine!  On my PC, the radio pips, and the hands on the analogue clock I
> wrote, appear to be precisely in sync, but I expect that means within a few
> tens of milliseconds.


:-) Perhaps we both need to wear glasses :-)


>
>
>  A neat feature I added is that when the clock can't synch it won't show a
>> time. It will show -- : --.
>>
>
> Shouldn't that be the classic video recorder display?  A flashing 00:00 or
> whatever?


00:00 flashing... done!

I'll post a video for you to see soon. You'll tell me if you like it or
not. You'll be my beta tester :-)


Yes, I would be interested, but others might as well, so perhaps put it up
> on the Web somewhere?


I'll do that as soon as possible.

Cheers,
Miguel


More information about the time-nuts mailing list