[time-nuts] Low-long-term-drift clock for board level integration?

Bill Woodcock woody at pch.net
Mon Feb 20 01:48:15 UTC 2012

Hash: SHA256

On Feb 19, 2012, at 5:07 PM, Bill Hawkins wrote:
> If you are going to correct it with NTP, a simple crystal
> oscillator will do.

Yeah, my assumption was that something like a DOCXO or a VCTCXO would be about the best I'd get within budget.  But I'm very new to this, and just beginning to dip into all the articles about SC cut versus AT cut, etc.

> If you're using NTP, why do you need to
> initially set it to GPS accuracy?

We need the GPS fix for location, and get the time for free, so figured we'd use it.

> Your best solution is to maintain the GPS antenna and only use
> GPS to discipline a good crystal oscillator.

That would be nice, of course, and we'll get the best GPS antenna we can afford and fit, but we can't have an externally-cabled antenna, or require that people put these on windowsills, or anything like that…  There will be too many of them, and the level of clue of the people plugging them in out in the field is likely to be too low.

> Do you plan to regulate the ambient and power environment to
> some degree of accuracy?

Yes…  They'll all be indoors, which helps, and temperature-regulated crystals seem to be relatively widely available, if in a dizzying number of styles.  Regulated power is something that we need to just have someone spec out, or use a reference design, if one exists.  Anyone have pointers?

> Note that 10 microseconds is 1 part in 10E5.
> The folks on this list deal in parts per 10E12.

So that's something I've been having a hard time understanding…  If that's the amount of inaccuracy _per oscillation_, then at the time-scales I'm dealing with, it would quickly accumulate and become unuseful…  that is, 10 microseconds of drift per second is almost a second of drift in a day, whereas I need, ideally, something I can discipline to within 100 microseconds _total_, using just a single GPS fix, plus NTP over the long haul.

Now, I don't know whether what I want is possible or not, but that's why I'm asking these questions.

Or am I misunderstanding the parts-per-foo notation?

> Can you use a standalone receiver to always generate a 10 MHz
> signal or pulse per second signal that is distributed to all
> of the measurement devices in a facility?

There will only be one device per location, and we can't have any external stuff plugged into them.  Else yes, I'd just use GPS and be done with it.


Version: GnuPG/MacGPG2 v2.0.17 (Darwin)
Comment: GPGTools - http://gpgtools.org


More information about the time-nuts mailing list