[time-nuts] cheap 5V OCXO in 14DIP has about 1E-9 drift per day

Tom Van Baak tvb at LeapSecond.com
Sat Apr 9 21:03:18 UTC 2011


A note on notation.

That ISO convention is useful for ohms because few of us have
keyboards or phones with a cute Greek omega symbol. In this
case a short-hand 1k8 avoids the need for a special character.
It also avoids a tiny decimal point, which is relevant if the label
is to be inked on a small component.

But time and frequency is a little different. We use s for seconds
and Hz for Hertz and keyboards support this. Thousands scaling
is typically done k, M, and G. Thus there is no need for special
fonts or resistor/capacitor-style ISO label formats.

Once in a while, in professional papers you will see the equivalent
of s superscript -1  in place of Hz. I presume this is done since a
unit of inverse seconds is redundant when you already have an
official SI unit of seconds. Would Heinrich Hertz mind?

For small units we use m (milli), u (micro), n (nano), and p (pico).
It is common in email and on this list to use the Roman u instead
of the proper Greek mu for the micro prefix. When I write for the
web I try to use µ instead of u for micro. But here in plain
text email you will typically see "us" for microseconds.

One nit to watch out for is that mHz and MHz differ by 9 decades.
For those of you who use "Word" or spell checkers watch out that
it doesn't change your MHz and Hz to mhz and hz.

As an aside, again especially in email, you will find many posters
use the computer programming inspired mantissa/exponent format
for large or small numbers. For example the mid-term stability of a
rubidium may be quoted at 2.5e-13 or the frequency drift rate of
a quartz oscillator is given as 5e-10/day. So far no one has had a
problem with this notation, though I'm not sure how "legal" it is in
ISO circles.

/tvb





More information about the Time-nuts_lists.febo.com mailing list