[time-nuts] Frequency Stability of Trimble Mini-T
SAIDJACK at aol.com
SAIDJACK at aol.com
Thu Oct 16 23:57:11 UTC 2008
Hi Hal,
>First, we understand metastability. We can measure it and predict it. It
it
>mattered, we could test each individual part.
That was exactly my original point. The discussion got off topic, I just
wanted to point out that it is a matter of probability, and the math to
calculate this probability of failure is well known and understood.
The math says that you cannot prevent a metastable event, even if there are
1000 flip flops. The probability is never zero. And since the processes are
stochastic, the first failure could happen within 2 seconds of power on, even
with 10000, or a million cascaded flip flops. You can never prevent a
metastable event going through all of the flops. You can just make that probability
approach zero. Actually, if you had a million cascaded flip flops one of them
would probably fail due to other reliability reasons :)
>Second, the failure mode is exponential in a parameter we can control. So
>given a particular set of parts to pick from, it's reasonable to make a
>design with a probability of error small enough so that other things are
much
>more important.
Right! My original point again. The question was: how often do crystal jumps
happen. How often do metastable events happen. Both may be just a matter of
statistics and probability.
If we don't see a crystal jump within say one month of operation, we can
probably say the probability of one happening is extremely low.
But I have seen jumps happen after 3 months of continuous, documented
operation without any jumps, so again we may never have certainty that
jumps/metastable events will not happen. We just have to make them happen so seldomly
that they won't matter to the application at hand. But again if it jumps after 3
months, no one can guarantee it won't jump after 10 minutes.
bye,
Said
____________________________________
From: SAIDJACK
To: SAIDJACK
Sent: 10/16/2008 16:43:33 Pacific Daylight Time
Subj: Fwd: [time-nuts] Frequency Stability of Trimble Mini-T
>From iPhone
Begin forwarded message:
From: Hal Murray <_hmurray at megapathdsl.net_ (mailto:hmurray at megapathdsl.net) >
Date: October 16, 2008 15:40:12 PDT
To: Discussion of precise time and frequency measurement
<_time-nuts at febo.com_ (mailto:time-nuts at febo.com) >
Subject: Re: [time-nuts] Frequency Stability of Trimble Mini-T
Reply-To: Discussion of precise time and frequency measurement
<_time-nuts at febo.com_ (mailto:time-nuts at febo.com) >
We have done some extensive research trying to correlate jumps to
external phenomena (radiation, vibration, thermal effects, etc).
There is no straight forward correlation.
It's similar to asynchronous switching inside a digital computer.
You can add levels of flip flops to synchronize across two
asynchronous time-domains, ...
I think that's misleading in two ways.
First, we understand metastability. We can measure it and predict it. It
it
mattered, we could test each individual part.
Second, the failure mode is exponential in a parameter we can control. So
given a particular set of parts to pick from, it's reasonable to make a
design with a probability of error small enough so that other things are
much
more important.
If it absolutely, positively can't take any hit, then some more work is
involved
I would say that the first step is to put a number on "absolutely,
positively".
If you aren't willing to take some risk, you won't get off the drawing
board.
--
These are my opinions, not necessarily my employer's. I hate spam.
_______________________________________________
time-nuts mailing list -- _time-nuts at febo.com_ (mailto:time-nuts at febo.com)
To unsubscribe, go to
_https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts_ (https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts)
and follow the instructions there.
=
**************New MapQuest Local shows what's happening at your destination.
Dining, Movies, Events, News & more. Try it out!
(http://local.mapquest.com/?ncid=emlcntnew00000001)
More information about the Time-nuts_lists.febo.com
mailing list