Searching \ for 'DS32kHz TCXO' in subject line. ()
Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=ds32khz+tcxo
Search entire site for: 'DS32kHz TCXO'.

Truncated match.
'DS32kHz TCXO'
1999\01\04@230410 by

PICsters,

I was sampled four of the new Dallas Semi DS32kHz parts just before
Christmas.  These are temperature-compensated 32,768Hz oscillators  with
150uA nominal operation and 1uA battery back-up (uncompensated) operation.

True to my word, I tested them a little and thought y'all would like to
know the results.

My first challenge was the fact that they were in a Ball Grid Array, thus
I did some micro-soldering to be able to prototype the circuits.

My test was *very* primitive, but I think telling.

I set up the four oscillators:
A)  +5V at 72 degrees F
B)  +4.5V at 72 degrees F
C)  +5V at 40 degrees F
D)  +5V at 120 degrees F

I gated all the outputs through dual-in AND gates and into 32-bit counters
(they would roll over, but it was the difference I was looking for).  I
held one input LOW on each AND gate and then cleared all the counters.
Then I raised the AND gates with the same switch and let them all start
counting.

288 hours later, I stopped all the counters and read them.  The difference
was 128,918 counts from the highest (D) to the lowest (C).

At first I was not very impressed.  I calculated that this represented
447.632 counts an hour.  And this would mean:

447.632 * 24 * 365 = 3,921,256 counts/year, which still didn't impress me
until I realized:

3,921,256 / 32,768 = 119.67 seconds per year!  Two minutes a year in a
data-logger clock is something I can live with!  And if it happens that
one was fast and the other slow, then they are within 60 seconds a year!!

The compensation circuitry works a treat, methinks.  Now if the things are
also accurate.  I will be testing them against the clock at WWV in
Colorado as soon as I figure how to trigger off of the radio signal (the
NET?).  Any ideas would be appreciated.

The data sheet can be found at:

http://www.dalsemi.com/DocControl/PDFs/32khz.pdf

If its accuracy is as good as its compensation then it's a winner, I
think.

BTW, (A) and (B) above were only off 1,407 counts.  Amazing.

Michael

**********************************************
Outside of a dog, a book is man's best friend.
Inside of a dog, it's too hard to read anyway!
Groucho Marx
**********************************************

In a message dated 99-01-04 23:12:41 EST, you write:

<< I was sampled four of the new Dallas Semi DS32kHz parts just before
Christmas.  These are temperature-compensated 32,768Hz oscillators  with
150uA nominal operation and 1uA battery back-up (uncompensated) operation.

True to my word, I tested them a little and thought y'all would like to
know the results. >>
Thanks, Michael, for this report.  It is this kind of information that makes
these discussion groups so worth while.

Max

More... (looser matching)
- Last day of these posts
- In 1999 , 2000 only
- Today
- New search...