Text 31, 63 rader
Skriven 2004-09-04 08:29:00 av Greg Mayman
     Kommentar till en text av William Kitchen
Ärende: UNKNOWN IC
==================
 -=> William Kitchen said to Greg Mayman
 -=> about "UNKNOWN IC" on 09-01-04  00:27.....

 WK> Well, I wish I could claim this little bit of cleverness, but alas,
 WK> it's  something I found online. I am, however, describing it in my own
 WK> words and  adding a small modification that makes it even simpler than
 WK> the original  version. It would be hard to come up with a simpler or
 WK> cheaper solution, and  though it may not be quite the sophisticated
 WK> electronics project you had in  mind, it appears to meet all of your
 WK> stated requirements. 

Completely offtopic! Moderate that fella, Mr Moderator!!!!

I have a feeling that I read somewhere that ice is an insulator.
If that is right the electrodes could be buried in the block.

Experiments needed...

Thanks, William. That is SOOOOO much simpler than what I had in
mind, and as you say, it could be calibrated empirically!

This is a typical example of lateral thinking.

I remember a similar case that ocurred in the 1950's when one of
our hospitals wanted a timer for their X-ray machines, to give
times up to 1 second in 1/10 sec increments. The people doing the
job worked for about 6 months to develop a system using a
thyratron (gas triode) timer, with several regaulated power
supplies, that took up to 20 minutes to warm up and stabilise.

But one of my friends took an ordinary rotary telephone dial
(remember them?) modified the cam to give a continuous closure
instead of a string of pulses, and showed that this would give
better results, more consistently, and with zero warm-up, than
the electronic monster.

My purely electronic idea for the freezer alarm was as follows:

An NTC thermistor, powered via a variable resistor (for temp
calibration), feeding the schmitt trigger input of a PIC
microcontroller. The micro would be in SLEEP mode for about 10
seconds, then as it powered up, it would pull one of its ports high
to apply power the the R+NTC string. After a millisecond or so,
the schmitt input would be tested to see whether it had gone high. If
it had, indicating that the temp was below the trigger point, the
unit would go immediately into sleep mode. Otherwise the alarm
would be sounded for say 500 mSec, then the unit would go into
SLEEP mode again.

This intermittent mode keeps the power drain way down.

Of course calibration of the unit would be very tedious...

   ,-./\
  /     \ From Greg Mayman, in beautiful Adelaide, South Australia
  \_,-*_/    "Queen City of The South"    34:55 S  138:36 E
       v

... I used to have an open mind but my brains kept falling out
___ Blue Wave/386 v2.30

--- FLAME v2.0/b
 * Origin: Braintap BBS Adelaide Oz, Internet UUCP +61-8-8239-0497 (3:800/449)