GREAT Comment!

=D6mer Yalh=FD wrote:

> I have a board with 16F628A running with internal osc. Also, a 32KHz xt=
al is
> connected for basic time keeping.  It sleeps all the time and wakes up =
about
> every 500ms and does some stuff for short time then sleeps again.  Duri=
ng
> the awake time I turn on a led and turn it off just before going to sle=
ep.
>=20
> I have discovered that the chip does not wake up from sleep (the led do=
es
> not blink) for a while (4 seconds to 20 seconds or so) under 2 seperate
> conditions:
>=20
> 1. If I touch (with my bare finger) to the cap that is connected to the=
 RB7
> pin of the pic (where the xtal's one pin is connected) the unit does no=
t
> wake up
> Or
>=20
> 2. If I "breathe" on the same cap above.

I'll bet that you don't have a good GND return path for the GND of the=20
crystal capacitors. You have to be very careful about layouts. Your=20
finger offers a different (better?) gnd that the circuit offers.

But, having said that, it is also an issue when moisture gets into a
system. The impedance is so high, that almost anything can affect it.

>=20
> Note that I do not know the exact word for what I am doing so I said
> breathing.  Let's say that you want to see if your breath smeels, and y=
ou
> "breathe" to your hand and smeel it.  That is what I do to the cap (exc=
ept I
> don't smell it afterwards).  When I did this, I was actually trying the=
 get
> something off of the pcb, and was not testing my breath.  I also did th=
is
> test right after brushing so probably this is not because of my breath!=
 :-)
>=20
> I have put this board to a test in a fridge with -16 deg C and +85 deg =
C
> (+/- 5 deg C).  And the board still works.  Also, this behaviour has be=
en
> observed with many of the same boards and many of different kind of boa=
rd.
>=20
> The cap is 22pF.

Theoretically 22pF should be OK, but I always make it 33pF; it's slower=20
to start up, but not as noise sensitive.

>=20
> Why does this happen?  Bad board design?

Sometimes, i.e. the GND issue above.

> Has anybody else experienced this before?

Yes. It was a layout problem (the GND return of the crystal caps ran=20
around the board before it found its way to the PIC GND pin.

> How would I know this wouldn't happen after the unit is shipped?
> Is this a humidity problem?  Would this problem occur again after it ra=
ins
> on the unit (in IP68 box) and then there is sun right after the rain?

Yes, in some cases I spray the oscillator section of the board with a=20
protective plastic coating. NOTE: make sure the can is grounded before=20
you spray.


> P.s. The unit will be in an IP68 enclosure and the problem occurred wit=
hout
> the enclosure, no way to test it inside the enclosure.

You might want to consider a sealed case, with a packet of moisture=20
absorber. I like the clay type best, very easy to recycle.

> I am sorry for the lengthy message but I am really curious as to what i=
s
> going on here and to find out if I can take any precautions.

--Bob
--=20
              --------------
                Bob Axtell
        PIC Hardware & Firmware Dev
          http://beam.to/baxtell
              1-520-219-2363

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body