My hunch is that the circuit relies on the batteries being very local,
and the wire length to the power supply is causing instability. You
could test the hypothesis by putting batteries on as long a cord as the
power supply and see if you get similar results.

In that case, the fix would be to put a small bypass cap where the
batteries should be.

Cheerful regards,

Bob

On Wed, Jan 23, 2013, at 08:27 AM, William Bulley wrote:
> According to RussellMc <apptechnz@gmail.com> on Wed, 01/23/13 at 11:11:
> >=20
> > > This "enhanced" design has a serious failure mode - when it works at =
all,
> > > and that seems to be unpredictable - it "runs" at an accelerated rate=
..
> > >
> > > Sometimes it doesn't light up at all!   :-(
> > >
> > > How can this be?
> >=20
> > It seems very likely that you have higher voltage or spikes present.
> >=20
> > I'd expect that placing a large capacitor on the plug pack output would=
 help.
> >=20
> > Thought: Is plug pack definitely DC and not AC? (Unlikely) .
>=20
> Not higher voltage (measured using DC multimeter) - stable 4.5 VDC...
>=20
> Spikes - can't say, have yet to use oscilloscope on the power rails...
>=20
> I thought of the large capacitor, but have not done so.  Thanks.
>=20
> The wall wart or plug pack is stamped or labeled with 4.5 VDC... and
> was ordered as such from DigiKey.  It is definitely DC because:
>=20
>    1) the DC multimeter so confirmed
>=20
>    2) I specified 4.5 VDC when I ordered from DigiKey
>=20
>    3) the sticker/label on the wall wart or plug pack says so
>=20
> This is such a weird problem...  :-(
>=20
> What bugs me is this: clearly the logic circuitry must have some way
> (oscillator, bi- or multi-stable circuit, etc.) of "telling time", but
> why would 4.5 VDC from wall wart or plug pack be so different from
> 4.5 VDC from three AA cells - enough differenct so as to affect the
> operating frequency of said oscillator?
>=20
> If the display was powered from AC mains, I could see how a logic
> circuit could sense the zero crossings (for example) and use that
> as a (relatively stable across six hours) time base.  But since the
> display is powered (as designed) from three AA cells, there must be
> an internal time keeping device or circuit, yes?
>=20
> Or could they depend on the time constant of an RC circuit (with a
> very large value of C)?  Only tearing the tapestry open will give
> me the anwser to this question.   :-(
>=20
> Regards,
>=20
> web...
>=20
> --=20
> William Bulley                     Email: web@umich.edu
>=20
> 72 characters width template ----------------------------------------->|
> --=20
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist

--=20
http://www.fastmail.fm - One of many happy users:
  http://www.fastmail.fm/help/overview_quotes.html

--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.