dplatt@coherent.com (Dave Platt) (01/24/89)
I had an interesting experience this weekend which throws some light on
some of the circuitry in the Mac II, and on a possible design glitch in
the Kensington Turbo Mouse ADB.
I arrived home late Saturday night, after attending the MacWorld Expo
(sore feet!) and the Netters' Dinner (excellent food!). My wife told
me, with some apprehension, that our Mac II had apparently turned itself
on while she was out shopping, and that she was unable to shut it down!
She had tried the "Shut down" menu item, and the Mac had "given me a box
I've never seen before"; she had tried the power-off button on the back
of the Mac, and the machine had refused to power down.
I checked, and sure enough something strange was definitely occurring.
When I selected "Shut down", the "You may now turn your Mac off safely"
dialog box appeared... quite normal for a Mac SE, but _not_ for a II!
When I pushed the power-off button, the machine shut off and immediately
powered itself back on!
I started fiddling with the ADB cables and devices, and soon determined
that the machine would not shut off if my Kensington Turbo Mouse was
plugged in (either directly or via a daisy-chain). Plugging this device
into the ADB would cause the Mac to power itself up, and to refuse to
power down.
The next day, I brought my other trackball (same model) home from work,
and determined that it was behaving correctly... the problem was clearly
located in the offending trackball rather than in the cables or the Mac
II itself. I popped the case off of the defective trackball, removed
the PC board, and began investigating.
It became clear quite quickly that there was _no_ way that the active
circuitry in the trackball could be jamming the "power up" line on the
ADB... the line in question isn't connected to anything other than the
two ADB sockets, via a single trace on the PC board.
I began looking for a short circuit in the plugs and the PC-board traces
in question. My DVM said that no short-circuit existed; the "power me
up" trace had no connection to ground or to any other ADB line.
At this point, I plugged the trackball circuitry back into the ADB...
and, lo and behold, the Mac did not spring to life! I snarled a bit...
intermittent problems are on my list of least-favorite phenomena. I
began wiggling the plug and socket around with my fingers... and, when
my fingers pressed the socket's PC-board solder connections in the right
way, the Mac booted up spontaneously.
After fiddling around a bit, I determined that pressing my fingers
between the "boot me" and "shield" traces would supply a _very_ weak
connection between these points; I measured it at between 5 and 10
megohms. Apparently, even this very high resistance lets enough current
flow to trigger the "Power-on button has been pushed" detector.
I examined the tiny bit of PC-board to which the socket was connected.
The "ground" and "power me up" traces come within about 1/2 millimeter
at the point at which a short length of flat-wire cable connects the
socket to the main PC board. Bridging this gap with one finger, and
pressing lightly, was sufficient to cause the Mac to power up.
Our Mac lives in our back den... a room that gets quite cold at night.
The weather here has been cool and foggy during the last week or so.
When we opened the den up on Saturday morning, warmer and moister air
from the main portion of our house may have caused a small amount of
condensation to occur inside the trackball casing... and, apparently,
this condensation supplied enough of a current-path through the "power
me up!" line to cause the Mac to boot, and to refuse to shut down later
in the day. In effect, the moisture was "leaning on the power-up button".
I mixed up a dab of nonconductive 5-minute epoxy, coated the affected
PC-board pins and traces, let it harden, and sealed the trackball back
up. End of problem.
I now understand why electronic-equipment manufacturers often spray
their PC boards with a silicone-resin or urethane "conformal coating" to
seal out moisture. Unfortunately, Kensington didn't do so in this case.
--
Dave Platt FIDONET: Dave Platt on 1:204/444 VOICE: (415) 493-8805
UUCP: ...!{ames,sun,uunet}!coherent!dplatt DOMAIN: dplatt@coherent.com
INTERNET: coherent!dplatt@ames.arpa, ...@sun.com, ...@uunet.uu.net
USNAIL: Coherent Thought Inc. 3350 West Bayshore #205 Palo Alto CA 94303