MiniSport Laptop Hacker - Vol 14                25 May 1993
Copyright (c) 1993 Brian Mork

>>> ADMIN
Things have been pretty quiet this time around.  I got only one message
back from the field.  In the meantime, I managed to bring up an Internet
node in-house, and I'll start porting these newsletters over there, too.
Look in the micro.zenith and rec.radio.amateur.packet forums.  If you
read these MiniSport Laptop Hacker newsletters, please make it a practice
of kicking back at least a one line message indicating what parts are
particularly useful or not useful.

>>> POWER SUPPLY HARDWARE
Well, for those of you that have been following the MLH series, you'll
know that one of my power supplies has been giving trouble.  It finally
gave up the ghost.  Simultaneously I heard from another ham who had near-
ly identical symptoms:  The Power & NumLock LEDs flash on for about 1/2
second and then the computer power downs by itself.  From Internet, I
heard from one person who bought his for a real price, with a guarantee.
Its power supply died this way, also, and the seller replaced it.  He's
had no problems since.  Too many similar stories!  I wanted to know
what's going on!  In the meantime, I'm looking to buy broken MiniSports.

I tore mine open and started from the exterior working in.  There are a
handful of components that condition power and select where power is com-
ing from --the AC adapter or the battery.  This initial power supply sec-
tion is documented below.

My two ZL-x computers are different:

          Older                              Newer
     ---------------------------        -----------------------------
     * Power supply piggyback           * Power supply all on one card
       card with number PA2020P25         with number PA2020P65
     * Motherboard 1PC1606AMB-40I       * Motherboard 1PC1606AMB-41I,
       with green patch wire from         no patchwire from/to Test
       TP43                               Point 43.
     * U47 BIOS is "BIOS V3.3D"         * U47 BIOS is "444-804"
     * Disk controller WD37C65BJM       * Disk controller is same part
       has manufacture date in            number dated 44nd week in
       22nd week of 1989.                 1989.

The power conditioning sections are physically different, but electrical-
ly they're the same.  I expect my older one was a "fix-it" prototype that
later made it onto the production board.  Here's the schematic of the
piggyback section:
              +------------->|-----+
              |  ........     D1   |
              |  .      .          |
              |  .  \   .          |   +---~~~~----~~~~-----BJ
ylw "BA"------+----+ \----------+  |   |    L1      L2
(battery)        .      .       |  |   |
                 .      .       |  |   |
red "AD"------+------+---->|----+--+---+--------------------MIDL
(9v pwr)      |  .  _)  .   D2
              |  . (_   .
           D3 |  .  _)  .
              ^  . (_   .K1
              |  .  _)  .
              |  . (_   .
              |  .   )  .
blk "G"-------+------+--------------------------------------GND
(ground)         ........


 D1,D2 - two B8300s or integral N9K-60QC04        BJ named because the
 D3    - ceramic bubble (part no. unknown)        older power supply has
 L1    - approx 24 turns on 1/4" form             a Blue Jumper wire.
 L2    - 1 turn on core
 K1    - SPDT DIP relay, bottom    M    AD n/c    MIDL named because it
                          view     *    *  *      is the MIDDLE wire
                         pinout:   |              connecting the piggy-
                                   *    *  *      back board.
                                       GND BA
Operating from battery power, power comes in through D1 and out both the
BJ and MIDL connections.  When the 9v is available, current comes in
through D2, again going out BJ and MIDL.  My suspicion is that one of
these carries the main power current and one is a monitor tap, i.e., in-
dicates when AC sourced power (9v) is available.  Additionally, K1 pulls
the contact closed, providing current back to the battery pack for charg-
ing.  If you listen carefully when 9v is connected, you can hear K1 click
closed.  the glass diode (D2) provides a shorted path for the current
spike coming from K1 when it clicks off.

Notice that this circuit is in operation whether or not the computer is
turned "on."  It might be more appropriate to think of the computer as
always being on;  when you press the on button, the power supply just
"upgrades" to full output status.  The rest of the power supply (all
stuff between what I presented above and the header I described back in
issue ????) lives on an approximately 5"x4" circuit board.  Reverse engi-
neering of this board has been slowed by my inability to identify three
chips.  There are two of each of the following: MB3778 (16-pin IC), K612
(3-terminal + heat sink tab), and D4049G (16-pin IC).  PLEASE FORWARD ANY
INFORMATION YOU HAVE ON THESE THREE CHIPS.

73, Brian, ka9snf@wb7nnf.#spokn.wa or Internet bmork@opus-ovh.spk.wa.us