Re: 4400/200 won't turn on?

[Mouse <mouse%Rodents-Montreal.ORG@localhost>]

>> [...4400/200 that won't turn on...]

[Martin Husemann]
> Failing PSU would be my bet.  Often can be repaired quite cheaply (or
> even DIY, but high voltage parts, not a first time electronics
> project).

It would not be a first-time project; I've been playing with
electronics since...oh, since I was a kid, probably since I first had
the dexterity to handle the tools.  Certainly decades.

But see below.

[Romain Dolbeau]
> IIRC, the 4400 has an ATX power supply or some other PC-originated
> standard.

The voltages are.  (The PS markings say +5, +3.3, +12, -5, -12, and +5
standby.)  The connectors...may be.

It has four six-pin connectors that remind me of peecee power supplies
of the generation before ATX - I haven't unplugged them because I'd
rather not risk dirtying currently-good connections.  It also has a
small three-pin connector, light-duty enough that it's obviously
carrying very little power, marked "SOFT PWR" in the silkscreening;
this one carries ground, standby +5V (max 20mA, according to the PS
markings), and a third wire which I conjecture is power-on control.

> Some Macs also had on-board batteries for various functions including
> keyboard-based booting, but my memory is fuzzy.

[Martin Husemann again, different message]
> Oh yes, empty battery would do it too, definitively worth to try
> first.

[Michael]
> Check the onboard battery.
> Older macs are notorious for not powering up with a dead battery.

My port-macppc subscription seems to have gone away - or perhaps I
unsubbed long enough ago I don't recall doing it.  I saw none of the
above on-list mail.  I've got a resubscribe in progress; in the
meantime, I'm replying by scraping the port-macppc archive.

I also got an off-list mail also recommending I check the battery
(thank you, you know who you are!).

It took a little doing, but I eventually found the battery.  In the
case of my machine, it's a small black plastic near-cube, connected to
two pins of a four-pin connector (which has only three pins actually
populated).

The battery was marked as being nominally 4.5V.  After removing it
from the board, it measured somewhere between 1.4 and 1.5 volts (with
my Fluke, a high-impedance voltmeter), so, quite dead.  Its markings
gave me no reason to think it rechargable.

I took a small external power supply driving 5V and connected it up to
those pins, with a 1K series resistor because I'm paranoid.  Then the
machine proved to be willing to turn on just fine, so the battery
theory appears to be the winner.

I'm not sure what do here.  The negative battery connector pin has
sub-1-ohm resistance to power supply ground, so I could drive the
positive battery pin from the power supply's standby +5V line (probably
with a capacitor to support brief power-off periods and resistors to
avoid overloading the 20mA max).  But that would not survive long
periods unplugged.

In the immediate term, I'll probably go with the capacitor and
resistors.  If I were confident I could find one easily, I'd get a
small battery holder designed for three AAA cells or some such and use
that.  There's plenty of physical space to mount such a thing.

Anyhow, at least now I can get it to power up. :-)  Thank you very
much, all of you!

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse%rodents-montreal.org@localhost
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B