OT: solution to PC systems problem??

[Debby and Gene Stopp]

Watch out if you're using Win95 - any interruption in the mouse connection
will cause 95 to loose it and never find it again, requiring you to tab your
way over to shutdown and starting over. NT doesn't do this - apparently it
polls for the mouse and if it gets unplugged and then plugged in later,
cursor control will resume after about ten seconds. NT is quite a tolerant
OS. 98 I haven't tried, even though I'm staring at it right now...

The missing keyboard at boot time is easy to get around - if there isn't a
"halt on no errors" option in the BIOS, then there's probably a "quick boot"
option in the BIOS (the latter on many of the newer AGP 100-meg Super 7 ATX
mainboards). Once booted, both 95 and NT will tolerate a disconnected
keyboard that gets reconnected later, no problem.

I've had good luck with the Linksys active switch, which works seamlessly
with both NT and 95, and uses momentary front panel switches that you could
probably easily remote. One drawback is cost, about $200 for a 4-port
keyboard/mouse/monitor switch. But it's cheaper than 4 monitors and takes up
less room. And last but not least, it's not nearly as fun as hacking up a
bunch of old keyboards!

- Gene


-----Original Message-----
From: The Dark force of dance <batzman at all-electric.com>
To: 'DIY' <synth-diy at mailhost.bpa.nl>
Date: Friday, August 28, 1998 9:55 PM
Subject: RE:OT: solution to PC systems problem??


>Y-ellow Y'all.
> And thanks to the many suggestions. I've managed to find a great deal of
>the information I needed.
>
>To all that suggested those commercially available switch boxes. Thanks but
>they aren't a very elegant solution in this case. There's nowhere to put
>one and they aren't remote-controllable. They also generally don't deal
>with multiple formats. IE: PS/2
>
>To sum up. (In case anyone else needs to do this kinda thing)
>The keyboard is relatively easy to do. An old keyboard on the end of a
>cross over switch. In this case a set of relay contacts. Just the board
>from the keyboard module will do. You don't actually need the keys. As long
>as both it (the dummy) and the real keyboard are powered up, it's possible
>to swap between them without the PC noticing. Just as long as you're not
>sending data at the time you switch over. This shouldn't be too much of a
>problem.
>
>So when both machines boot, they will see a keyboard on the end of the
>system and receive the necessary ACKs back from it. After that it doesn't
>much matter. There may be a problem with the status LED information but in
>this case it's not a problem because the master keyboard (as it were) has
>no LEDs. It's an IR keyboard. (Thanks Philip) :)
>
>The mouse is another problem all together. I'm not entirely sure if the
>same approach will work. However a Serial mouse and a PS/2 mouse are
>entirely different entities. What I realized I can in fact do is use a
>serial mouse. I don't _have_to_ use a PS/2 Mouse. The PS/2 mouse is
>virtually hooked directly across a UART. It's powered by 5 volts and not 12
>as in a serial mouse. It's data signal is also 5 volts as opposed to +/-12
>in a serial mouse. And of course that means it's signal is positive logic
>instead of being inverted. In short, it would be a really big job to
>convert the PS/2 Mouse up to Serial. Not impossible but impractical. But as
>I said, I shouldn't have to.
>
>What I still don't know is from where the serial mouse actually takes it's
>power. I assume that what it does is get power from one of the signals on
>the serial port. Since an RS232 port supplys no power as such. One or two
>of the lines must be taken high/low and the mouse get's it's power from
>that. The mouse being CMOS should draw only a small amount of current and
>can get away with it. But I can't work out how this is done exactly.
>Obviously time for some raw science here to find out.
>
>Anyway, thanks for all the pointers. I appreciate it.
>