PS/2 Interface (was: Wang 300 Calc]

Tony Duell ard at
Sat Sep 15 17:40:14 CDT 2007

> I dunno, Mike.  I'm generally capable of operating most any computer, 
> but my eyesight is terrible (and not getting any better with age).  
> I'd sure as heck would like the assurance that whatever plug I'm 
> sticking into a receptacle at least isn't going to result in "magic 

Magic smoke is generally released by : 

1) Overvoltage on one or more pins (e.g. connecting an RS232 signal to a 
TTL input)
2) Reversing power lines
3) Shorting TTL outputs together (although this may not do any damage if 
you're lucky)

Since the PS/2 keyboard and mouse interfaces are electrically very 
similar, I can see no reason why you'd let the smoke out if you pluged 
them into each otehrs' conenctors, unless oyu plug a PS/2 keyboard into 
one of thsoe seiral pouse adapaters and then into an RS232 port. But 
that's a PS/2 interface, it's an adapter for a mouse that can either be 
PS/2 or RS232. Plugging a true PS/2(only) mouse into such and adapter 
would do damage too.

I am suprised nobody has complained yet that IBM used a DB25 socket for a 
TTL-level parallel port. After all, plugging an RS232 cable in there will 
do a lot of damage. Or that HP used a 50 pin Microribbon connector for an 
RS232 port on some of their machines -- and evenbrought the sysem power 
lines (+5V, +12V, -12V) out on it. Connecting a SCSI device to that will 
not be pretty!

I guess it may come down to the unix philosophy in the end. 'We don't 
stop you doing stuping things becuase that would stop you doing clever 
thigns too'. There may very goof reasons for wanting to connect a device 
to an interface connecotr the manufacturers didn;t think of...

Of course no sane person uses mains connectors for anything but mains. 
Note I said 'sane'. There was at least one UK audio amplifier (I think it 
was a Pye) that used the same type connector -- a 2 pin Bulgin thing -- 
for both mains input and loudspeaker output. Swap them round, and you 
needed a new output transfdormer. I also once had to repair an infrared 
spectrometer where the manufacturers had used such a Bulgin mains 
connector for the analogue output signal. There was no other mains 
connector on the electronics box, so the operator had, not suprisingly, 
connected mains to that. In factm the electronics got its mains supply 
down a multi-core cable (also carrying motor control signals, etc) from 
the optical unit (which did have a mains cable). Needless to say applying 
mains ot the output let some smoke out (but suprisingly lttle -- I think 
one resistor was burnt out).

> smoke" (see my earlier comment on wall wart perversion) even if I 

Hmmm... I can see both sides to this. 

On the one hand, it's a very bad idea to use the wrong PSU ('wrong' in 
voltage, or polarity, or...). So having unique connectors that only fit 
the right device would seem to be beneficial.

But ther other side is that wall-wrts get sepearated from the devices 
the power. They may fail. And they're often mains-voltage-specific (a 
problem for Europeans wanting to use US stuff...). Some years ago, I 
bought one of those packs from Maplin of the 16 most common power 
connector 'tips' and a cable with a moudled socket on the end to take 
them. The other end of the cable I wired to a couple of 4mm plugs to fit 
my bench supply. THis has proved invaluable over the years for powering 
up units where the original wall-wart is missing or can't be used. 

So I can see the beenfit of a stanard connector too, one that I can match 
up with one from this kit.

Also, wall warts fail. Often the output cable fractures internally, close 
to the moulded connector. If you can't get the connector, you may be 
stuck trying to get a new wall-wart. Try that for a 20 year old unit 
sometime :-)

On the whole, I guess I'm in favour of standard connectors (2.1mm 
coaxial, 2.5mm coasial, maybe 2.5mm and 3.5mm jack plugs), but giving the 
power requerements and polarity on the unit iteslf somewhere (either on 
the nameplate label, or on the silkscreen of the PCB inside, or...). Not 
in the manual only, as that can get lost too.

> didn't quite read the label on the box correctly.
> Making this vintage--does anyone remember the competitions the Eagle 
> used to run, taking a bunch of people off the street, presenting each 
> with a boxed Eagle (III, I think) and timing them to see how quickly 
> they could set the thing up.  IIRC, the record was in the sub-one 
> minute range.

How pointless, consdiering how rarely most people need to set up a new 
computer, compared to doing other things with it. 

It's similar to the 'easy to learn' claims of a lot of products. I don't 
much care about that. What I want to know is whether, once I've leart to 
use it, will it save me time/effort/whatever. I am quite prepared to take 
time to learn something that will, and equally, won't waste any time at 
all on somethign that won't.

Incidetally, I was once timed to see how long it would take me to 
dismantle an ASR33 into the main sections (cover, base pan, typing unit, 
keyboard, punch, reader, call control unit) and put them back together 
again. I think I got it down to under 5 minutes in each case.


More information about the cctech mailing list