slick bits in computers (WAS: VAXen Rule!)

[Tony Duell]

> 
>  >   Not sure I'd agree on the reliability...I've spilled my fair share of
> > coke into keyboards, and contact-closure keyswitches just don't cut it
> > for me unless they're sealed.
> 
> You can make switches that are cheap and nasty, and you can make
> switches that will last 100 years - and everything in between. If you
> want a fair comparison, match the HP keyboard against a quality
> "mechanical switch" keyboard, not a Chinese piece of crap. Lets face

For some reason, the HP9830 did not use this trasnformer keyboard, it 
used a matrix of (IIRC Cheey brand) keyswitches. And yes, I have had to 
replace _one_ of those on my HP9830 (which is 33 years old, so I am not 
really complaining).

I do wonder, though, if the transformer keyboard would be that much more 
expensive to produce. The encoder/controller board was about 20 TTL chips 
then, now it could be triviailly fitted into a PLD or ASIC (if you did 
the latter, you could fit the comparator in as well). The keyboard 
'transformer' PCB is just a double-sided PCB, easy to make now, and the 
only components on that are a diode for each pair of keys (compared wit 
ha diode per key for a decent mechncial keyboard)

As regards the mechanical bits, there are 4 for each key : 

The keycap (same as for a mechancial switch, basically)

A plunger (plastic moulding) with a metal disk heat-staked to the end

A spring

A mounlded plastic housing.

No worse than the parts for a mechanical switch, and shouldn't be any 
more expensive if produced in quantity.

> it, HP stuff of that era was pretty much the pinnacle of engineering.

One reason why I like their _old_ machines.

-tony