25 Aug
2015
25 Aug
'15
12:06 p.m.
I'll admit to being surprised at the cheapness of the keyboard. Here I
expected to find a true mechanical type, like the early IBM machines - what
I found is a step-down from the C-64.
Depends one what you mean by 'early IBM machine'. The well-known type M
used on the 5170 PC/AT [1] is of this sort of design
[1] Yes I am well aware that the 5170 is not an early IBM machine in terms of IBMs
total computer production. But a lot of people now think it is ancient.
The switching side is what I'd term a 'membrane sandwich. There are three
layers in the stack, with upper & lower layers forming the switch contact
pairs. Sandwiched in the middle is a perforated insulator sheet, with a
hole pattern that matches the switch contact pattern. Pressing on the right
area of the stack unites the upper & lower contacts, temporarily closing
the switch.
Another well-known keyboard with this type of design is the DEC LK201 (as used on
the VT220, etc).
I have taken both LK201s and IBM type Ms totally apart. The latter I managed to
reassemble with screws. The former I never managed to get to work properly again,
the plastic is just too thin to take screws, there is not enough to re-melt and no glue
seemed to hold.
So I suppose if a spilled liquid infiltrated the
stack, perhaps drying and
leaving a residue, it may have short-circuited any number of contact pads.
But IIRC, I was measuring +very+ low resistance in the shorted pins. I'm
not sure how even dried Coca-Cola could create that hard of a short.. so I
need to take some more careful readings. If they are truly hard-shorted,
I'd tend to look for something other than a contamination issue.
What bothers me about this design it that it relies on the the springiness of the
plastic to open the contacts again. This seems like something that might fail with
time so the sheets remain in contact.
-tony