An experiment in replacing the tired foam in an IBM Model F Bigfoot.
So far I have learnt from various forum posts here on DT that there are two sorts of foam in Model F keyboards. This foam is supposed to press the barrels against the back iron plate and circuit board, and sits just behind the perforated front iron plate. I have two Bigfoots, each with one or other variety of original foam.
My first Bigfoot is older, and has Ensolite foam, a sort of PVC material that appears to be more resistant to ageing than the later variety.
The second Bigfoot has more flimsy foam, with a thin black vinyl backing. This stuff does not appear to age quite so well as Ensolite. The black vinyl feels a bit tacky, and the foam has lost most of its springiness.
I searched in vain for a fresh supply of Ensolite. I decided instead to use closed-cell polyethylene foam, three millimetres thick. I've seen this foam used in floating buoys, and it is quite resilient stuff: if it can survive years at sea, it should last for quite a while inside a keyboard. It is easy to work with too, cuts easily with sharp tools.
I bought a pair of hole punches, 12mm and two millimetre diameter, and far more foam than I really needed, and set to work. I honed the hole punches to a much keener edge, so that they cut the foam with a gentle twist of the punch. No need for hammering.
Using the original sheet of foam as a template, the holes are cut with a gentle twist of the hole punch tool.
The corners of rectangular shapes are first punched through with the 2mm tool.
All cuts complete. The slits along the top and bottom allow the fastening tabs on the front iron plate to pass through to the rear iron plate.
The original and the new foam, side by side.
Inserting the barrels. The fit was appropriately snug.
The space-bar is tricky, the flipper has to be held down in place because the space-bar must be hooked up to the stabiliser wire from inside. It's not possible to hook up the stabiliser wire from the outside, after attaching the front and back iron plates together. With the space-bar in position, the spring is compressed with a loop of floss attached to a suitably heavy spanner.
The two iron plates are pressed together with four clamps, one at each corner of the plates, with wood dowels spreading the pressure and protecting the chromate finish of the plates. These need to be left to press for a few hours, so that the foam compresses, otherwise trying to slide the plates against each other to re-engage the hooked tabs is too difficult, and could risk accidental slippage and damage.
Sorry, I didn't manage to take photos of the manoeuvre. I'll try to illustrate this with another edit later.
The conclusion: physically, the foam seems to hold the barrels in place suitably well, they don't wiggle. Electrically, I can't tell if the foam has the right properties: I didn't test with the original foam, an oversight that I shall keep in mind for future experiments. The result is a keyboard that works, but appears to have de-bouncing problems:
Bigfoot with polyethylene replacement foam:
Code: Select all
thhe quick brown fox junmps over the lazy dogg the quick brown fox jjumps over thhe lazy ddog the quick brown fox jumps over the lazy dog thhe quick brownn fox jjumps over thhe lazy dog thhe quick brownn fox jumps over thhe lazy ddog thhe quicck brownn fox jumps over the lazy dog the quick brown fox jumps over the lazy dog
Code: Select all
the quick brow`n fox jumps over the lazy dog the quick brown fox jumps over the lazy dog the quick brown fox jumps over the lazy dog the quick brown fox jumps over the lazy dog the quick brown fox jumps over the lazy dog the quick brown fox jumps over the lazy dog
I wonder if the polyethylene is prone to static? Or is the pressure on the barrels too great? Or is the scan rate in Soarer's Controller too fast? The original cable may be a source of interference. I will try attaching the arduino to the logic board with much shorter wires, to see if that improves the spurious back-tick and bounced keys.