The main theme continues – it was tougher than the factories and I expected to meet IBM’s exacting standards and tolerances from 35 years ago.
Instead of delivering functional keyboards and meeting the original factory timeline (I’ve been typing on the fully working prototypes without issue this year), I want to make sure these keyboards live up to the quality standards of the originals for the clickiest, most musical typing experience possible, even at the cost of time (rejecting parts) until these standards are met. Eventually they have been met for the other critical parts, many of which finished production and are waiting in boxes at the factory for assembly (tens of thousands of barrels/flippers, the die cast zinc cases, capacitive PCBs, controllers, custom IDE/floppy style ribbon cables have finished production). Check out a few blog posts back for a detailed production update by part.
Dimensions on the keys that are off by just 0.5mm do not function well (I received the latest key samples this week). Spring free length variances are too large for a more consistent feel.
I cancelled the spring contract last month due to poor part tolerances and I expect to cancel the key mold contract as well and move it to another manufacturer this month. While this is going on I will have assembly start on the rest of the components to avoid any bottlenecking so we keep on schedule.
The springs are going to be made in the US. Almost none of the US-based spring factories had the equipment to manufacture springs to IBM’s 1980s tolerances and automatically discard all springs that do not meet the tolerances (at least one does – the one I am working with). Some US factories could only meet 4x IBM’s original tolerances!
The factory said that these springs’ tolerances are as tight as those of their aerospace customers.
I will be ordering sample springs this coming week for evaluation and then if they are good I will order all the springs from this factory – everyone will get the US-made springs.