Monday we spent at the hotel, as much of Ghana observes the end of Ramadan. On Tuesday, myself, Aimee, Benjamin, and Edward Antwi journeyed to Suamee magazine with a hastily assembled machine #2 and extra coin mechanism. Our mission was to find, interview, and contract with local businessmen to produce the design in quantity. Benjamin's acquaintance with a local fabrication shop yielded us metalworkers, who were happy to assemble the sheet metal components. After some back of the blueprints math and amortization considerations, Edward arranged for the construction of up to six cases. We then traveled to another shop, this one a machinist, hoping to find someone capable of duplicating the Beaver coin mechanism. He was intrigued by the device, but agreed that any replication would be difficult if not impossible with subtraction techniques. While he offered sand casting as a possibility, we shared a concern that the accuracy afforded by sand casting simply would not suffice. This was confirmed later in the evening by phone after our machinist consulted other specialized sand casters.
Thursday, we returned to the shop for our final day of assembly. As we would be leaving the remaining components for seven additional machines with Edward and the universities, it was essential that we complete machines #2 and #3 to a state where they could be used as prototypes for a new generation of devices. To this end, we assembled and tested the two over the mid-week, discovering a new issue or two along the way -- Particularly, both machines had to be disassembled multiple times for issues with the coin mechanism, as well as issues with the axle mounting locations.
Beaver Coin Corporation holds a special place in my heart. While the ratcheting sound of the one-way gear is soothing and has grown melodic to my ears over the past year, I have a particular hatred for whoever designed the hole pattern which holds the device together. In short, two cast halves are mated about a rotating assembly, with four screws about the perimeter. If any of the screws are loose, the thing sits funky and eats coins or itself. Confused? here.
See, all four bolts thread into tapped threads in cast "sleeves," threading about 6 or 7 threads of contact. All except the Evil Bolt of Doom, which threads into a sleeve calmly only threaded at the bottom half. This leads to anywhere from 0 to about 4 threads of engagement, which meant that if the spacers were slightly too thick (which they were) or if the case was slightly bent (which also happens) this screw wouldn't catch, and often would happily a few miserable threads out of its sleeve and promptly strip the hole entirely. I quickly learned to install all the other screws first to appropriate torque, and then install the Evil Bolt of Doom to a much lower and cautious torque.
Somebody should really fix that, it's annoying and wasted hours of assembly.
We also encountered problems with vending in these two machines, which we addressed but didn't meet our standards set with the first machine. Most of the issues stemmed from a loose and inaccurate drilling of the mounting holes for the axle. With these holes off-location, the axle would sit at a noticeable angle to the inventory chutes, leading to starkly different results from each. To rectify this, we drilled the holes oversized but centered, and then created a "shim" of Gorilla Tape, which we added in a one-layer-thick wrap to the axle, to match the oversized diameter.
Our unresolved consistency concerns with machines #2 and #3 stem from their lack of consistent quantity of condoms vended. Neither machine jams, nor steals your money, but neither consistently vended a steady number from both chutes. We left the "proof to the readers," so to speak -- in hopes that rising to solve the problem will lead Edward's team to engineer better designs for these problem areas.
Friday's flight back was largely uneventful, though the tedium of airport security was a bit more than I'm used to -- the Accra airport leads one through three security checkpoints and at least as many passport inspections, finally dumping you in a securities--off glass boarding gate, where you sit quite helpless wondering what's on the departure signs behind you, and where exactly is this aircraft anyway, weren't we supposed to board a few minutes ago? Once in the air, however, we made excellent time and caught New York sleepy and unaware -- like a grumpy bear not quite awake -- and made our way quickly through the American side of the security show. By mid-morning, we were all safely headed back to Albany.
