80-Bus News

Having got the plates, I got out the steel rule, pop punch and electric drill and started work. I had decided where the plugs for keyboard, serial I/O, mains, video and other connectors were to go on the end plates, so these were painstaking chopped out, the smaller ones using a selection of Q-Max punches, the larger ones drilled and then opened up with a metal fret saw and filed to shape. One end plate would carry a 5″ Muffin fan to keep the whole lot cool whilst the other end plate had all the holes for the sockets and the mountings for the power supply. The fan presented a problem. I could not find any metal mesh to cover the fan opening, and leaving it open would leak RF like a sieve, so I marked out about 500 3 mm holes over the intake area of the fan and spent an entertaining evening drilling them all out. As the box in its finished form would be almost airtight, I had to drill a similar number of ventilation holes in the other end plate to exhaust the air sucked in by the fan. The intake and exhaust holes for the fan had to be fairly small, as I didn’t want to take any chance of RF leaks. Large holes let RF out!! Making all these holes neat and tidy was no easy task.

With the two finished end plates, I started assembling the box. The two side plates were fitted to the base plate in turn, held in place with engineers clamps and a row of holes spaced at 1″ was first piloted and then drilled out to 3.3 mm through the side plates into the base plate flanges. The two side plates were then pop rivetted to the base plate. Similarly, the two end plates were clamped to the now fitted side plates and drilled and rivetted in a like manner. This made a nice substantial box without a lid, it was even reasonably square. The box went back to Smiths and a tight fitting flanged lid was then made to fit. The whole lot cost about a tenner, but I dare say that was a special price to me as they owed me a couple of favours. Anyway the box was a lot cheaper than buying a ready made frame and box.

The seven slot mother board was made from 8″ wide Veroboard and the positions of the mounting holes marked and drilled on the base plate. The lid was drilled for a number of PK self tapping screws. The time came for a dummy assembly and to see if I had forgotten any holes. The whole lot seemed to fit together, the fan, the PSU, the mother board, the sockets and all, so the whole thing was taken to pieces again. The outside of the box was then painted in dark blue Hammerite paint making sure to mask things like earth points, etc.

The fateful day arrived, the whole lot was reassembled in the box, wired up and switched on for the first time. One thing was immediately apparrent, the fan make an awful racket and the amount of air rushing through the box was overkill in the extreme, more like a wind tunnel than a computer box. On the basis that large diameter slow moving fans are quiet, some way of slowing the fan down had to be found. Simple, a 0.33uF 600V paper capacitor in series with the fan dropped the supply to about 90V, the fan ran a lot slower and was very much quieter. Was the cooling now adequate? Having left the thing on for several hours with all the cards in revealed a temerature rise of all of 10 degrees C so I guessed it was all Ok.

Did it cure the RFI? Well no, not entirely. There was still a lot of keyboard polling noise which inexplicably did not go away when I detached the keyboard. It went away when I detached the printer!! So the printer cable was replaced with a screened cable, leaving the keyboard ribbon cable as before. Still not quite right, the interference shot up when the disks started up so