Re: Disk configuration

On 1/8/2015 8:06 AM, Jim Oberholtzer wrote:

I asked one of the engineers in Rochester why there were so many screws (I
forget the exact count but it was > 25) on the side panel. That panel had to
come off to get after any of the cards installed on the system. His
response was electromagnetic emissions. The system was putting out too
much. So they started putting in screws until it got low enough to be
legal. It was a very unscientific way of solving a potentially serious
issue.

I can't speak to what a Rochester engineer said, but the technique is
very scientific. The distance between the screws should be on the order
of less than .10 wavelength. From a manufacturing cost perspective
that's probably too many drilling / tapping / fastening operations so
they reduce the count until they are just within FCC rules.

I solder the entire seam for my 3GHz transverters (wavelength approx
9cm) rather than screw the top on. Tapping aluminium for 4/0 screws
isn't my favourite chore; double sided PCB material works just fine as
an RFI shield. But when I need to work on the internals, it's a pain to
desolder. Which is why IBM use screws... :-)