On 05/24/2016 05:44 PM, Chuck Guzis wrote:
On 05/24/2016 02:21 PM, Paul Berger wrote:
The CROS cards used in a 360/30 where the same size as an 80 column
card on purpose so you could you a keypunch machine to program the
microcode.
But I believe that the CROS cards were mylar, no?
On the 360/30, there were 12 word lines printed in silver
ink on one side of a .003" thick Mylar card, and bit line
boards, with traces running at right angles to the word
lines. There were 60 bit lines per board/card. The word
lines had 60 little "pads" hanging below them, and these
were at the exact location of the holes in standard IBM
punch cards. So, if you punched out the pad, one of the
capacitor plates would be missing. If the pad was not
punched out, then it formed a capacitor between the word
line and the bit line. Air bladders applied even pressure
to the stack of mylar card and bit line board. With only 12
words/card, it took quite a lot of them to hold the full
microcode.
On the 360/50 and /65, the data pattern was etched into a
bunch of wiggly traces. For each word, there was a driven
line and a balance line. If the driven line was wide across
from the 1's bit line pad, you got a 1 in the control store
bit. if the non-driven line was wide across from the 1's
bit line pad, you got a zero.
The mylar sheet was not punched, so changing the microcode
required replacing a whole bunch of etched circuit boards.
Jon
