CDC CYBER 170 Consoles

[Tom Hunter]

I have been working on CDC CYBER 170 mainframes between 1977 and 1988. In
2002 I wrote an open-source emulator for the CDC 6000 and CYBER series (see
my website http://www.control-data.info/). In 2013 I also developed the
open-source VHDL firmware emulating the console controller for these
systems. The firmware runs on a Xilinx Virtex 6 FPGA on a PCI Express
(PCIe) board. The off-the shelf Xilinx board carries a small custom
"piggy-back" board with 4 DACs and 4 op-amps to interface to a DD60or CC545
console. This PCIe board was used by Paul Allen's Living Computer Museum
(LCM) in Seattle from 2013 onward in a PC running my 6000/CYBER series
emulator to drive a real DD60 console. The CC545 has a very similar
interface and my emulated controller would work with it too.

For many years I have been trying to find one of these vector drawn CC545
consoles to use with my emulator but I haven't been able to find one.
Recently I decided to build a clone of it myself. Bitsavers has a hardware
manual with schematics:


http://www.bitsavers.org/pdf/cdc/cyber/cyber_170/62952600L_CYBER_170_Display_Station_CC545-CDEF_Hardware_Reference_26Mar1979.pdf

The CC545 console achieved unusually fast deflection with an
electromagnetically deflected CRT. I am trying to understand the tricks
they used to get these high speeds. Part of the magic is a dual-yoke which
provided gross positioning within 2 microseconds to anywhere on the screen
using the first yoke (this is VERY fast) and then painted the character
using a second yoke around that base position with 0.1 microsecond per
stroke (this is VERY fast too). The two yokes work in an additive manner.

The reason there are two yokes is that you need quite low
inductivity/impedance to be able to drive the symbol vectors at 0.1
microsecond per vector with up to 24 vectors making up one character. The
gross position yoke needs to create a large enough magnetic field to sweep
across the entire screen so has a higher inductivity/impedance but the
magnetic field has 2 microseconds to stabilise.

The older DD60 console used electrostatic deflection which is much faster
by its very nature. Traditional CRT oscilloscopes were all
electrostatically deflected because of the speed advantage over
electromagnetic deflection.

The CC545 manual on Bitsavers has a good description of the circuits and
schematics, but unfortunately Section 8 with the "Parts Data" has not been
scanned. I really would like to know the types of transistors used in the 4
deflection amplifiers as well as the details of the dual-yoke and possibly
the CRT data.

It would also be very useful to see details of the design of the dual-yoke
and possibly the inductivity of each of the coils. This dual-yoke is most
unusual and very different from what is used in TVs, CRT monitors and even
vector drawn games like Vectrex or early vector drawn Atari arcade games.

Could somebody please help?

Thanks

Tom Hunter