3 Feb
2023
3 Feb
'23
4:14 p.m.
On Fri, Feb 3, 2023 at 2:41 PM Will Cooke via cctalk
<cctalk(a)classiccmp.org> wrote:
Yes. The thresholds shift with temperature. Some
companines (DEC?) used temperature-compensated amplifiers to address the issue. IBM, on
some models like the 1620, kept the cores at a constant (elevated) temperature.
Most of the core memories that I've worked on, including quite small
ones like the HP9100B, have a thermistor physically next to the core
array to control the threshold.
I also seem to remember there were MAINDECs for various PDP8 and PDP!1
core units that did the worst case sequence for heating the cores and
checked the memory was sill operational.
-tony