MCSC@YaleVM.YCC.Yale.Edu (03/30/89)
A classroom of PS/2 model 50's has been plagued with erratic but persistent problem with clocks losing time -- i.e. several hours to about a day, apparently when they are powered off overnight or longer. It's not consistent but reoccurs frequently ( more than 1/week.) One, 2 or most machines may show the slippage each morning. Batteries have been exchanged, DOS3.3 ( and patch) reinstalled, to no avail. Very little software installed or being used; no memory resident programs, terminal emulators, or LAN. Any suggestions would be appreciated: even hearing that other have experience the same problem would be a relief. Richard S. Crane BITNET: CRANER@YALEVM
tchu@bbn.com (Tom Chu) (03/30/89)
In article <359@YaleVM.YCC.Yale.Edu> MCSC@YaleVM.YCC.Yale.Edu writes: >A classroom of PS/2 model 50's has been plagued with erratic but persistent >problem with clocks losing time -- i.e. several hours to about a day, >apparently when they are powered off overnight or longer. It's not consistent >but reoccurs frequently ( more than 1/week.) One, 2 or most machines may show >the slippage each morning. >Batteries have been exchanged, DOS3.3 ( and patch) reinstalled, to no avail. > > (rest of message deleted) I don't know if this has been mentioned or not, but the entire line of PC-ATs and PS/2s are notoriously bad for not keeping the time correctly. From what I remember, this problem occurs randomly and differs from machine to machine. When I was involved with the design of PCs, we did a test and set every clock on various PCs to a time. We recorded the time and found that different machines had big differences in the time. The best clock is the Compaq 286 deskpro. We believed the problem is either with the crystal tuning circuitry or the quality of the crystal itself. The reason for this was that a customer complained that his clock was too fast. There is also a known problem among PC designers about the battery circuitry. There is a certain condition that can occur such that the battery life can be shortened substantial. What happens is that the battery is also driving the +5V of the clock when the system is on. T. Chu