obrien@rand-unix@sri-unix.UUCP (08/11/83)
This message is empty.
scw@ucla-locus@cepu.UUCP (08/11/83)
From: Steve Woods <cepu!scw@ucla-locus> We went a bit further than that in terms of monitoring . . . "overtemp" stuff on the VAX power supplies is purportedly low enough to avoid a fire, but high enough to fry your VAX real good. I used to work on an 11/40 that got fried when the Air-conditioning people turned off the chilled water in our wing of the building; the machine devloped wierd flakey intermentent problems after a few months, such as hanging for 20-30 min then continuing (the only way you could tell [if no one was on it] was the system date/time would be *WRONG*), DEC Field service worked on the thing for 3 weeks and they couldn't fix it until they went and replaced **EVERY** board in the CPU. It seems that what happened was: each board had at least 2 chips that got moved out to the end of their life span by being cooked. Moral: it is a *****VERY***** good idea to have a thermal shutoff switch/circut in you power supply.
dmmartindale@watcgl.UUCP (Dave Martindale) (08/14/83)
The three VAXes in the math building at Waterloo are protected by a very simple overtemperature shutoff. The main circuit breakers feeding the electrical panels which supply the machines' power are a special sort known as "shunt-trip" breakers. They will trip on overcurrent like any ordinary breaker, but can also be tripped by the momentary application of 120V to a pair of terminals on the breaker. We have thermostats which close at a set temperature (25 or 30 degrees Celsius) located in the machine rooms and connected to these trip terminals. Thus if the airconditioning dies for any reason at any time, all power to the machine will be shut off before the room temperature gets high enough to cause damage. We haven't had any airconditioning failures yet, but it's cheap insurance.