hi,

>Subject: DEC 3000/M400 cooling
>To: None <port-alpha@netbsd.org>
>From: der Mouse <mouse@Rodents.Montreal.QC.CA>
>List: port-alpha
>Date: 07/16/2004 00:01:50 
>I have a
>
>DEC 3000 - M400, 133MHz
>8192 byte page size, 1 processor.
>total memory = 192 MB
>
[...]
>It now seems to suffer from some kind of thermal problem: after 
>running for something like ten minutes, it locks up hard.
[...]  Power-cycling it brings it back,
>Holding the motherboard with the back-panel connectors facing you,
>component side up, the most obvious heat generator is of course the
>CPU.  But looking to the left of the CPU, there are some smaller chips,
>then two big chips, then a bunch of some 15 to 20 smaller chips.  This
>latter bunch run hot enough to worry me.  And on the motherboard,
>underneath the daughterboard with all the back-panel connectors,
>approximately under the area behind the ISDN/Ethernet connectors,
>something is getting very hot - the back of the board (the only
>accessible part) gets hot enough that I could probably injure myself if
>I left my hand on it very long - despite my having placed a fan
>specifically to cool that area.
>
>Is this normal?  It certainly appears that the built-in cooling in the
>case wasn't keeping it cool enough, despite there being nothing
>obviously wrong with it (I looked for clogged dust grilles and such).
>Also, it used to work fine when all put back together in the case.

I'm not an AXP-man, I just hardly remember that yours must be one
of those heavyheavy large desktops - isn't it?

my thoughts go the following path: possible little spare margin of
cooling power by original design -> aging -> cooling power weakened
-> over time, machine runs hotter increasing temperature gently ->
some DC-Voltage filtering chemical capacitors dry out -> noise + peaks
on DC-Supply increased over time now up to a level which causes on-chip
protection diodes (or whatever circuits there be) to heat slightly more, 
possibly even to a latch-up condition finally ending up in the error 
condition you describe

you also mention hot circuits near the connectors to the outside 
world, right?  may it be that the venerable machine suffered an 
a tad to strong dircharge (think surge, ESD or the like) ?

am I way too off and out in SciFi land with my thoughts?

good luck
Stephan