Re: DECstation 2100: power trouble?

["Maciej W. Rozycki" <macro%orcam.me.uk@localhost>]

On Wed, 15 Apr 2026, Dave McGuire wrote:

> > Curious.  Do you happen to recall whether you had anything connected to
> > the keyboard and mouse connectors?
> 
>   Man this was 1993, those neurons are all but gone.  But I don't think we
> used the mouse terminator plugs for the mouse port.  We definitely didn't have
> keyboards or mice connected.  I do recall "setenv console 8".  We always had
> good batteries in our machines. (this was a production environment in a
> company)

 Now that you mention that it seems to match my recollection (and console 
logs I've obtained over the years confirm it) that if you set the variable 
to use a serial port console, then the absence of a keyboard/mouse does 
not cause a self-test failure.

 But I've left them plugged in, because if the system has been off long 
enough for the battery pack to drain (a couple of months to a year, from 
experience; most likely from self-discharge rather than actual current 
drawn by the RTC chip), then you'll get these self-test failures again 
(and if you do leave the VSIMM in, then you'll actually have to attend the 
system to boot, and mine is in a rack at a remote lab half a day travel 
away).

 Some other DECstation systems do fail self-test with certain firmware 
revisions if a keyboard or mouse is missing even if the serial console has 
been chosen, as long as there's a video option installed.  That has to be 
considered a firmware bug I suppose.

  Maciej