Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!cmcl2!phri!roy
From: roy@phri.UUCP (Roy Smith)
Newsgroups: comp.protocols.appletalk
Subject: problems with KFPS-2 on DELNI
Message-ID: <2848@phri.UUCP>
Date: Fri, 14-Aug-87 11:10:27 EDT
Article-I.D.: phri.2848
Posted: Fri Aug 14 11:10:27 1987
Date-Received: Sun, 16-Aug-87 04:58:35 EDT
Organization: Public Health Research Institute, NYC, NY
Lines: 34
Keywords: power supply


	We've been having problems with our new KFPS-2 crashing.  At first
we suspected software (we're running KIP/CAP, after lots of you folks
debugged my config files for me; thanks!) but now it looks like power
supply problems which may be widespread (hence this note).  We've got S/N
695; apparantly the problems are confined to a certain S/N range but I
don't know what it is.  I also don't understand (and neither does tech
support) why our brand new KFPS-2 has that S/N since currently shipping
units have S/N's over 1000.

	From what I gathered by talking to Kinetics tech support, on some
KFPS-2's, the power supply will start to drift if the kbox is on a DELNI.
On a xciever, the kbox supplies +5V to the xciever, but the DELNI has its
own power supply so doesn't draw any +5V current from the kbox.  If I got
the story straight, this causes the +/-12V in the kbox to drift.  This
sounds plausable, but I'm not a power supply guru.

	Supposedly, putting the kbox on a xciever should make the problem
go away.  In our case it just made it worse (we used an InterLan NT-100).
On the DELNI, the kbox crashed maybe once a day, on the NT-100, it was more
like every 15 minutes.  Unfortunately, moving back to the DELNI didn't get
us back to once a day crashes, so I suspect the NT-100 did something to the
kbox.

	BTW, by "crash" I mean the following.  We would be running OK in
both directions (Ether-AT and AT-Ether) when everything would stop working.
Ping would show the kbox dead.  Sometimes just doing a restart from
FastPath(tm) Manager would do the trick, sometimes it required a reload,
and sometimes the Manager wouldn't even see the box unless you cycled power
on it.
-- 
Roy Smith, {allegra,cmcl2,philabs}!phri!roy
System Administrator, Public Health Research Institute
455 First Avenue, New York, NY 10016