Article View: cu.slug
Article #490Sickly 3670
From: reti
Date: Fri, 01 Nov 1996 00:00
Date: Fri, 01 Nov 1996 00:00
39 lines
2243 bytes
2243 bytes
Date: Thu, 31 Oct 1996 10:27 EST
From: Tim Bradshaw <tfb@aiai.ed.ac.uk>
I have a 3670 which has become ill. When powered on it either fails
completely to wake up (fault light stays on, screen never lights), or
it gets to the FEP prompt, but when doing Initialize Hardware Tables
it says one or both of `Checkword error in Datapath' or `Checkword
error in Sequencer'. And of course it won't boot after that (it gets
as far as starting to load the world, but then says something about
memory errors (VM error? I'm not quite sure).
I've replaced the CPU (datapath, memory control, sequencer) with known
good cards (they work in a 3640). I haven't replaced the FEP or IO
cards (I'm not sure they're compatible with a 3640?)
Yes, they are and the FEP board sounds like a good candidate to solve your
problem. The FEP uses the rest of the hardware to initialize registers
and load words into memory, if it is broken things can be pretty strange.
The disk definitely works (tried two).
I suspect that it may be trouble with the backplane somehow -- it's
alsways been a bit sensitive to having the cardcage disturbed -- but
I've reseated all the cards a good few times with no luck.
Perhaps some other card is hurting the backplane: is this possible?
I've tried it with only a minimal set cards in, and with several
different memory cards.
The supply voltages look OK to me.
So I don't know what to do now other than junk it -- can anyone help?
(I'm not completely convinced I'm subscribed to this mailing list (I
have tried!) so if anyone has ideas could you CC to me as well.)
Thanks!
--tim
Technical Details
Message-ID:
<846878714.19646@news.Colorado.EDU>Path:
arc1-ca.pugleaf.net!unknown!not-for-mailShow Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 10077e,2a9cc3750da2c7b5 X-Google-Attributes: gid10077e,public From: <reti@wilson.ai.mit.edu> Subject: Sickly 3670 Date: 1996/11/01 Message-ID: <846878714.19646@news.Colorado.EDU> X-Deja-AN: 193768820 distribution: cu sender: daemon@lace.colorado.edu x-note1: message-id generated by recnews x-note2: mail msgid was <19961101190956.6.RETI@STHELEN-VLM.AI.MIT.EDU> organization: University of Colorado at Boulder newsgroups: cu.slug