home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.amiga.hardware
- Path: news.kei.com!ub!dsinc!scala!news
- From: dave.haynie@scala.com (Dave Haynie)
- Subject: Re: My old A3000 is down HELP PLEASE
- Sender: news@scala.scala.com (Usenet administrator)
- Message-ID: <1996Jan15.213225.19529@scala.scala.com>
- Date: Mon, 15 Jan 1996 21:32:25 GMT
- Reply-To: dave.haynie@scala.com (Dave Haynie)
- References: <4d2k7l$2br@netserver.univ-lille1.fr>
- Nntp-Posting-Host: gator
- Organization: Scala Computer Television, US Research Center
-
- In <4d2k7l$2br@netserver.univ-lille1.fr>, thomin@univ-valenciennes.fr () writes:
- >A3000 mother board rev 7.3, daughter board rev 7, no extents
- >When I turn it on:
- >- with KS1.3, no problem
- >- with KS2.x, screen turn to pale yellow, then to black
- >What does it mean ???
-
- It means that you're triggering an exception of some kind in 2.x
- that's not present with 1.3 -- the yellow screen is the result of any
- 680x0 trap taken before the system is far enough along to display a
- System Error or Guru message.
-
- A common source of this problem in A3000 family machines is some kind
- of bus timeout. On the A1000/A2000 architecture systems, all bus
- cycles get a zero-wait-state transfer acknowledge if they don't
- explicitly ask for more time. That's impossible on the A3000, where
- zero-wait-state cycles would happen to fast to override, bus cycle
- control is more distributed (Gary, RAMSEY, DMAC, Buster, and the FPU
- each have independent control of when they end a cycle sent to them).
-
- So the default behavior of the system is to let a device end its own
- cycle. However, if you access something that doesn't exist, this would
- mean a never-ending cycle. Since we didn't want machines to lock up
- simply due to small programming errors, Gary has a timeout
- mechanism. When you power up, Gary waits about 1/4 second, then
- quietly ends the cycle. In order to actually detect when these cycles
- happen, though, Gary has a second mode that generates a bus error
- after 8us. Until the system is up running real 2.x trap handlers, this
- will trigger a yellow screen. It's never supposed to happen during
- start up.
-
- The most common cause of this problem on an A3000 is simple -- booting
- up without the daughterboard installed. Due to an oversight in the
- A3000 motherboard design (eg, I screwed up), one of the Zorro II
- "cycle termination override" signals is not pulled-up on the
- motherboard, only on the backplane. So without the backplane
- installed, this signal tends to float low, and so you get a Zorro II
- cycle with infinite wait-states, which gets overridden by the Gary
- chip, with bus error, before bus errors are properly trapped.
-
- Dave Haynie | ex-Commodore Engineering | for DiskSalv 3 &
- Sr. Systems Engineer | Hardwired Media Company | "The Deathbed Vigil"
- Scala Inc., US R&D | Ki No Kawa Aikido | info@iam.com
-
- "Feeling ... Pretty ... Psyched" -R.E.M.
-
-