Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!usc!rutgers!cbmvax!daveh From: daveh@cbmvax.commodore.com (Dave Haynie) Newsgroups: comp.sys.amiga.hardware Subject: Re: More on my sick 2500/30 Message-ID: <22162@cbmvax.commodore.com> Date: 4 Jun 91 19:52:31 GMT References: <9106040619.AA27553@wugate.wustl.edu> Reply-To: daveh@cbmvax.commodore.com (Dave Haynie) Organization: Commodore, West Chester, PA Lines: 74 In article <9106040619.AA27553@wugate.wustl.edu> FINEBERG@WUMS2.WUSTL.EDU (Charlie Fineberg) writes: > To recap: On both machines the problem is not > Denise > Paula > Gary > Buster > Rom > 68000 >because I swapped them from one machine to the other and both machines >behave the same as they did before. I have tried swapping the two CIAs on >my 2000 and still have had no luck. I am still hunting and very >frustrated. Anyone want to sell me an old Agnus and chip puller? >Anyone want two A2000's cheap? Aaargh. Chip swapping, while the easier way to track down problems, is also about the worst way to go about it, unless you have a very obvious indication that a particular chip is bad. If you don't, you'll only be lucky if you find the problem. Obviously, not everyone's a computer designer (though, of course, our lab and production technicians, along with an unknown number of field service people, undoubtedly debug these things a zillion times faster than the folks who built them, since they have practice and know what goes wrong, where all we know is how the thing works) with access to 'scopes and analyzers. First thing you do is check out anything that you might have changed in the system. Did you plug in a new peripheral? Hook up that printer last night with the power on? Recently add plug-in memory? Change Agnus chips? Have any thunderstorms or power surges? Computers do sometimes "just fail", if a marginal part finally makes it under the margin for good. But most of the time, something causes the problem. Being that volt meters are common, though, I always recommend that early on, you check the computer's voltage, at a remote point on the motherboard or, better yet, an expansion card if you have one. The voltage should read just about exactly 5V, with everything turned on. A computer could work just dandy for years, just on the edge of the margin, then get flakey as soon as one more little thing is attached. Zapping your system somehow is annoying, since it can get many different parts of the system, but some are more commonly clobbered than others. If you mis-plug something in a floppy or parallel port, you can damage either CIA chip. If you mis-plug the video port, you might zap a small decoupling resistor that links Denise to the main voltage supply. I've even heard of someone zapping the protection diode on the "+12V_User" supply, which is had at the video or serial ports. It's also technically possible to destroy Agnus (via the video port) or Paula (via the serial or mouse ports) from the outside, though this happens less often than a zap of an 8520, if for no other reason in that there are about 10 CIA pins for every one Agnus/Paula pin at these ports, so the CIAs are a bigger target. After that, remove EVERYTHING that's hooked to the system except for the monitor, keyboard, and floppy. If it still fails, you can attack the problem methodically. Of course, it's much better with a 'scope, then you can follow things very closely. You check to make sure you have all the proper clocks running. You check to make sure the OVL pin is going low out of the 8520 (this can be checked with a logic probe, even). Check the TICK signal going into the CIA, that can prevent the system from coming up. Check the RESET* signal and make sure its high. Check for activity of some kind on the 68000's AS* and DTACK* lines. Check the ROM chip select for activity. And finally, even experienced hackers can't solve every possible problem. Unless you know the system inside and out, and have the equipment to trace though the whole operation, you may not be able to determine what has gone wrong. At some point, you need your computer back in working order. You may have to take it to a service center, plain and simple. >Charlie Fineberg Biochemistry Dept., Wash. U. Sch. of Medicine (WUMS) -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "This is my mistake. Let me make it good." -R.E.M.