hanes@raksha.eng.ohio-state.edu (Mark D. Hanes) (01/19/90)
This is a follow-up post to my previous post concerning a green screen on power-up along with a blinking power LED (11 blinks, pause, and repeat). Does anyone out there know how the RAM's are checked on power-up. Are they written to and then re-read for verification? Or what? The problem does not seem to be a mis-seated Fat Agnus. It has been re-seated, had the connector pins bent out and re-inserted, even replaced the socket itself. All custom chips have been checked by swapping one at a time with known working ones. The clock signals seem ok, and the refresh signals coming out of Agnus seem ok, as well. The problem is not a short with the RF shield either, nor a short between the board and the disk drive. Thus, I think the problem is a defective 256kx1 DRAM chip. I was curious how these are checked on power-up in order to get a better idea of what is failing. If anyone has any additional suggestions or an answer to the above question, please e-mail me at hanes@phao.eng.ohio-state.edu Thanks again, Mark. -=- Mark D. Hanes hanes@phao.eng.ohio-state.edu Electrical Engineering The Ohio State University
collins@pnet02.gryphon.com (Steven Collins) (01/20/90)
I don't recall the layout inside an A500, but in an A1000 it is VERY easy to cut traces underneath any of the socketed chips by removing them with a screwdriver. A real chip puller is greatly prefered... I 'fixed' and A1000 for a friend of mine once this way, but eventually found it by poking around with a logic probe... There is an excellent file called 'the flake report' about things that go wrong on amigas. It has many helpfull suggestions on possible problems and diagnostic procedures. All I can find is my hardcopy right now, but it suggests that a ram chip that is failed 'open' can be found by piggybacking a known good ram chip on top. It also suggests that a bad ram chip that is failed 'short' might be hotter to the touch than good rams. I will upload the whole thing, if I can find it and no one objects to inflamably... steve collins UUCP: {ames!elroy, <backbone>}!gryphon!pnet02!collins INET: collins@pnet02.gryphon.com