Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!apple!usc!samsung!uakari.primate.wisc.edu!unmvax!ariel.unm.edu!hydra.unm.edu!jcmiller From: jcmiller@hydra.unm.edu (Jeff Miller) Newsgroups: comp.sys.sgi Subject: Non Volatile RAM checksum error!! Message-ID: <1990Sep7.071805.22982@ariel.unm.edu> Date: 7 Sep 90 07:18:05 GMT Sender: news@ariel.unm.edu (USENET News System) Distribution: usa Organization: University of New Mexico, Albuquerque Lines: 45 I have had this repetitive problem with my SGI 4d/20 system that I am rebuilding. Every time I try to boot, cold-start or reboot the system, I get the following sequence of events: (Actually, I usually get just the error message, with some or all of the other stuff. The following is what I always get after powerup) ---------------------------------------------------------------- Console DUART test PASSED Memory walking bit test PASSED Memory address uniqueness test PASSED Non-volatile RAM checksum is incorrect: Initializing the non-volatile RAM parameters. Interrupt mask registers test PASSED Graphics subsystem test PASSED Non-volatile RAM checksum is incorrect: Initializing the non-volatile RAM parameters. ---------------------------------------------------------------- I suspected that the problem might be a low lithium battery, but my DVM quickly cleared up that question. Besides, my clock keeps *perfect* time, even with the power off for 2-3 days (more, perhaps, but that is about as long as I care to leave it off 8-) I suppose that this could be a software problem, since I have not been able to find any information about how to initialize the NVRAM in the limitted number of manuals that I own. My problem is that the system keeps telling me that it is, "Initializing the non-volatile RAM parameters," and yet the system complains that the parameters have not been initialized. A friend said that the problem might be a bad NVRAM chip on the motherboard. Can someone tell me whether or not the NVRAM is really an EEPROM or just simply a battery-backed CMOS static RAM? Perhaps it is combined with the clock/calendar chip, as it is with some of Motorola's clock chips. (The clock chip in my system happens to be made by National Semiconductors --> DP8572AN <-- I have no spec sheet on it...) Any comments, solutions, remarks, etc. would be GREATLY appreciated! jcmiller@hydra.unm.edu