>Date: Fri, 8 Nov 2002 18:12:39 -0500 >From: Anna Silliman <[EMAIL PROTECTED]>
>It's odd that I had exactly this same problem >today, and thanks Kyle for the advice because mine >seems to be fixed. > >I want to know if a dying battery could have caused >all this horror? > > My 7300/Sonnet G3/OS8.6 had been >rock steady for ages. Then one day the monitor settings >didn't stick, and after the next restart stalled at >the welcome screen. However, with extensions OFF it >would load fine. Same situation after putting in a new >battery; and even a fresh new system folder didn't >solve the problem. > >I think it was the 30 minute wait suggested by Kyle >that did the trick. > >But why??? The solution you used usually solves corrupted NVRAM, so you most likely had corrupted NVRAM. NVRAM stands for non-volatile RAM and it is a small amount of memory (8KB) which is powered by the battery even when the machine is unplugged. Many settings are stored in NVRAM and these settings play a large part in the starting up process of your computer even before the machine looks for any disk drives. Some of the settings are used before the machine even makes the start up bong. If you have corrupted NVRAM, any number of problems can occur, the worst of which is a machine which will not bong at start up and so appears to have a fatal hardware problem. Removing power as Kyle suggested deprives the NVRAM chip of power and so resets its contents, eliminating any corruption of the contents. The root cause of your corrupted NVRAM could have been a marginal battery, as a low battery might *almost* maintain the NVRAM contents leading to lost bits in places. Failed OSX installs also lead to apparently dead machines from corrupted NVRAM rather often. At least I've seen a lot of reports that turned out to be that. Irrelevant aside for the hardware folks. The NVRAM is physically stored on a little chip which is an 8KB SRAM chip in an SOIC package. Apple used some Sony, some Samsung and some Winbond chips for this, depending on the batch. Several manufacturers made compatible chips, including Toshiba and Mitsubishi. I tried replacing the chip with a 32K pin compatible chip (NCs changed to address lines) but it appears that Apple either didn't implement the additional address lines (it was a long shot, shrug) or the 8KB of space is programmed into the ROM (perhaps 8K of address space is mapped to the NVRAM?). The 32K chips work, but they don't provide any extra capacity for the NVRAMrc file, as far as I can tell. There might be some subtle hack possible that I'm missing. Jeff Walther -- PCI-PowerMacs is sponsored by <http://lowendmac.com/> and... Small Dog Electronics http://www.smalldog.com | Refurbished Drives | -- Sonnet & PowerLogix Upgrades - start at $169 | & CDRWs on Sale! | Support Low End Mac <http://lowendmac.com/lists/support.html> PCI-PowerMacs list info: <http://lowendmac.com/lists/pci-powermacs.shtml> --> AOL users, remove "mailto:"; Send list messages to: <mailto:pci-powermacs@;mail.maclaunch.com> To unsubscribe, email: <mailto:pci-powermacs-off@;mail.maclaunch.com> For digest mode, email: <mailto:pci-powermacs-digest@;mail.maclaunch.com> Subscription questions: <mailto:listmom@;lowendmac.com> Archive:<http://www.mail-archive.com/pci-powermacs%40mail.maclaunch.com/> --------------------------------------------------------------- >The Think Different Store http://www.ThinkDifferentStore.com ---------------------------------------------------------------
