At 14:02 -0500 01/26/2003, PCI PowerMacs wrote: >Date: Sun, 26 Jan 2003 09:05:06 -0500 >From: Linda Hutcheson <[EMAIL PROTECTED]> >Subject: Bad memory help! > >I ran Ramometer and came up with 3 error messages after 28 passes. > >Memory Error at Address $IDE354EC expected $07FD2ADO but read $07FD2ACO >Memory Error at Address $IDE414EC expected $003FE2956 but read $003FE296 >Memory Error at Address $IDE364EC expected $003FE2956 but read $003FE296 > >Unfortunately I have no idea what any of this means.
I have not heard from anyone who knows how to translate the addresses which RAMometer reports into which memory stick has the problem. I think that first character in the addresses you reported should be a 1 and not an I. The digits used above are in base 16 with A - E representing the digits from 10 - 15. The Address is the location in your memory which has the problem. the expected/read numbers are the numbers stored in that Address and the number which came back out of the Address. When what comes out does not match what goes in, it's bad. You clearly have one or more defective RAM DIMMs. How many DIMMs do you have? You need to do some testing with the RAM Sandwich method to determine which specific DIMM is bad. You see, RAMometer cannot test the memory which is occupied by the operating system. So even if a DIMM passes RAMometer, if that DIMM was storing part of the OS, then you don't know that that DIMM was good, because a section of the DIMM was not tested. The OS usually occupies space at the top and bottom of address space, so if you put one known good DIMM in A1 and another in A4 (assuming this is an x500 or x600 machine) then the DIMM(s) in A2 and A3 should get a good test. Leave the B bank empty. At this point I'm assuming that you don't know if any of your DIMMs are good. So what I would do is test just three DIMMs as described above at a time until you find a batch of three that will pass more than 30 iterations (you really need to run 1500, but this will do for now) of RAMometer. You can't be sure the two outer (A1 and A4) DIMMs are good, but the one in A2 or A3 is good, if RAMometer passes. Once you find a group of three which passes RAMometer, put a sticker on the middle DIMM (the one in A2 or A3) to mark it as good and move it to A1. Continue testing with other DIMMs in A4 and A2 or A3 until you get another set which passes more than 30 iterations. Now take the middle DIMM from this batch and mark it with a sticker as well. Install it in A4. You now have two DIMMs which you know are good enough to pass more than 30 iterations of RAMometer. By putting them in A1 and A4 they will soak up the OS space (provided they're not 8 or 16 MB DIMMs--too small) . Now test all the rest of your DIMMs one or two at a time by installing them in A2 and/or A3 and running RAMometer again. Eventually, you'll get a failure again, and that will be the one that is causing your problems. To really test your RAM you should do the above with 1500 iterations of RAMometer (including the part to identify good RAM for the outer slots). RAMometer can take as many as 1200 - 1300 iterations to find a failure. I haven't heard any reports above 1300 iterations, but I like to use 1500 as a margin. Most failures turn up in the first 50 iterations though. 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:[EMAIL PROTECTED]> To unsubscribe, email: <mailto:[EMAIL PROTECTED]> For digest mode, email: <mailto:[EMAIL PROTECTED]> Subscription questions: <mailto:[EMAIL PROTECTED]> Archive:<http://www.mail-archive.com/pci-powermacs%40mail.maclaunch.com/> Using a Mac? Free email & more at Applelinks! http://www.applelinks.com
