It happens regardless of the type of drive connected, and it does not happen when any drive is connected to the power supply and JUMPERED INCORRECTLY, or with a slave drive that's jumpered incorrectly. The drive does not have to be *in use*, in fact there is no way of even accessing it during this test. It's only when the Promise BIOS correctly sees and identifies what's connected to it that the errors occur. There is only one 'lytic cap near the Promise chip, I can cap value it and ESR it and see what happens, but since others are having this problem I doubt it's bad. But, could be a "bad batch" of caps Asus had. I'll send you image of the area off-list. Well, here's the link of the photo (large, 381k). http://orpheuscomputing.com/images4/P4C800E_Deluxe-Promise_chip.jpg
Speaking of catastrophic failures: don't anyone think an optical drive's laser isn't powerful. LOL! I've come across CDROM drives with holes "laser burned" into the tops of the drive! Now that's one way to "REALLY mess up your media". -Clint God Bless Us All Clint Hamilton, Owner http://OrpheusComputing.com ) ----- Original Message ----- From: "Elmer L. Tibbetts Jr." <[EMAIL PROTECTED]> Clint, That still sounds like a filter cap problem because you are only seeing the problems when the drive is in use (drawing power). Do you have a spare controller or drive to try to isolate the problem to one or the other? These types of problems are always "fun" to isolate, not like one drive I had a while back. I noticed the access light being on when I wasn't accessing it I was thinking possible virus (don't ask) then I remembered that that drive didn't have an access light (right about the time that the smell hit). The voice coil driver chip on the hard drive interface board had failed catastrophically and was burning a hole through the circuit board. It did make trouble shooting real simple but didn't do my blood pressure any good in the short term ("interesting times"). Have a good one, Elmer At 17:52 25-11-03, OrpheusComputing.com wrote: >Hi Elmer, thanks. Haven't heard from you in a while. :) I was >just about to post an update to this. I've narrowed the >problem down to the Promise controller. I get NO ERRORS with it >enabled, but with nothing connected. If I connect ANYTHING to >the SATA or IDE connectors on the Promise controller chip, I >get thousands of errors. The device MUST BE ACTIVE to get >errors. In other words, if the drive is not plugged in to >power, no errors. If the drive is jumpered INcorrectly, but >plugged into the power, no errors. When it's DISabled, I get >no errors. When it's ENabled I get no errors if nothing is >connected. I only get errors with the controller enabled and >with something connected. It can be an SATA drive, OR an IDE >drive, doesn't matter, just as long as it's connected to the >Promise connectors. >-Clint > >----- Original Message ----- >From: "Elmer L. Tibbetts Jr." <[EMAIL PROTECTED]> > >Clint, > > It sounds like you might have a power problem. Very >carefully >check all of the electrolytic capacitors on the mother board >near the >memory slots and voltage regulators for signs of leakage or >swelling, >because this sounds a lot like a ripple/dirty DC power problem. >Pay >special attention to the area around & under each capacitor as >well as the >tops, what you are looking for is a (normally dark brown or >black) residue >(not old flux - that "should" have been removed prior the board >being >shipped) or any sign that the capacitor has been out gassing. >If it has >then it will have to be replaced to bring the amount of >capacitance >(filtering) back up. These kinds of problems are easily >avoidable IF the >board is properly designed in the first place but it's not >"cost-effective" >to do that any more. > > One final note, it is possible for an electrolytic >capacitor to >fail without showing any external visual changes, in this case >the only way >to isolate the problem is to pull the suspected capacitors and >test them on >the bench (with multi-layer boards I just isolate it down as >far as I can >using a scope/analyzer and then replace the capacitors starting >using the >"best guess" algorithm (a mix of engineering, visual >inspection, and luck) >until it works. (It sounds strange but it tends to be much >faster & >cheaper then trying to either isolate the bad component(s) with >reverse >engineering & test equipment.) But if the board is still under >warranty >this is probably not an option ... yet. > >Hope this helps, > >Elmer > >At 10:55 24-11-03, [EMAIL PROTECTED] wrote: > > >Hi all. I'm going to try and post this here since no one has > >been able to answer it in various technical mobo forums. I > >realize it's a pretty technical problem, but there are a few > >techie members here that may be able to help. > > > >I've been wondering for over a month now, why with the Promise > >controller enabled on my Asus P4C800-E Deluxe (Rev 1.02) I get > >thousands of errors on Memtest86 AND the Microsoft RAM > >diagnostics utility? I've changed video cards (nVidia & ATI), > >power supplies (top rated Fortron-Source FSP400-60PFN & cheap > >Leadman 400 watt), **MEMORY, everything and it still happens. > >NO PCI cards are installed. It DOES NOT matter if anything is > >even connected to the Promise controllers! It just has to be > >enabled in the BIOS and the errors are incredible. When I > >disable the Promise controller, no errors! Legacy USB is > >DISABLED in the BIOS. Overclock settings and memory settings > >have no effect on the errors. Whether FSB is 200mhz or >255mhz, > >I get the same results. Whether Turbo and PAT are on or off, >I > >get the same results. Regardless of voltages, same results. > >Regardless of ANY settings, the results are the same; Promise > >on, thousands of errors. Promise off, no Memtest or MS RAM > >diag's errors. The mobo is not in a case. **Memory tried is > >Geil PC4000 Ultra Platinum, and some generic PC2700. ONE SLOT > >OR TWO SLOTS used, same thing, even switching slots with one > >stick. I CAN run the memory tests on my P4B533 flawlessly > >using both types of memory. So, there is nothing wrong with > >the memory, and even it was it would not explain why only > >errors with the Promise enabled. CPU is a retail 2.4C (that's > >a P4 2.4ghz with 800mhz quad-pumped bus). The HD on the > >Promise controller can be accessed and used fine. > > > >Since I have two chips which look like VR's (voltage >regulators > >or FET's) marked "9115" on this mobo that are about to melt, > >I've been trying to get an RMA for SEVEN WEEKS now, and after > >that long and over 60 emails and form fill-outs the Asus site, > >NO REPLIES. They won't even reply to address the Promise > >controller memory error issue. One of these "9115" chips are > >next to the AGP slot, the other next to the DIMM slot closest > >to the outside edge of the mobo. I'm wondering if that issue > >and the Promise issue are related, as in a screwed up mobo. > > > >BIOS version is 1011. I am aware of v1012 & 1013 and the > >issues surrounding them (Asus pulled both 12 & 13 from their > >website). I have both 12 & 13, but of course am not going to > >try them for one; they were Beta, and I've heard of some fried > >mobo's with 13. > > > >FWIW, This mobo has the Intel ICH5R SATA controller, as well >as > >the Promise SATA and IDE controller. I had planned on using >an > >ATA133 HD on the Promise controller in the IDE non-RAID >setting > >for downloaded and other backups, and I have a WD Raptor SATA > >drive as the main HD. This was the reason in finding a board > >with a "3rd IDE" connector so no devices would have to be > >slaved. (Conventional 2 IDE ports have CDRW and CDROM drives, > >ICH5R has the main Raptor SATA HD, Promise has the IDE HD > >backup). These are not add-in PCI cards, but integrated > >motherboard chip controllers just like the typical IDE chip > >controllers, except for being RAID capable. I might add, the > >ICH5R is a dedicated bus chipset, and the Promise chipset runs > >on the PCI bus. > > > >So........what's going on here? Is it normal to get all these > >memory errors with the Promise controller enabled, and if so, > >WHY? If anyone here should happen to have one of these >mobo's, > >would you PUH-LEEEEZE take a few seconds and boot from the > >Memtest floppy or MS RAM test floppy and check the errors with > >the Promise controller enabled and disabled? If someone would > >do this, then I would know if this normal or not and if normal > >I could (after TWO MONTHS) finally get this SOB in the case >and > >completed. As far as the 9115 chips, I can put heatsinks on > >them if there is no Promise problem. I'd also appreciate it >if > >any P4C800-x users would touch these "9115" chips and see if > >they are this hot. Any input or suggestions from anyone > >appreciated. (BTW, all works fine when it boots into XP > >whether the controller is enabled or disabled, however memory > >error issues potentially indicate serious stability problems > >that must be addressed, not to mention a possibly screwy > >Promise chip which I'm trying to ascertain). > >Thank you! > >-Clint ============= PCWorks Mailing List ================= Don't see your post? Check our posting guidelines & make sure you've followed proper posting procedures, http://pcworkers.com/rules.htm Contact list owner <[EMAIL PROTECTED]> Unsubscribing and other changes: http://pcworkers.com =====================================================
