From: Big E on 12 Dec 2008 14:18 Thanks Paul. Tried moving the PCI cards all around to different slots to no avail. The USB card has an NEC chipset and reports the proper IDs that compare to the .inf files. The OS just won't recognize the drivers or something in the ID tranlation which takes me back to my previous post... If the onboard USB is assigned as PCI slot 0 and now neither of my other two PCI cards are recognized in any slot 1 - 3, could the PCI bus have gone bad? Seems unlikely but sometimes you never know. I'll probably have to shell out for one of those new-fangled kick-butt boxes and relagate this one to lowly file server status. Seems a shame since it's been a fantastic runner.
From: Big E on 13 Dec 2008 14:16 We're getting into an area where I don't know the nuts and bolts, but with logic and your help, I'm learning. If it has screws, plugs and knobs, I'm all over it otherwise. I'm not a big fan of hot fixes either for similar reasons. I looked at the USBMan website info but I didn't try anything there yet as none of the PCI slots seem to be working correctly at this point so the problem seems bigger than the SiS 7001. Everest recognizes both the onboard SiS USB and the NEC USB card for what they are as far as I can tell though I'm far from being an expert. The video card (Gigabyte GeForce FX 5200) shows up correctly as well. The audio card (Echo Mia) is another story altogether as Microsoft would have no data for that one. It shows up as Multimedia Controller [NoDB]. I'll try a reinstall on that one later. As far as the CPU chip goes, I'm guessing that the folks who assembled the mobo only had 3.4GHz chips on hand. The board is designed to run at 2.8GHz, 800MHz FSB with a Prescott chip, hence the discrepancy. Clock info from CPU-Z and Everest as follows: Per CPU-Z CPU core speed 2806.5 Mhz Multiplier 14.0 Bus Speed 200.5 MHz Rated FSB 801.8 MHz Per Everest: CPU Clock: floats around 2804.65 to 2808.13 MHz Multiplier: 14.0 FSB: 200.33 to 200.58 MHz Memory bus same as FSB Seems to me not too far off the mark. The Award BIOS date is 12/28/04 and the version number matches the last one released by Foxconn though their website says it was released in 2007. Go figure. I'm reluctant to use a third party update - I've seen those royally hose up a machine. BIOS SuperSpeed set at 200 MHz (allowable range 200 – 232). On the Frequency/Voltage Control page, items listed are as follows: Auto Detect DIMM/PCI Clk: Enabled Spread Spectrum: Disabled CPU DRAM Frequency: SPD I haven't found anything that specifically states at what speed the PCI bus is running. AGP must be OK, otherwise I probably wouldn't be able to see what I'm typing. The installed memory is currently now 2GB PC-3200 400 MHz; was 1GB 400 MHz when I started this thread. What's next? Thanks.
From: Paul on 13 Dec 2008 17:01 Big E wrote: > We're getting into an area where I don't know the nuts and bolts, but with > logic and your help, I'm learning. If it has screws, plugs and knobs, I'm all > over it otherwise. > > I'm not a big fan of hot fixes either for similar reasons. I looked at the > USBMan website info but I didn't try anything there yet as none of the PCI > slots seem to be working correctly at this point so the problem seems bigger > than the SiS 7001. > > Everest recognizes both the onboard SiS USB and the NEC USB card for what > they are as far as I can tell though I'm far from being an expert. The video > card (Gigabyte GeForce FX 5200) shows up correctly as well. The audio card > (Echo Mia) is another story altogether as Microsoft would have no data for > that one. It shows up as Multimedia Controller [NoDB]. I'll try a reinstall > on that one later. > > As far as the CPU chip goes, I'm guessing that the folks who assembled the > mobo only had 3.4GHz chips on hand. The board is designed to run at 2.8GHz, > 800MHz FSB with a Prescott chip, hence the discrepancy. Clock info from CPU-Z > and Everest as follows: > > Per CPU-Z > CPU core speed 2806.5 Mhz > Multiplier 14.0 > Bus Speed 200.5 MHz > Rated FSB 801.8 MHz > > Per Everest: > CPU Clock: floats around 2804.65 to 2808.13 MHz > Multiplier: 14.0 > FSB: 200.33 to 200.58 MHz > Memory bus same as FSB > > Seems to me not too far off the mark. > > The Award BIOS date is 12/28/04 and the version number matches the last one > released by Foxconn though their website says it was released in 2007. Go > figure. I'm reluctant to use a third party update - I've seen those royally > hose up a machine. > BIOS SuperSpeed set at 200 MHz (allowable range 200 – 232). > On the Frequency/Voltage Control page, items listed are as follows: > Auto Detect DIMM/PCI Clk: Enabled > Spread Spectrum: Disabled > CPU DRAM Frequency: SPD > > I haven't found anything that specifically states at what speed the PCI bus > is running. AGP must be OK, otherwise I probably wouldn't be able to see what > I'm typing. > > The installed memory is currently now 2GB PC-3200 400 MHz; was 1GB 400 MHz > when I started this thread. > > What's next? > > Thanks. > OK. I hadn't considered the possibility of a mismatch between the Vcore regulator capabilities and the processor's power class. I'm having trouble remembering the right technical language now, but basically the BIOS checks the conditions, finds say an 04B processor, realizes the Vcore is only designed for 04A, and can select the fallback value for the multiplier. So instead of x17, it is using x14. This is also why 2.8GHz processors have no fallback option, because they're already at x14 multiplier. (Prescott PRB=0 and PRB=1 bit, indicates power requirements) http://download.intel.com/design/Pentium4/datashts/30056103.pdf If it was a Prescott, the geometry would be listed as 90nm instead of 0.13u. When Prescotts came out, they had a pin definition changed on the bottom of them, which detected whether the motherboard was "Prescott Ready". The processor itself, would refuse to start, if the pin was in the wrong state. Thus, for many people, there isn't any fallback case to consider, because it won't run at all. I don't know what happens, if you bust off, or insulate the pin as appropriate, to override the behavior. (The pin on the bottom of the chip, is a different issue than the PRB enumeration from an internal processor register.) In checking the info from Foxconn, it does say the motherboard supports Prescott. But it doesn't say anything about power class. And looking at the motherboard picture in the manual, the board has a two phase regulator ??? So that would explain Foxconn throttling back on the multiplier. Kinda a pointless design feature - if you support Prescott, why go for a gutless Vcore ? Prescott was the king of power consumption, and should have a decent regulator. (This may, in fact, be the first motherboard I've run into, that made that kind of design decision. Intel provided for this possibility, but I cannot say I've seen many motherboards using the feature.) So that provides a partial explanation of the frequency, and based on your info that the input clock is 200MHz, there is no reason to suspect the PCI or AGP clock. I just figured someone had messed around, to get it to run at 2.8GHz. (200/6 = 33Mhz, 200 is a canonical frequency, and dividers would be set properly for it.) The "Multimedia Controller [NoDB]" is coming from Everest and its own internal database. Which is perfectly normal. I have hardware here which cannot be identified. What I like Everest for, is giving the DeviceID, so I can look up the hardware myself. The official PCI registry is not available to the public, so instead we have to rely on the publicly maintained list of DeviceIDs. (Either that, or buy a current copy of Everest from Lavalys :-) ) Leftmost column is VEN, next is DEV, and the third level is SUBSYS. http://pciids.sourceforge.net/v2.2/pci.ids You can verify the DeviceID and other numbers, do map to the product (if it is in that file). And the most important part, is that the VEN/DEV/SUBSYS appears in the INF file of the driver installer for the product. There has to be a match, for it to work. The purpose of looking in Everest, is to check that the DeviceID type information, agrees with the INF file of the installer you're using. I'm guessing at this point, that the hardware is fine, and something is going on with drivers. There is an example here, of a place where corruption can occur. "XP asks for drivers" http://www.uwe-sieber.de/usbtrouble_e.html First link involves "driver.cab", the second "the data is invalid" (a registry settings issue). http://groups.google.ca/group/microsoft.public.windowsxp.hardware/msg/f79c4c044536f6af?dmode=source I'm not suggesting that you run off and try any of those! Those are just examples of corruptions that can occur in Windows, and either screw up the installation of hardware, or result in requests to install drivers again. I'm thinking it is a software problem, but I don't recognize the symptom of finding "extra" unrecognized entries in Device Manager. Paul
From: Big E on 13 Dec 2008 23:17 Paul - Hmmm... After all this jacking around, the onboard SiS Open Host Controllers show up correctly in under USB Controllers in Device Manager again instead of Other Devices as was the case there for a while but, alas, the drivers can't be found by Windows when I enable them. I'll have to compare the device IDs to what's in the .inf file and then to the list on Sourceforge. At a quick first glance, the VEN/DEVs are correct. I didn't look at subsys yet or the compatible ID list in Device Manager. One question - Is the Matching Device ID shown in Device Manager/Details from the .inf files? The USB Controllers and audio card that won't install don't show one. All the other devices do. I won't be able to dig any further into this for a couple days but I'll post what I find. If it will help, I can upload screenshots to MediaFire and post the links. I really appreciate your thoughts on this. Thanks
From: Paul on 14 Dec 2008 03:00 Big E wrote: > Paul - > Hmmm... After all this jacking around, the onboard SiS Open Host Controllers > show up correctly in under USB Controllers in Device Manager again instead of > Other Devices as was the case there for a while but, alas, the drivers can't > be found by Windows when I enable them. > > I'll have to compare the device IDs to what's in the .inf file and then to > the list on Sourceforge. At a quick first glance, the VEN/DEVs are correct. I > didn't look at subsys yet or the compatible ID list in Device Manager. > > One question - Is the Matching Device ID shown in Device Manager/Details > from the .inf files? The USB Controllers and audio card that won't install > don't show one. All the other devices do. > > I won't be able to dig any further into this for a couple days but I'll post > what I find. If it will help, I can upload screenshots to MediaFire and post > the links. I really appreciate your thoughts on this. > > Thanks I hadn't really looked at it in that much detail. On trying a few menu entries from the Details in Device Manager, it looks like the info is derived from something. I don't typically rely on Device Manager for that anyway, because Everest makes it easier to quickly review the entries (fewer clicks). If you look at the Everest info, it almost looks like all devices in the Devices:PCI are having their config space data shown. So it looks more like an attempt to read each hardware device directly. USB 1.1 drivers should have been present in the original WinXP (dunno about completeness of all brands and instances of USB devices though). SP1 added USB 2.0 drivers. If the USB drivers won't install, some other software failing is preventing it. The trick is to figure out what is doing it. Any error messages, or things you can find in log files, might help. You'd be surprised how many .log files there are floating around in there. For example, if you're having trouble sleeping at night, start reading the "setupapi.log" file :-) Mine is half a megabyte. I see a few USB entries. Paul
First
|
Prev
|
Next
|
Last
Pages: 1 2 3 4 5 Prev: CPU test Error Next: Why don't camera reviews cover the data connection to the PC? |