Archive-name: asus-faq.270995 Last-modified: 27th of september 1995 *************************************************************************** * ASUS: FREQUENTLY ASKED QUESTIONS * *************************************************************************** ABOUT THIS DOCUMENT This question-and-answer list is posted monthly to the Usenet newsgroup alt.comp.periphs.mainboard.asus. As soon as the author has finished reading the group news.group.questions and learned there how to set up a faq it hopefully will become official. Presently it is not official! It is simply posted from time to time to the newsgroup. Please send corrections, additions and comments to Magnus Thelander (magnus.thelander@jetset.ct.se) or...well still searching for anybody who is willing to help setting up this faq. ;-) This information is abstracted and condensed from the posts of many contributors to alt.comp.periphs.mainboard.asus, supplemented by infor- mation from ASUS support and vendors of asus products. You can get the latest released copy of the faq through ftp from the ASUS internet server or one of it's mirrors. For adresses refer to the document mentioned in Question 01. You should find the file under pub/ASUS/TechInfo/faq.txt. Thank you, Jerry Lee from AsusTek for the distribution through your ftp server. DISCLAIMER Absolutly NO GUARANTEES are made regarding the accuracy of the following information. COPYRIGHT This compilation is by Carsten Fuchs. Distribution over the Internet or by electronic mail is unrestricted. Other use requires my permission. *************************************************************************** CONTENTS Changes since last posting: Edited contact information above. Frequently Asked Questions: Q01) How to reach ASUS and NCR, ASUS related ftp-sites and NCR related ftp-sites? Q02) Where can I find specification for ASUS products? Q03) Where can I find general info on PCI and chipsets? Q04) What is the better cache scheme: write-through or write-back? How can I upgrade my SP3g with a dirty Tag RAM? Does it increase the performance ? Q05) What type of simms can the PCI/I-SP3g use? Q06) How to connect a mouse to the PS2 mouse connector on the ASUS SP3g? Which mice are PS2 compatible? Q07) What can cause problems with hanging devices during boot on the on the NCR 53c810 SCSI controller? Q08) Are there problems with Quantum or Conner harddrives and NCR controller? Q09) Is it possible to have a SCSI drive as the primary (bootable) drive and an IDE as secondary? Q10) The PCI E-IDE Controller CMD 640B and OS/2? Q11) How to run an early version of the AMD 486DX4-100 at triple speed. Q12) What is the undocumented jumpersetting to run an early version of the AMD 486DX-100 with an ASUS SP3g at triple speed (100MHz)? Q13) What is the undocumented jumpersetting to run an early version of the AMD 486DX-100 with an ASUS AP4 at triple speed (100MHz)? Q14) Why does the ASUS PVI-AP4 with an 486DX4-100 only have a Norton SI of 132 instead of 198? Q15) How can I run external SCSI devices with an onboard SCSI controller? Q16) ASUS SP3g-floppycontroller and Floppy-Streamers? Q17) How to merge an NCR SDMS Kernel in a Flash BIOS file? Q18) What can I do when I've lost my boot-password? Q19) What are the error codes of the AWARD BIOS version 4.50? Q20) Is there a general problem with simultaneous EIDE and floppy access? Q21) How to solve Ctrl-Alt-Del Problem with the BIOS Update tr5i0107 for the ASUS TP54 Board? Q22) Is there an error in the Mainboard manual for the P55TP4XE on the jumper setting for write-protection the BIOS? *************************************************************************** Q01) How to reach ASUS and NCR, ASUS related ftp-sites and NCR related ftp-sites ? |> (Last modified 15 April 1995) |> |> This Document contains information about ASUS and NCR support. |> |> ==================================================== |> || Here is a list of Asus info on the Internet: || |> ==================================================== |> |> Official ftp-site from ASUS: |> - asustek.asus.com.tw (192.72.126.1) /pub/ASUS |> |> WWW Page from ASUS: |> (Use for Productinformation only, very slow...) |> - http://asustek.asus.com.tw/ |> |> Other sites with BIOS-Updates: |> - daemon.mit.edu /pub/asus |> - ftp.cis.nctu.edu.tw /Vendors/ASUS/ |> - leica.ccu.edu.tw /pub1/vendors/ASUS/ |> - ecp541.ecn.nl \pub\ASUS |> - ftp.dfv.rwth-aachen.de /pub/msdos/ASUS |> - ftp.mathematik.uni-marburg.de /pub/mirror/asus |> |> Write to with serious problems: |> - tsd@asus.com.tw |> |> Guys from ASUS: |> - jerry@asustek.asus.com.tw (reads this newsgroup) |> |> ============================================== |> || Here is a list of how to contact ASUS: || |> ============================================== |> |> ASUSTek Computer Inc. |> * Tel: +886-2-894-3447 |> * Fax: +886-2-894-3449 |> * Address: No. 150, Li-Te Road, Peitou, Taipei, Taiwan, R.O.C. |> * Marketing Information: info-twn@asus.com.tw |> * Technical Support: |> - Fax: +886-2-895-9254 |> - BBS: +886-2-894-4667 |> - E-mail: tsd@asus.com.tw |> - WWW: http://www.asus.com.tw/ |> - Gopher: gopher.asus.com.tw |> - FTP: ftp.asus.com.tw/pub/ASUS |> |> ASUS America Inc. |> * Tel: +1-408-956-9077 |> * Fax: +1-408-956-9088 |> * Address: 1635 McCandless Drive, Milpitas, CA 95035, U.S.A. |> * Marketing Information: info-usa@asus.com.tw |> |> ASUS Computer GmbH |> * Tel: +49-2102-445011 |> * Fax: +49-2102-442066 |> * Address: Harkort Str. 25, 40880 Ratingen, BRD., Germany |> * Marketing Information: info-ger@asus.com.tw |> |> |> =================================================== |> || Here is a list of NCR info on the Internet: || |> =================================================== |> (Look also at the ftp-sites listed above for NCR support on ASUS |> boards) |> |> NCR Drivers: |> - ftp.symbios.com /pub/ncrchips/scsi/drivers |> - ftp.hmpd.com /pub/ncrchips/scsi/drivers |> |> NCR's technical support can be reached through the eMail adress: |> - ncr.chips@ftcollinsco.ncr.com. |> Please, be aware that NCR's states that support for their control |> provided through asustek. |> |> ========================================= |> || Technical support Hotline for NCR || |> ========================================= |> |> Also, a technical support hotline is available for PCI-SCSI software |> and hardware issues. This hotline is mainly for end user questions but |> it also contains information about the latest drivers and BIOS that |> are available. |> |> In the U.S call (719)573-3016 |> In Europe call ++49-8231-9644-18 (++ means international access code) |> In Germany call 08231-9644-18 |> |> |> |> Thanks to all who helped collecting this info. |> Send comments and additions to me. |> |> Clemens. For the newest version of this document refer to "FTP-Sites with ASUS Support", which is frequently posted in the newsgroup. Clemens Hofmann (zxmeg05@student.uni-tuebingen.de) mantains this ducement, so please send any changes (i.e. new servers, new web sites, unavailable old servers,...) to him. Please, try to use the server the nearest to your location ! ---------- Q02) Where can I find specification for ASUS products ? There are several ways to get the specifactions on ASUS products: 1. Through the ASUS WWW server (http://asustek.asus.com.tw) 2. Through ftp on the ASUS ftp server or one of it's mirrors (adresses see Q01). The directory should be: ASUS/Info/Spec. Please use the server the nearest to your location! 3. Look for a document "ASUS products specifications" in this Newsgroup. Rick Lindsay (rlindsay@bga.com) is posting this regurlarly. This is a service for all those poor guys ;-), who can access the internet mail and news, but can not use ftp and WWW. ---------- Q03) Where can I find general info on PCI and chipsets ? Answered by jdouglas@mitre.org (sorry, could'nt find your real name in the posting and your host was unreachable by finger) and Rod Smith (rsmith@psych.colorado.edu): An excellent information source (that I won't be duplicating) is a weekly post by Pat Duffy on comp.os.os2.misc. His last posting are also available for anonymous ftp from: ftp.netcom.com in the directory /pub/ab/abe. Get everything with pci in the title. Intel operates a fax-back service in the US, at 800-628-2283 with lots of information on PCI available. There's also the Intel BBS, at 916-356-3600. ---------- Q04) What is the better cache scheme: write-through or write-back ? How can I upgrade my SP3g with a dirty Tag RAM ? Does it increase the performance ? Answered by Stefan Esser (se@ZPR.Uni-Koeln.DE) and Markus Kuhn (mskuhn@cip.informatik.uni-erlangen.de): The ASUS SP3G can drive write-through or write-back cache. You can use both modes with and without dirty tag RAM present. In write-through mode, dirty tag RAM has no function. The BIOS default is "write-back", resulting in some 5% performance loss, because if there is no dirty tag RAM, then there are lots of unnecessary memory writes, since each cache line refill will need to be preceeded by a write back of four DWORDs, even if none of them had been modified. In write-back mode, dirty tag RAM increases real-world performance by approx. 1% over write-through. The small positive effect of write-back on PC compatible hardware is caused by the lack of write-allocate. In case of the Saturn chipset, the number of clocks for some transactions is higher for write-back and the Saturn has already very efficient write posting buffers which compensate much of the inefficiency of not having a dirty tag RAM. Starting with rev. 1.8, the SP3G got a socket for a dirty tag RAM. With this RAM installed. Suitable chips for 256kB cache size are 64K x 1 RAMs with 15 ns access time, e.g. Cypress Cy7C187, AT&T ATT7C187-15, Matra MHS HME 65787-15, Motorola MCM6287C-15 and Toshiba TC5562-15 or any other equivalent. ----------- Q05) What type of simms can the PCI/I-SP3g use ? Answerded by Stefan Esser (se@ZPR.Uni-Koeln.DE): The SP3g can only use pairs of simm modules, because the SATURM II chipset alsways enables memory interleave, and to do so you need at least to modules. So if you plan for example 16MB for your system count the price of 2 * 8MB. The board can use simms with or without parity as it doesn't use the parity bit on the simms. With 4MB and 16MB modules you can either use pairs of single sides or double sided ( = density, I like that one Stefan). Even though people reported that the successfully mixed a single sided and double sided 16MB module on one memorybank. With 8MB and 32MB modules you can only use double sided (= density) simms. ---------- Q06) How to connect a mouse to the PS2 mouse connector on ASUS mainboards ? Which mice are PS2 compatible ? Most mice don't support the PS/2 signals. Consider all the cheap Taiwanese mice shipping with most clones. Some pointing devices from Logitech and the Sicos Colani mouse are known to be PS2 compatible. All combination mice (serial and PS2 compatible) need two sets of leads in their cable: one for serial, one for PS/2. Only the adapter that comes with the mouse should be used, as there is no real standard for which pins on the DB9 will carry the four PS/2 signals. To connect a PS2 compatible mouse, either ask your local dealer for a special cable, or make one on your own. onbaord PS2-mouse connector | pin # signal (view: component side) | -------------------------- | 1 GND | 2 CLK 1 3 5 7 | 3 DATA | 4 missing pin 2 o o 8 | 5 N/C | 6 missing pin (o's are missing pins) | 7 +5 V | 8 N/C | Thank you Wolfgang Esslinger (wolfgang@ims.uni-stuttgart.de) for the advice to label the connector as it is usual for a ribbon cabel. ---------------------------------------------------------------------- | PS2-Mouse Socket | (front view: not soldering side) | pin # signal ------------------------- | | 1 DATA 1 2 | 2 N/C | 3 GND 3 4 | 4 +5 V | 5 CLK 5 6 | 6 N/C [] | | [] is the hole | Tip from Markus Plauen (apatpgpauman@apa.co.at): If the connector on mainboard was round you could directly solder the PS2 mouse socket on it (the two missing pins are the hole of the PS2 mouse socket). Thank you Jim McKie (jmk@plan9.research.att.com) for buzzing it out on the P54NP4 and thank you Reid Trimble (reid@lvld.hp.com) for calling ASUStek to get the information for the P54SP4. ----------- Q07) What can cause problems with hanging devices during boot on the on the NCR 53c810 SCSI controller ? Answered by Paul Bash (pbash@netcom.com): My experience has been that anytime your bus hangs its is almost always one of two things: 1) bad termination 2) bad cables or excessive cable length. The NCR 53c810 chip is a SCSI-2 FAST device. Normal SCSI-2 runs at up to 5MB/sec synchronous, works fine with passive termination and can tolerate a total bus length of 6 meters, or about 18 feet. SCSI-2 FAST tightens the timings such that 10MB/sec synchronous transfers are possible. This requires *active* termination and it is generally recommended that your total cable length not exceed 3 meters). If your devices are hung on the bus external to your system unit, this also _requires_ the highest quality shielded cables you can get. The $5 or $10 K-Mart Blue Light special just won't cut it. Good cables are typically $35-50 for 1 meter lengths and active terminators vary from $20-50 depending on the connector (at least around Boulder, CO). Mixing not FAST SCSI capable devices (i.e. older CD-ROMs) on a bus with another FAST capable one (i.e. most new disk drives) can be problematic (even though SCSI specs states it shouldn't), so moving it to a different bus might be a solution. You can buy an ASUS PCI-SC200 NCR-based SCSI card from Dave Gee (dgee@netcom.com) for $75 so that option isn't too painful. One other thing to check for is that the devices near each end of the bus have "term power" enabled. Only one device is _required_ to provide power to the terminator on the bus, which usually is the controller. In some situations it doesn't hurt hurt to have enabled "term power" on _all_ devices (I typically just enable term power on everything). Practical advice from Markus Kuhn (mskuhn@cip.informatik.uni-erlangen.de): An optimally configured SCSI bus gets the TERMPWR supply from exactly one device, by convention usually the host adapter. Systems usually also work if several devices provide TERMPWR, however the termination is not as efficient and in critical situations, this might reduce the reliability slightly. On the ASUS SP3G (not SP3), the TERMPWR line is fed from a separate voltage controler at the NCR host adapter, so it is particularly recommendable with the SP3G to let only the motherboard deliver TERMPWR, because the motherboard delivers a particularly good TERMPWR (good = noise-free). If you have a device which comes with active termination, put it at the other end of the bus, because active termination is much more robust than the usual old passive resistor arrays. The SP3G also comes with active termination. ----------- Q08) Are there problems with Quantum or Conner harddrives and NCR controller ? Answered by Stefan Hinker (hinker@cip.informatik.uni-wuerzburg.de) and Rod Smith (rsmith@psych.colorado.edu): Quantums LPS Prodrive serie has reported problems with the NCR controller. Some people state that their Prodrive work fine under DOS and OS/2 but not under LINUX. Even though this does NOT apply to all drives. It seems that this depends on the model and the ROM revision, so if you own an older Prodrive it is allways worth a try. The newer Quantum Drives, Lightning serie, does not have any problems with the NCR controller. Practical advice for OS/2: Try to add the option /!SN to the OS2CAM.ADD driver in your config.sys. Practical advice for LINUX: Martin Haneke (haneke@zedat.fu-berlin.de) reported to get the combination NCR and Quantum LPS 540S work. 1. Be sure to have only one device in the SCSI chain with termination power enabled. 2. Disable parity checking on the Quantun and _all_ other SCSI device as well. 3. Enable shadow ram for the NCR Bios. Use the adress DOSCAM.SYS or whatever reports. Answered by Dirk Rogalla (dirk@kant.ping.de): Conner drives with an older Firmware-Version have reported problems. Conner is aware of this problem and is shipping Firmware-Flash-Updates on disk. ----------- Q09) Is it possible to have a SCSI drive as the primary (bootable) drive and an IDE as secondary? Answerd by Rod Smith (rsmith@psych.colorado.edu): I've not tried this on the Asus, but unless the Asus BIOS is VERY weird, it'd be simple to impossible, depending upon the OS. The standard PC boot sequence goes something like this: 1) Do various things like RAM checks. 2) Look for IDE hard drives and enter them into a table. 3) Pass control to any cards with BIOSes on them, including SCSI control- lers. A SCSI controller will, at this point, add its own drives to the list begun in 2. (Note that the SP3G's SCSI BIOS is not on separate card, but I'd expect they'd keep the sequencing here the same as is standard, to avoid subtle [or not-so-subtle] incompatibilities.) 4) Pass control to the boot sector on the first hard drive in the list, which must necessarily be an IDE drive if an IDE drive is present and listed in the CMOS. That last phrase gives the wedge that's needed under some OSes -- if you leave the IDE drives unlisted in the CMOS entries, they won't be added to the list that the BIOS maintains, and so the machine will try to boot off of the first SCSI drive. Depending upon the OS, the IDE drives may or may not then be detected, and may or may not cause further problems during the boot. Under OS/2, for instance, it will work IF you move the order of the IBM1S506.ADD and OS2CAM.ADD files so that the former (the IDE driver) appears later in the CONFIG.SYS than the latter (the SCSI driver). I sus- pect that it would work OK under Linux without further muddling about, but I'm not positive of this. To the best of my knowledge, this wouldn't work at all under DOS, since DOS uses the BIOS for IDE hard drive support. I've no idea about Windows NT or any other OS. SPECULATION: It might be possible to do what you want under DOS by using a boot loader program such as OS/2's Boot Manager or Linux's LILO, but you'd also need another OS's FDISK or some trickery and drive swapping. (Of course, if you can install LILO or Boot Manager, you've obviously already got that OS's FDISK, too....) My notion is to use only logical/ extended partitions on the IDE drives, except for the boot loader, if it needs a partition; and put DOS on a primary partition on the first SCSI drive. Since the BIOS (or DOS, or whatever) will assign drive letters to all primary partitions first, and since the boot loader will transfer control to whatever partition you want, DOS should then boot up on C:, which will be on the first SCSI drive, with D: on the first IDE drive. I can think of several possible reasons why this might not work, but I don't know enough details about the PC and DOS boot processes to know for sure. ----------- Q10) The PCI E-IDE Controller CMD 640B and OS/2 ? Answerded by Gary Farmaner (dialog@turing.toronto.edu): The OS/2 drivers can't properly handle the CMD640B. You will have to replace the original OS/2 IDE drivers with the new ATAPI compliant CMD drivers, if you connect a CD-ROM with ATAPI interface or want to use both IDE controllers. ----------- Q11) What is the reason for the problems with external and internal clocking with early version of AMD 486DX2 and AMD 486DX4 ? Answered by Roger Wallace (rwallace@world.std.com): For 3.3V Intel and AMD CPUs, 2x mode is selected with the CLKMUL pin connected to Vss (ground); 3x mode is selected by leaving the CLKMUL pin "floating" or returning it to a logic "high" level. Unfortunately in initial versions, AMD did not use the same external pin as Intel did for the CLKMUL function. As a result, we are seeing cases where an AMD DX4-100 always runs at 66 MHz because the motherboard happened to ground pin B13 (CLKMUL for AMD, no internal connection for Intel); alternatively, we hear of occasional AMD 3.3V DX2-66s running as DX4-100s and AMD 3.3V DX2-80s running as DX4-120s (!) -- these "events" occurring in motherboards which do not specifically accommodate the AMD chip, and leave the B13 pin unconnected. Hardware hacks (not for the faint of heart) to correct such problems in the general case: To make an AMD 3.3V CPU run in 3x mode (NOT readily reversible), simply cut pin B13 off the package. B13 will then be floating regardless of the "intentions" of the motherboard. B13 is on an internal row of pins, so this is a somewhat touchy operation. Those with 5V motherboards who are using piggy-back voltage adapters are actually better off than those with 3.3V-capable boards in this case: the "surgery" can be performed on the adaptor, with considerably less financial risk. To make an AMD 3.3V CPU run in 2x mode, tack solder some wire-wrap wire from pin A11 (Vss) to B13 (CLKMUL) of the CPU socket on the "circuit side" of the board. Be careful not to "short" any nearby pins or circuit traces. This one can be "undone", but don't change your mind too often. For the special case of the ASUS PCI/I-486SP3G and the ASUS PVI 486AP4 motherboards, there _is_ an undocumented jumper setup which will make the AMD chip run in 3x mode -- check Q15 or Q16 in this document. I'm not certain how the PVI-486SP3 handles the situation. Fortunately, the motherboard manufacturers are now catching up with AMDs eccentric pin-out, and are beginning to supply boards which handle the AMD CLKMUL pin properly. If you find yourself in the wrong mode with an _Intel_ chip, it is very likely that a board jumper is incorrectly set. ----------- Q12) What is the undocumented jumpersetting to run an early version of AMD 486DX-100 with an ASUS SP3g at triple speed (100MHz) ? Answered by Jason B. Faultless (jason@radar.demon.co.uk), Jess Johnson (jester@ccwf.cc.utexas.edu) and Doub Barlow (doug_barlow@novell.com): If you have a PCI/I-486SP3G and an AMD DX-100 then set jumper as follows: JP34 2&3 JP35 1&2 (This is an undocumented jumper setting JP36 1&2 confirmed by ASUSTEK inc CA) JP38 open The SP3G bios will not detect the chip as being a DX4 (unknown or dx-2 100 is normal), but this has no effect on function. A future BIOS update will fix this. ----------- Q13) What is the undocumented jumpersetting to run an early version of AMD 486DX4-100 with an ASUS AP4 at triple speed (100MHz) ? Answered by Rob Vugts (vugtsr@carin.ce.philips.nl) and Harm Hanemaaijer (hhanemaa@cs.ruu.nl): The jumper settings for AMD 486DX4-100 CPU on the Asus AP4 MB (rev. 1.6) are: JP17: open (These jumper settings were advised JP18: 1&2 and 4&5 by the Technical Support of ASUS) JP19: 3&4 The jumper settings for the Intel DX4 as described in the manual, can be used as well. ----------- Q14) Why does the ASUS PVI-AP4 with an 486DX4-100 only have a Norton SI of 132 instead of 198 ? Answered by Rob Vugts (vugtsr@carin.ce.philips.nl) and Harm Hanemaaijer (hhanemaa@cs.ruu.nl): This is due to the way the L2 cache timing is programmed in the BIOS. When both cache banks are filled with SRAMS (8 pcs data SRAMS) the timing of the L2 cache is programmed as follows: Cache Timing setting Normal Fast -------------------------------------------------- Cache Burst Write: 3-2-2-2 2-1-1-1 Cache Burst Read: 2-1-1-1 2-1-1-1 By setting your SRAM timing to FAST, (NORMAL is the default), cache writes are performed with zero wait states and Norton SI should report a CPU index around 198. However, there are some reliability problems with the Intel Aries chipset when cache writes are performed in 2-1-1-1 mode. In DOS there should be no problems, but when protected mode code is used, your system can become unstable. Intel errata suggested to use 3-2-2-2 write timing in stead of 2-1-1-1. Therefore, Asus has programmed the 'Normal' setting to be the default, in order to obtain a reliable system. This will decrease the performance of the DX4-100 CPU, but not as much as Norton SI suggests (Norton SI greatly exaggerates the cache timing difference!). For DX2-66 and DX-33 CPU's the decrease in performance is far less dramatic. In case only one cache bank is filled with SRAMS (4 pcs data SRAMS), setting your cache timing to FAST has no effect. Hence, with 4 SRAMS, cache writes are always performed in 3-2-2-2 mode. (That's why most mother boards are configured with only 4 SRAMS, for then no reliabilty problems can occur, even with SRAM timing set to FAST!). Another problem with having only one cache bank filled is that no interleaved cache access is possible (writing/reading one cache bank, while addressing the other one). This results in slower cache timing as well. For people who would really like to have 2-1-1-1 write timing I have the following suggestion: Fill both cache banks with either 8 x 32K 15 ns SRAMS or 8 x 64K 15ns SRAMS. Set your SRAM timing in the BIOS setup to FAST and check whether your system is stable. If not, switch back to NORMAL setting in the BIOS setup. Users with experience in reworking hardware can try the following method suggested by Intel and Asus: Workaround: Add a pull-low resister (68 ohms) to CLK2OUT signal (pin 109) of IB S82426EX chip. Rework method: On the AP4 motherboard, add 68 ohms SMT resister to the C54 location ( C54 is reserved no component mounted. It is located near the VESA slot, between the IB S82426EX chip and the MX8315 (U35) chip. Note that this may cause damage to your chipset after a long time. That's why Asus hasn't applied the rework method to the mother board! Note: That your warranty expires, even if you successfully apply the rework and that nobody else except yourself is resplonsible for any dammages ! ----------- Q15) How can I run external SCSI devices with an onboard SCSI controller ? Answered by Dominic maraglia (ezO42279@chip.ucdavis.edu) and Thomas Koschate (koschate@io.org): 1. Solultion: Buy a second SCSI controller on a plug in card. As you don't need a separate BIOS just to connect suplementat external devices, a NCR controller without BIOS will do (about 70 US$). Problematic is, that you loose another PCI slot and the high extra costs, as you allready own a SCSI controller. 2. Solution: There exist 50 pin header to micro-D 50 female adapter with the bracket. This will turn the end of your flat ribbon scsi cable into an external scsi port. It is not exactly cheap about 55 US$. Problem here is that it is nearly the same price than an entire card and that the micro-D connector is difficult to find. 3. Solution: Instead ot the micro-D you can use a 50 pin centronics and save Some money. You can either just buy the connecter, about 6 US$ or buy a ribbon cable with the connector already on it for about 14 US$. Practical advice for solution 2 and 3: If you set up your cable with the order internal-->controller-->external you can connect your turbo switch or any other switch to the scsi termination enable/disable jumper onboard (if the termination onboard is active, which is the case of the SP3g). This allows you to add and remove external scsi devices without opening the case, or buying an externai terminator. When the turbo switch is shorted scsi termination is disabled, when switch is open termination is enabled. This will save the cost for an external terminator which is about 18 US$. ----------- Q16) ASUS SP3G-floppycontroller and Floppy-Streamers ? Answered by Markus Kuhn (mskuhn@cip.informatik.uni-erlangen.de): If you experience poor performance with zour floppy-streamer, switch of ISA GAT mode in the BIOS setup. This should help, if no play around with other settings there (DMA, etc.). ----------- Q17) How to merge an NCR SDMS Kernel in a Flash BIOS file ? Answered s_stroed@ira.uka.de (Michael Stroeder) and jmayer@informatik.uni-kl.de (Joerg Mayer): If you follow this advice it is at your own risk. Be careful ! 1. Locate the offset of the SDMS Kernel. To do so search for the signature 55AAh in the BIOS image. This is the signature for a ROM extension. The kernel should be found at x000h, where x = 6,7,8,9,A,B,C,D etc. 2. Then start DOS debug and enter the following: n bios_old.bin % enter the name of the bios-file you want to patch l 7000:0 % load the file to this address n ncr.bin % the file you want to replace the old ncrbios with l 7800:0 % load this at offset 0x8000 relative to bios start n bios_new.bin % name of the new bios image rcx 0 rbx 2 w 7000:0 % write data beginning at 7000:0 q % exit debug bios_old.bin is the old binary image of your bios and bios_new.bin is the name you wish to give to the new binary image. So what this skript does is to replace the biosdata from offset 0x8000 to 0xc000 (that's 16k or 16384 bytes) with the new ncr image (please uncompress first :-) ). Explanation of the process in detail from Grant Farrell (Grant.Farrell@terc.csiro.au): The ASUS bios file (in your case awsg0302.bin) contains the ncr bios within it. This information starts at the address you found by searching for 55AAh. What you have to do is copy the new ncr bios over the old bios in awsg0302.bin. Now as awsg0302.bin is a binary file you obviously can not load it into your word processor to do this. Which is where debug comes in. What the debug script does is load into memory the current bios file (starting at 7000), then the new ncr bios (at 7800, which is the starting memory address where the ncr bios was put when you loaded awsg0302.bin into memory) and this overwrites the old ncr bios. Now you have your awg0302.bin file sitting in memory with the new ncr bios having written over the old one. Finally the debug script writes the file to disk as bios_new.bin (or whatever you want to call it) and you can then use flash.exe to reprogram your flash memory with bios_new.bin and you are in business. The reason that debug uses the address 7800 when your search for 55AAh turns up an address of 8000 is just a little querky characteristic of the way debug handles address. In debug, the address specified by the example is 7800:0. This is a form of segment addressing. This would map to physical address 78000h. Since you load the original BIOS into 7000:0 (i.e. physical address 70000h), you see that the NCR BIOS is at offset 8000h relative to the start of the Award BIOS. So as you have found your ncr bios is where it should be, all you have to do is edit the debug script (in a text editor, not a wordprocessor) and change the name of bios_old.bin to awsg0302.bin and ncr.bin to ncr307.bin, save the file and use the comamnd line debug < mergencr.txt DOS will echo all the commands on screen as they are executed so you can follow the process and at the end you should find a bios_new.bin file has been created. Have you updated your flash bios before? If you haven't, remember to change your bios jumper (if necessary on your board) and reboot without any memory managers (HIMEM.SYS etc) and to save an image of your old bios when you start up flash.exe, just in case. If anything goes wrong during the update and the program doesn't give you the OK, reprogram the bios with this saved image. You at least know this bios works. If everything goes well, turn off the computer, reset the jumper if need and get a kick out of seeing 3.07.00 come up on the scsi bios boot. Useful hint from Dirk Rogalla (dirk@kant.ping.de): Have a bootable diskette with DOS flash.exe and the new image handy As a final hint: I hope you have an eeprom programmer at hand in case some- thing goes wrong after erasing the eeprom has started and before the new data has been completely written! You won't be able to boot your computer (not even BIOS!) if anything goes wrong! ----------- Q18) What can I do when I forgot my boot-password ? Answered by Christian Mock (mc@ping.at): The password is stored in the Dallas clock chip, so I don't think a new flash rom will help -- until that one erases the cmos ram before doing anything else. I had my P54TP4 delivered with a SETUP password set -- I grabbed the Dallas clock chip data sheets by ftp (ftp.dalsemi.com), and found that there is a 12887 and a 12887A. the major difference is that the 12887A has a pin 21 which, when pulled to ground with power off, clears the cmos ram. ----------- Q19) What are the error codes of the AWARD BIOS version 4.50 ? Beep Codes: . = short beep _ = long beep . : no error, system is booting _.. : error in video card .. : error message shown on screen, no fatal error, system can be booted by pressing F1 to resume POST-Codes for AWARD-BIOS Version 4.50: To get this codes you need to plug a POST testing card with a hex display in one of the ISA slots. POST-Code | Description --------------------------------------------------------------------------- 01h | testing CPU flags 02h | testing CPU registers 03h | initialising timer, DMA and interupt-controller 04h | reading port 61h bit 4 05h | Initialising keyboard-controller 06h | reserved 07h | testing CMOS-interface 08h | testing first 64kByte of RAM 09h | initialising cache (1st time) 0Ah | loading array-table of interrupts OBh | testing CMOS-checksum or if INS-Key pressed OCh | initialising keyboard ODh | initialising video-card, identifying CPU-clockspeed OEh | testing video-memory, preparing screen to print POST-messages OFh | testing DMA-controller 0 10h | testing DMA-controller 1 11h | testing DMA page-registers 12h-13h | reserved 14h | testing timer 15h | testing interrupt-controller 0 16h | testing interrupt-controller 1 17h | searching faulty interupt-bits 18h | testing functionality of interrupt-controllers and timer 19h | looking for NMI-bits at port 61h, erasing them if OK 1Ah | printing CPU clockspeed 1Bh-1Eh | reserved 1Fh | setting EISA mode 20h | setting up EISA slots 0 21h-2Fh | setting up EISA slots 1 22h | setting up EISA slots n (n=2: 22h, n=3: 23h, ..., n=15: 2Fh) 30h | determination size of base and extended memory 31h | testing memory 32h | testing memory (EISA) 33h-3Bh | reserved 3Ch | setting up CMOS setup 3Dh | initialising mouse 3Eh | initialising cache if enabled in CMOS setup 3Fh | reserved BFh | initialising chipset with CMOS setup values 40h | virus warning 41h | initialising floppy-controller 42h | initialising harddrive-controller 43h | initialising periphal port 44h | reserved 45h | initialising coprocessor 46h-4Dh | reserved 4Eh | displaying errors if necessary boot again 4Fh | security check, asking password if enabled in CMOS setup 50h | writing back the CMOS from RAM, erasing screen 51h | enabling NMI, parity checking and cache 52h | searching for ROM in range C8000h-EFFFFh 53h | setting up system time 54h-5Fh | reserved 60h | virus security 61h | setting boot clock-speed 62h | setting up keyboard (numeric pad and keyboard rate) 63h | selecing boot device FFh | Booting BOh | fauly interrupts in protected mode B1h | NMI detected E1h-EFh | CMOS setup active (page #1: E1, page #2: E2,...) ----------- Q20) Is there a general problem with similtaneous EIDE and floppy access ? Answered by Scot Llewelyn (scotl@powerquest.com), forwarded by Brian Foddy (bkfoddy@soc.nwa.com): We have uncovered a very serious bug (we believe in hardware) that can cause severe data loss. On certain machines, a simultaneous floppy and hard disk access can cause a shift of a few bytes in the data going to/from the hard drive. It often shows up as CRC errors, sporatic system crashes, occasional corrupted partitions, and EXEs that don't seem to run reliably. All PCI EIDE systems that we have in-house have this problem, and our customers have reported similar problems on other hardware. This bug has cost us enough time that we have developed a program to reliably test for this bug. We are releasing it as freeware in the hopes that we can save a few people the headaches we've been through. Hopefully the hardware manufacturers will take notice and fix this problem. This program has been uploaded at FTP-OS2.NMSU.EDU/INCOMING as DMATEST.ZIP. FAQ-Questions: Experiences with the programm ? Which ASUS products are concerned ? Can it be fixed with a BIOS update ? ----------- Q21) How to solve Ctrl-Alt-Del Problem with the BIOS Update tr5i0107 for the ASUS TP54 Board ? Answered by Gordon Newell (gnewell@ibm.net): After updating the BIOS to tr5i0107 on the TP54 there seem to problems to make warmboot with Ctrl-Alt-Del. To avoid this you should flash the _whole_ BIOS (BIOS, boot block and ESCD). To do so choose advanced and the option 2. ----------- Q22) Is there an error in the Mainboard manual for the P55TP4XE on the jumper setting for write-protection the BIOS ? Answered by Steve Pitcher: I'm afraid I'm rather pessimistic on this point. I recently installed a P55TP4XE. The Users Guide shows JP5 as the jumper to write protect the bios. First of all, the book shows a diagram of JP5 & JP6 right next to each other. JP4 is shown in a different diagram, but is also right next to JP5. On my board, JP6 is missing. You can see where it was. They've hard wired the two contacts together. This is not a problem... except that its easy to look at the diagram of JP5&6, and then look at the MB and see two jumpers and assume they are 5&6... But on my board, these two are JP4&5, not 5&6. Look carefully at the labels on the MB. Secondly, the Users Guide suggests that by shorting pins 1&2 of JP5, the BIOS will be protected against intentional or accidental writing. I corrupted my BIOS with pins 1&2 of JP5 shorted. I then re-wrote the BIOS with 1&2 still shorted. It is my impression that shorting pins 1&2 of JP5 does NOT write protect the BIOS. The Users Guide says that shorting pins 2&3 will allow writing of BIOS. I find the BIOS to be writable in either position. Shorting pins 1&2 *does* appear to protect the Boot block and/or the ESCD. If you read the Readme files provided with V1.05 and V1.07 of the BIOS, V1.05 consists of a simpler procedure to write only the BIOS. This procedure works with pins 1&2 of JP5 shorted, that is, write protected. The longer procedure for V1.07 is supposed to write, not only the BIOS, but also the Boot block and the ESCD. This procedure to write the boot block and the ESCD fails with pins 1&2 shorted. This procedure succeeds in writing the boot block and the ESCD if pins 2&3 are shorted, as documented. So it appears that shorting pins 1&2 protects either the boot block and/or the ESCD, but *NOT* the BIOS. This has been my experience, over the past week or so. I'm interested in the experiences of others! ***************************************************************************