TELVM

Well there is always some potential fiddling remaining , but the 98 is now running fairly well in the dino . I installed XP Pro SP3 in the second partition and that went very smoothly, without any problems neither with the USB 2.0 card nor with a recently inserted RTL8029AS Ethernet card. Working now on the internet connection. I'm surprised how decently the old PII Deschutes 350/100 @434/124 with 768MB RAM moves XP ; I was expecting a barely bearable slowness at best, but it runs reasonably well with not too much disk scratching . To flog the old horse a bit more I tried this ancient trick to increase voltage in Slot 1 CPUs. It worked, raising from 2.0 to 2.2V ... ... but although the dino POSTs and boots at 466/133, both 98 & XP freeze on desktop. Time for a PIII Coppermine closer to 1GHz I think . BTW, I'm searching for some game recording software (à la FRAPS) that works under Win98, without success thus far. ¿Where can I find it?

Shrinking both vcache and ramdisk size didn't make any difference, dino keeps freezing on boot as before. Seems system arena exhaustion is not the problem. I'm exploring another ramdisk software, Cenatek Ramdisk9X/Me v1.5: Has some nice options, formats in FAT32 with 512K clusters, and works seamless thus far with 128MB of ramdisk and 256MB of vcache. I like this software, it's very much like its Dataram descendant which I use in modern comps with Win7. Son unless you tell me that xmsdsk.exe is decisively faster, I think I'll stick with the Cenatek. Keeping in mind that, in this hardware context, 'fastness' is a very relative concept :

Used your config.sys and modded the msdos.sys, got to DOS prompt, did a DIR OK. Then a couple of seconds after entering WIN command, it just froze with no message at all .

@ Jaclaz: Thanks a lot for the link to Letter Assigner, it's exactly what was needed and works like a charm . Also for the grammar lesson . @M()zart: Thanks for your advice, USB 2.0 runs fine now in the dino and pendrives are recognized instantly. Seems it was some IRQ/drivers conflict, now solved.

Sure, here they go: Dino System Files Got another little drawback, I made two FAT32 partitions in the HDD, the first 32GB one for Win98SE and a second 80GB one for future WinXP. Problem is, these partitions are now named C: and D:, and the CD/DVD driver was displaced to letter E: . That's not good, some old games only work fine if CD driver is D: . There is any way to switch letters so that 80GB HDD partition is E:, and CD/DVD is D: again? Were that not possible, can we 'hide' or 'deactivate' somehow the 80GB partition for Win98SE eyes only, so that ignores it, doesn't assign letter, and gives letter D: directly to the CD/DVD? Dino System Files.zip

No joy man, system keeps crashing on startup. However something bizarre: Starting up step by step I observe the line with xmsdsk.exe passes OK; then, some lines later, the crashing comes after some mouse related line . But if I REMark the ramdisk line, comp starts OK and mouse works OK . Could it be that XMSDSK.EXE is conflicting somehow with the mouse?

I'm now having some trouble trying to create a 128MB ramdisk using XMSDSK.EXE . I put both the HIMEMX.EXE and XMSDSK.EXE files in C:\Windows , and added these two lines to the top of CONFIG.SYS: DEVICE=C:\WINDOWS\HIMEMX.EXE INSTALL=C:\WINDOWS\XMSDSK.EXE 131172 K: /C1 /T /Y But for some reason the comp just freezes while initializing it . (If I REMark the ramdisk line, HIMEMX.EXE loads OK) And now gentlemen time for some laughs . A bit of retro benchmarking, just for fun: Behold my highly sophisticated astronautics-grade PII cooling solution : Works like a charm however, keeping the Deschutes 350/100 overclocked to 434/124 quite cool:

At last! Halleluyah! Got USB 2.0 pendrives working with the dino! Out of desperation I started changing many things almost aleatorially, and by pure serendipity this is the winner combination: - USB 2.0 PCI card moved to slot 3 (center one of the five). - Installed Spanish version of NUSB 3.3, then installed VIA drivers for VT6202 USB 2.0 PCI Card. - Under Device Manager - USB, disabled Universal Host Controller Intel 82371AB/EB PCI to USB. - In Device Manager - System Devices - PCI Bus - IRQ Control ... ... leave only 'Get IRQ Table using MS specification table' ticked. - BIOS · ACPI: Enabled · Power Control by APM: Yes · PNP OS Installed: Yes · Resources Controlled by: Auto · Reset Configuration Data: Disabled · Assign IRQ for VGA: Yes · Assign IRQ for USB: Yes · PCI Latency Timer: 32 PCI Clocks · Slots 1 to 4 Use IRQ: Auto · USB Keyboard Support: Disabled

Sadly it was not possible to install Win98SE with 'setup /p i' to disable ACPI in this comp. After entering the command installation began, but always froze and hung some seconds later. I had to reinstall with ACPI enabled. If the problem is conflict from ACPI and USB controllers sharing the same IRQs, how about disabling some non-used devices (joystick, COM, printer ports) to free IRQs, then reassign so that no IRQ is shared by ACPI and USB?

Thanks a lot for your help and patience Dencorso. The system went too FUBAR , better to start afresh. I'm reinstalling as per your last instructions.

Well the system is screwed up, big time. I'm going to reinstall Win98 trying disabling ACPI during installation with setup /c i setup /p i .

Roger Wilco . But must I mark hexa or decimal before entering the '2', i.e. :

Tried as you said. Found several USB 'ghosts' when in safe-mode, then deleted everything under 'USB controllers' in device manager. Then disabled 'Assign IRQ for USB' in BIOS. On restart Windows, asking for Win98 installation CD, reinstalled everything again as in the picture I posted above: But when I inserted a pendrive its light began flashing AT LAST! Then came a message 'new hardware found' and I let Windows search for the most suitable driver for the new device, installed it, and finally the pendrive was recognized, and it appears as 'E: extractable disk' in Windows Explorer . HOWEVERRRRR ... the instant I click on E: , Windows frozes and hungs . No BOD, but I have to reset. May be I could try the procedure again, and this time cancelling the installation of any USB drivers, except the VIA add-in USB2.0 ones (which I think are the last two lines under 'USB Controllers' in the picture)? Just to be sure: Hexa, decimal, or doesn't matter? Sorry Dencorso, I'm not sure exactly which following settings are you referring to?

Anyone at hand. Done done done and no change in device manager, same exclamation mark problem and pendrives not recognized. Both mouse and keyboard are PS/2 in the dino. The closest thing to 'disable onboard USB controller' I can find in BIOS is 'USB Keyboard Support', which is set to 'Disabled'. Perhaps disabling some of the items under 'USB controllers' in the device manager? I mean this style:

Thanks a lot for that info Dencorso , we could be in the right path now, I installed those drivers and device manager no longer shows a 'question mark' error: Concentrador de raiz USB = USB Root Hub Controladora de host universal = Universal Host Controller I suspect there are too many things in there, should I delete some items ? However when I insert a USB flash drive in the VIA VT6202 PCI USB card, yet another 'USB Root Hub' appears, with an exclamation mark error : I right click properties on the exclamation error and it says: Which I think translates as: 'Due to hub transfer request failure, a hub has been disabled. Check the cables and devices in your computer and then restart '. Seems we are almost there but there is still some bug, any ideas?

No joy either with the 'russian' drivers . Now that you mention it, I've got this thing from the catacombs: However no drivers, no idea of brand or model, and it just generates an 'interrogation problem' in device manager when I insert it in the board. Thanks jumper, I disabled memory hole and enabled auto detect RAM clock. But enabling spread spectrum could be a bad idea, several people tell me it can bring more trouble than peace . Searching the web I found at last some good drivers for the Riva TNT2 M64: 1999-vintage Detonator 2.08s (the 'legacy' drivers that nVidia's website still offers didn't work good at all). These old drivers work like a charm . I also found an ancient version of CPU-Z that gives CPU frequency with decimals, i.e. 434.3 MHz .

Thanks Dencorso, I disabled that prior to flashing BIOS following advice on the procedure elsewhere in the net. The board now 'can see' the full 120GB in the HDD, I'll try the Ranish utility. Now that you mention it, as a matter of fact I started the resurrection with a Coppermine 733/133 ... which unfortunately was mortally wounded when the old & cheap PSU that came with the dino went up in smoke the first time I plugged to the wall . Luckily the mobo and the RAM survived the holocaust, but the PIII was KIA . I kicked off into jovian orbit the ******* saboteur PSU, installed a decent one, and inserted the nex best vintage CPU at hand. The Deschutes 350/100 mounts the L2 cache externally at half speed, and lacks SSE, but I'm surprised how decently this old soldier performs . It even overclocks stably to 426/122 without fiddling with the voltage . Better fossils are coming, including a couple slotckets to expand the potential CPU possibilities . None of several 2.0 flash drives I've tried is recognized, so I think the problem, whatever it is, comes from the dino . I've tried formatting the flash drives with FAT32 and FAT16 and no joy. I'm aware Win98 can't understand NTFS. My russian is very limited and Google translator refuses that page, could that be USB drivers for Win98?

Thanks a lot for helping LoneCrusader. However after installing NUSB 3.3 the dino still refuses to recognize any of my thumb flash drives . May the problem be that all the flash drives I try are USB 2.0, and this board can only work with USB 1.0? Could it help somehow formatting the flash drives in FAT16? On the bright side of life I managed to reflash the Award BIOS without triggering the revolver , so if it works as advertised the board should now be able to exploit the full 120GB in the HDD. But now comes the second hurdle: I understand that Win98SE's fdisk.exe is limited to 64GB. Do you recommend using uncle Bill's solution? The HDD is currently jumpered to limit it to 32GB, and Win98SE is installed in a 32GB FAT32 partition in C: . Can I now just de-jumper the HDD and create a new partition (say E:) with improved fdisk or another tool, on the 'rediscovered' 88 GB? Or would it be better to re-format completely the HDD and reinstall Win98SE? By the way, I've captured some images of the BIOS. I haven't the foggiest on how some parameters should be set , so if you see some absurd or catastrophic setting please advice.

Thanks for helping, but first of all there is a surreal problem . The dinosaur can't read from USB thumb drives, seems some USB drivers need to be installed previously. As it is not connected to internet, and no other comp at hand has floppy disk drive, I was planning to feed it files from USB flash drives. Guess the mobo came originally with some CD including USB drivers, but now it is MIA. Perhaps these Intel drivers for 440BX chipset could do the job? I could download them in other comp, burn to a compact disk, and feed the dino from the CD driver. Play old DOS and Win9x era games mainly, then perhaps some light internet browsing (mail, forums, wikipedia, stuff like that; no movies or heavy duty).

Patched BIOS' for older PC's to overcome HDD limitations. If the MoBo is one of them, you can use the patched. I have on a couple and they work fine. Mind you, always be safe and back up your original. Thanks submix8c, that's exactly what was needed , I started a new thread here.

Any help in making this dinosaur function properly will be appreciated . In those last years of the XX century when this machine was born I knew little about software, and very little about hardware, so this is gonna be a challenging archaeo-compu-logical exercise for me . I'll use this PC initially to play DOS and Win9x games, but later I'd like to add some more serious function, maybe with WinXP on another partition (although I expect people in this forum to convince me to remain in Win98 ). The hardware - Mobo: Tekram P6BX-An (chipset Intel 440BX) - CPU: Pentium II Deschutes 350MHz/FSB100 - RAM: 768 MB (3x256 SDRAM 133MHz) - GPU: Riva TNT2 64 32MB AGP - HDD: Seagate Barracuda 7200.7 120GB - 1x DVD drive - 1x 3.5" diskette drive The SO: Windows 98SE (spanish language version) After many years peacefully sleeping in hibernation, the venerable machine is again alive & looking for trouble : And the first problem is ... 32GB HDD limit. To fully use the 120GB HDD (currently jumpered to 32GB) I understand there are two hurdles in the road: 1) Award BIOS pre-summer '99 bug (32GB) 2) Fdisk.exe limit (64GB) I've got a fresher v1.08 BIOS version for up to 128GB HDD. I've also got some different versions of the AWDFLASH.EXE utility. The maneouver is explained at the end of the mobo manual, here: Mobo Tekram P6BX-An MANUAL.pdf I remember that in those days flashing BIOS was a bit like playing russian roulette , so if anyone more experienced could recommend the fail-safest procedure, which version of AWDFLASH.EXE would be best for this mobo, etc. may be I don't crash jumping the first hurdle .

Wow, thanks a lot Dencorso ! How easily one feels at home and amongst friends in this forum . So let it be written the title of the thread, I'll start it once I get the archaeological jewel and precisely catalog its innards. We'll talk again on the mobo manual then. In the meantime I'll have a look at the programs you recommend and will keep exploring this forum, tons of great info here around . BTW, a sister of mine is currently living in Rio (Ipanema), could it be you are neighbours?

Thanks Fredledingue. I'm well aware of the advantages of ramdisks, also been using them for years. And I agree with you, its main handicap (volatility) we can turn into virtue by using the ramdisk as automatic and merciless garbage disintegrator every shutdown. In addition to temp files, browser cache and other stuff, I also move the entire Firefox profile into ramdisk. This speeds up even more the browsing experience, and spares the SSD a lot of writing in the modern comp. Seems it is possible to run Win98 off a ramdisk, this looks like a success story: My link Thanks for the good advice Dencorso, I'll follow it and try XMSDSK 1.9I . There exists any early version of Crystaldiskmark, or similar benchmarking stuff, that runs in Win98? There is another thing I would like to comment with you, if you don't mind. Next weekend I'm taking posession of a 1999-vintage comp I'll try to restore running Win98SE. It's going to be a lot of fun, but also I know I'm very rusty with both the hardware and the software of those times. There are some hurdles I've foreseen (32GB HDD limit, 768GB of RAM) and surely there will be more. So I was thinking of opening a thread on the subject here in MSFN-Win95/98/Me, where maybe you the much more experienced people could give advice one step at a time, if that wouldn't be asking too much. Would it be appropriate to start such a thread? If so, where exactly? Maybe in the '9x member projects' section?

Thanks gentlemen. It has been over a decade since last time I fiddled with Win98 and I'm extremely rusty, if I ask something particularly obvious/stupid please indulge me . So for Win98 we've got ... XMS ramdisks: - Ramdrive.sys - XMSDSK 1.9I - Srdisk ? Cenatek Win9xMe 1.5 ? Non-XMS ramdisks: - RLOEW's Anything else? Maybe some old version of the Gavotte could work in Win98? What I have in mind is testing to select a fast ramdisk for speeding up a bit a 1999-vintage computer I'm restoring.

(Thought best to push this old thread that starting a new one) How about the Cenatek RAMDisk9xME 1.5? Any experience with this in W98? http://www.brothersoft.com/ramdisk9xme-82623.html I use its Dataram descendant in W7 and it works like a charm. If we got less than 1GB of RAM (768MB in my case), there are any advantages or disadvantages in using a XMS or non-XMS ramdisk with W98?