Afaik the P5A 1.06 won't work correctly with the mobiles 🙁

I'll go look the jumpers anyway, I might have them somewhere...

Tetrium wrote:

Afaik the P5A 1.06 won't work correctly with the mobiles 🙁

I'll go look the jumpers anyway, I might have them somewhere...

Oh yeah, that's right. Oops. They run really slowly with them?

Edit: I tried 2.0 volts for the CPU, but I see no BIOS POST.

retro games 100 wrote:

Tetrium wrote:

Afaik the P5A 1.06 won't work correctly with the mobiles 🙁

I'll go look the jumpers anyway, I might have them somewhere...

Oh yeah, that's right. Oops. They run really slowly with them?

Don't know, I never tried it as I don't own any of the 1.05 or later P5A's

Edit:
Btw, I reckon there might not be any undocumented jumper settings. There are potentially 16 possible options and it seems all are already documented! 😁

There "might" be some other options by removing a jumper but don't try that, I ruined a motherboard that way!

Edit2: I'd say just use any of the mobiles on 2.0v and at 550Mhz or so.

Edit3:On another note, iirc the newest version of that chipset could cache more memory anyway so you could use a regular K6-2?
And a regular K6-III is of course also an option.

OK, thanks a lot for the info. I notice that Vogons user 5u3 talks about the Asus P5A 1.06 and a K6+ chip, here. He says that it runs slowly. However, I can't get my mobo to POST with a K6+ chip. It works OK with a K6-2.

When it POSTed with this K6-2, I noticed that the BIOS POST message said something about a "Beta BIOS", so I guess I'm using the latest BIOS? I wonder if there's a hacked BIOS I need, for K6+ chips? I want to see this slowdown for myself, just out of curiousity! 😀

retro games 100 wrote:

OK, thanks a lot for the info. I notice that Vogons user 5u3 talks about the Asus P5A 1.06 and a K6+ chip, here. He says that it runs slowly. However, I can't get my mobo to POST with a K6+ chip. It works OK with a K6-2.

When it POSTed with this K6-2, I noticed that the BIOS POST message said something about a "Beta BIOS", so I guess I'm using the latest BIOS? I wonder if there's a hacked BIOS I need, for K6+ chips? I want to see this slowdown for myself, just out of curiousity! 😀

Perhaps your board already has that beta BIOS file? I really don't know but if you need it, try that unofficial K6+ page.

Btw, I know removing some jumpers for the ASUS TX97 series did decrease the supplied voltage, lowering it to 2.0v. I just don't know about the ASUS P5A and as I said, removing jumpers "might" damage something.

I checked the Asus website, and I do have the latest BIOS. It's Beta 1011 005. What is the unofficial K6+ website address please? I wonder if I need a BIOS upgrade, to get the K6-III+ to POST. Thanks a lot.

Here it is 😉
http://web.inter.nl.net/hcc/J.Steunebrink/k6plus.htm

But most links to the flashers are dead. I downloaded all the BIOS files earlier (in case the site ever decided to disappear) but finding out which flasher you need gave me some real headackes. Feel free to enlighten me! 😉

I have temporarily abandoned the Asus P5A 1.06 mobo. I am now using a DFI P5BV3+ rev B3 mobo. I notice on that excellent K6+ webpage, there is a link to a Beta BIOS for this mobo. Unfortunately, the link I need is dead. The table entry is:

DFI P5BV3+ 11/30/2000 BP5Bx1130.ZIP There are 3 different versions of this beta BIOS. One for each Rev (A, B, or C) of this board.

Please can you supply me with the file called BP5Bx1130.ZIP? Thanks very much if possible! 😁

retro games 100 wrote:

I have temporarily abandoned the Asus P5A 1.06 mobo. I am now using a DFI P5BV3+ rev B3 mobo. I notice on that excellent K6+ webpage, there is a link to a Beta BIOS for this mobo. Unfortunately, the link I need is dead. The table entry is:

DFI P5BV3+ 11/30/2000 BP5Bx1130.ZIP There are 3 different versions of this beta BIOS. One for each Rev (A, B, or C) of this board.

Please can you supply me with the file called BP5Bx1130.ZIP? Thanks very much if possible! 😁

Sorry but I only got the ones that were available for download on the page itself. This page is really old and the external links were already dead when I downloaded those files 🙁

That's OK. The DFI board is using a BIOS dated 01/03/2000. I think the latest is 11/30/2000. I think this BIOS file may be on driverguide. I might try to flash the board with this file...

However, I think I got the K6-III+ CPU working in the DFI board, without the need for this very latest BIOS. I ran SpeedSys. Does it look right to you? Is it using the + chip correctly? The BIOS POST says K6-III, but not K6-III+. Perhaps that is OK? Thanks a lot for any info on this.

My next job is to overclock the + chip. I want to see if it can do 600. If not, 550 is OK. Actually, the DFI manual only provides configuration info about a 5.5 multi, so it looks like I cannot test a 6 multi. The multi is configured using 3 dip-switches, and there are 8 options. They go from 2 to 5.5. This info can be seen here.

My next job is to overclock the + chip. I want to see if it can do 600. If not, 550 is OK. Actually, the DFI manual only provides configuration info about a 5.5 multi, so it looks like I cannot test a 6 multi.

Set the multiplier for 2x. All the later K6 chips (400mhz and above, IIRC) remapped the 2x multi setting to 6x.

retro games 100 wrote:

My next job is to overclock the + chip. I want to see if it can do 600. If not, 550 is OK.

Yes it can 😀 I used 120x5 and 6x100 (at 2V @ P5A). And it remains painfully slow as any other K6-2/3. Good for old stuff though 😀

Thanks very much for the info, people. I set the DFI's mobo jumpers to 2x multi, and attempt to boot it up. I see no BIOS POST on the screen, but this is not a problem - I just increase the mobo's CPU voltage from 2.0 to 2.1. Now we are in business. 😀

Everything in DOS appears to work, including Quake timedemo full screen. I get 95.9 FPS. I am using a GF3. BTW, I am using 1 stick of CL3 PC133 SDRAM. I have set all of the BIOS settings to maximum, except for CAS which is 3 (not 2). 3DBench 1.0c = 454.1. Doom timedemo full screen = 2134/542. My next job is to get Windows 98 working. I'll do that soon.

Edit: I will also try "5x & 120 = 600". That sounds like a fun challenge.

I set the BIOS CAS value to 2, even though the SDRAM I am using is CAS 3. It worked. Perhaps it works because the stick is PC-133 speed, and you can get away with this CAS speed increase, because the mobo is only running at 100 FSB? BTW, the DFI mobo has no option to increase the FSB beyond 100, which is a shame.

A lot of PC133 was officially specced for CAS3 @ 133, and CAS2 @ 100, so it's not surprising that it works.

Regarding flashing BIOS chips - If there is no suitable flash utility, or if you are unsure as to which version of the flash utility to use, I think that UniFlash is good. Also, I am now thinking about getting a Willem BIOS flashing device. If UniFlash cannot flash a BIOS chip, would the Willem unit be suitable?

With the CPU running at 600 MHz, and the BIOS timings completely maxed out, I did a clean Windows 98 installation. Everything worked. I ran VIA's 4-in-1 service pack, version 4.43, and it messed everything up. I then did another clean Windows 98 installation, and discovered that the 4-in-1 "ATAPI Vendor" option was to blame.

I ran 3DMark 2000 with a GF3 (43.x driver), and noticed a few minor graphical glitches on the screen. The benchmark then quit back to the desktop on the "adventure tests". I lowered the multi to 5.5 (from 6), and that fixed the problem. I then increased the multi back to 6, and increased the CPU's voltage to 2.2. I reran 3DMark 2000, and it worked. I've got a fairly heavy tightly fitting copper cooler attached to this K6-III+ CPU. Sandra 2002 Pro reckons the CPU temp is 28C.

One thing I have noticed is that when I am using Windows Explorer, when I browse for files, I see that some of the file icons are missing. There's just an empty space where the file icon should be. Also, I couldn't get SuperPi to run without it repeatedly crashing. This system is not completely stable @ 600 MHz inside Windows 98. DOS maybe, but not Windows.

After a short rest, I reran SuperPi, and its .exe icon was visible inside Windows Explorer, and also it worked OK. But perhaps after the system gets hotter and hotter, the system will start to become unstable again? Just after I took the screenshot below, Sandra went mad and unexpectedly quit. It's likely it did this because the OC'ing is making the system unstable, but it could be the VIA mobo + 4.43 service pack. Unfortunately, the screenshot has chopped off SuperPi's 16K to 1M test results. I think they are a bit meaningless, because of the system's instability.

Is it worth putting 2.3 volts through this K6-III+ chip?

There's definitely something wrong with Windows Explorer. If I set the multi to 4, and the CPU volts to 2 (wish I could go lower, but I can't), and the BIOS timings to normal - then if I click on Start, then Run, then expand the "Run browse" window so that it's quite large, then if I browse about inside various folders on a CD-R disc, some of the files icons are missing.

I wonder if it's the service pack which is causing this to go wrong. I uninstalled the ATAPI Vendor option. I did not install the Power Management. I installed the mini port driver thing. I will uninstall that, and see if that makes a difference. Also, I could uninstall all of the VIA 4-in-1 version 4.43 service pack components, and try 4.35 instead?

If anyone is using one of these boards, or very similar, what version of VIA's 4-in-1 service pack are you using, with Windows 98?

I think I have got this system working. I uninstalled all components of VIA's service pack version 4.43. I then installed version 4.35. It went wrong because after a reboot the desktop froze, and so I uninstalled the "ATAPI Vendor" component in safe mode. I wonder why that component isn't working on this DFI mobo? Anyway, with just the VIA AGP + mini port service pack components installed, the system appears to be normal.

I tried a crazy stress test: I ran SuperPi 1M test, and at the same time I ran Prime95, and also at the same time I ran lots of Sandra 2002 Pro tests, and then while all of this was running at the same time, I clicked on Start -> Run -> Browse, and browsed a CD-R disc. When I looked in to various folders on the disc, I could see all the icons next to the files. The system appears to be working normally.

I'll now try some overclocking again...