K6-2+, K6-3+ support for BCM IN5598 w/ added functionality

Don't ask how to over-clock.
Uranium235
BIOS Newbie
Posts: 28
Joined: Thu Jul 29, 2004 6:26 am
Location: Lancaster, PA

KachiWachi, this is where I got the 50ns EDO:
http://cgi.ebay.com/BRAND-NEW-64MB-EDO- ... dZViewItem

I'm sure they have more available if your interested. :)
Last edited by Uranium235 on Sun Nov 13, 2005 8:39 am, edited 2 times in total.
KachiWachi
The New Guy
Posts: 1451
Joined: Fri Mar 29, 2002 10:32 pm
Location: Pennsylvania, USA

I would need 32MB SIMMS though... *sigh*
Uranium235
BIOS Newbie
Posts: 28
Joined: Thu Jul 29, 2004 6:26 am
Location: Lancaster, PA

Here’s another update and I should have nearly everything sorted out now:
1) I had a failing motherboard to begin with. It is dead now. I believe the board was tired and had seen nearly 6 or 7 years of continuous use. A powersupply attached to it had failed over a year ago possibly sending a deadly voltage spike through the board when a capacitor inside exploded. The board continued to work afterwards and the incident was forgotten.
I was able to acquire a brand new OEM replacement board to continue testing the K6-2+ 450 and the patched BIOS. It made sense to get this board after acquiring the 50ns EDO. I had debated on getting a Super7 board but that would have required PC100 that I didn't have.

2)My original board came with a NEC/Packard Bell BIOS that I believe had the bootblock write-protected. Although the patched BIOS flashed successfully, I don't think all the BIOS settings were fully functional. Primarily some of the memory timings that were unhidden did not work.
The new board I acquired had a very early Packard Bell version 1.00 BIOS that was not write protected. The patched BIOS was able to completely overwrite this early Packard Bell BIOS making all BIOS settings fully functional. I was able to hot flash the old BIOS chip, which I keep as a spare.

3)The Clear CMOS jumper didn't work properly on either board. I've found that to properly clear the CMOS I needed to also remove the battery and disconnect the power from the motherboard in addition to moving the Clear CMOS jumper for maybe 10 minutes.

4)The ACPI IRQ routing bug that I experienced is not present in the patched BIOS after all. This bug appears to have been carried over from the former BIOS possibly due to the inability to overwrite the bootblock. In fact, the patched BIOS doesn't even appear to implement ACPI which I don't have any problem with.

5)The patched BIOS still reports my hardrive as UDMA-5 even with the 40-wire IDE cable but it operates completely stable after switching cables with UDMA enabled.

6)The new board still doesn't like to run the L3 or external cache at the 66 or 75 FSB with the K6-2+ but after some testing I've found this board is significantly faster with the K6-2+ even without the motherboards cache enabled. Performance seems to be on par with a Super7 board. 400(6x66) seems to be the most stable clock for this K6-2+ 450 and this motherboard. Maybe a different CPU would perform better than the one I have. I thought about trying a low voltage K6-IIIE or underclocking a K6-2+ 533 to see if they would run at 450 more stable.
I’ve also tried a regular K6-2 at 450(6x75) with this patched BIOS and it performs very well with the tightest memory timings.
I'm a little suspicious of the CPU voltages put out by this board. I think this board may be overvolting but I can't be sure. I'm considering getting the Processor Protector here:
http://www.autotime.com/pprotect.html
or just testing the CPU socket with DVOM to confirm the voltages.

7)I’ve also been testing Windows XP with SP2 on this board and they work fine together. I just needed to download the newer SiS XP UDMA IDE204a drivers to be able to enable UDMA2 for the IDE controllers.

8)There appears to be an even newer Packard Bell BIOS available for this motherboard than I previously thought. It is version 1.09 dated 11/15/1999 and appears to have hardrives supported up to 65GB. It lacks the memory settings available in the patched BIOS unfortunately and still doesn’t support Write Allocation.
It can be downloaded here if anyone is interested:
http://members.driverguide.com/driver/d ... rid=117568
I believe this was the last BIOS ever put out for this board by anyone. I’d be interested to know if there was another newer BIOS out there.

I'm going to park this machine where it's at for now and may try testing some more things out with it at a later time.
Last edited by Uranium235 on Tue Mar 08, 2005 4:36 pm, edited 1 time in total.
Uranium235
BIOS Newbie
Posts: 28
Joined: Thu Jul 29, 2004 6:26 am
Location: Lancaster, PA

Good news for IN5598/PB850 users... Complete success with a 1.7volt K6-III+ CPU!! This is the CPU this motherboard needs.
The CPU voltage can be set as low as 1.8volts so .1volts really won't matter.
This low power K6-IIIE+ can do 500mhz(6x83) stable without breaking a sweat even with the board's L3 cache enabled. :D

BTW, BIOSMAN has posted the patched BIOS to support the K6+'s on his website.
Here:
http://web.inter.nl.net/hcc/J.Steunebrink/k6plus.htm
Last edited by Uranium235 on Mon Nov 14, 2005 4:27 am, edited 1 time in total.
Uranium235
BIOS Newbie
Posts: 28
Joined: Thu Jul 29, 2004 6:26 am
Location: Lancaster, PA

I was going to let this thread die but I've been asked to follow up with some additional information.
The IN5598/PB850 is still running strong with the 1.7volt K6-3+ at 500(6x83). The great thing about the SiS5598 chipset is that PCI bus can be run async of the FSB, so the 83 FSB can be used with complete system stability. Just set jumper JP18 to bridge pins 1-2 for PCI ASYNC at the 83 FSB.

Here is the undocumented setting for the 83 FSB:
JP14=on, JP15=off, JP16=off

There are undocumented CPU voltage settings also:
1.9volts; JP8=close, JP9=open, JP10=open, JP11=open
2.0volts; JP8=open, JP9=close, JP10=open, JP11=open
2.1volts; JP8=close, JP9=close, JP10=open, JP11=open
2.3volts; JP8=close, JP9=open, JP10=close, JP11=open
2.4volts; JP8=open, JP9=close, JP10=close, JP11=open


I recommend disabling the integrated video, which sucks anyway, and using a PCI video card. I use a 64MB Geforce MX4000. I also get better performance disabling the integrated ISA ES1869 audio and use a PCI Soundblaster Live Value.

I need to note that the 2.0volt K6-2+ 450 I originally tested on this board completely died on another motherboard, a VA-503+, and had never worked right on the other board either. I acquired that CPU used and it could have been failing all along screwing up my testing. Therefore, I cannot say that a 2.0volt mobile AMD K6+ won't work on this board, but only that this board has not been fully tested with a known good 2.0volt K6+ CPU.

I believe the voltage regulators are weak on this board and that someone may have better success with the lower voltage embedded K6+ CPU's. One indication of the weak voltage regulators is random reboots sometimes after high CPU utilization. I haven't experienced reboots with the 1.7volt K6-3+ but have experienced the reboots with higher speed K6-2's.

I would be very interested to know if anyone else has had success running a K6+ CPU on this motherboard with Jan Steunebrink's patched BIOS. Please post your results.

EDIT: One good thing I should note is this motherboard seems to have no compatibility problems with any PCI video card I've tested on it. This includes MX4000, MX400, Radeon 9000 and Radeon 9100.
Oldskoolmaniac
New visitors - please read the rules.
Posts: 3
Joined: Sat Feb 04, 2017 3:36 am

Sorry for digging up an old thread, but really need to now how did you jumper this board for 6x multiplier I can only get 5.5x83 and how do you enable the l2 cache for the k6 III+, is the external cpu cache setting in the bios?
Denniss
BIOS Guru
Posts: 3153
Joined: Thu Mar 21, 2002 8:16 pm
Location: Near Hannover (CEBIT) Germany
Contact:

AFAIR these CPUs understand a 2x multiplier set by jumper as 6x multiplier
Oldskoolmaniac
New visitors - please read the rules.
Posts: 3
Joined: Sat Feb 04, 2017 3:36 am

I got to boot up my k6 3+ 450Mhz its overclocked 6x83=500MHz but when ever external cache is enable it locks up before booting devices and memory count repeats 3 times before it continues through the rest of the setup.
edwin
The Hardware Archivist
Posts: 6286
Joined: Wed Mar 20, 2002 7:11 pm
Location: Netherlands
Contact:

does it work at its rated 450MHz?
edwin/evasive

Do not assume anything

System error, strike any user to continue...
Oldskoolmaniac
New visitors - please read the rules.
Posts: 3
Joined: Sat Feb 04, 2017 3:36 am

yep even when its a its stock 450 it does not boot to windows when external cache is enabled
Post Reply