The original ver 3.0 is available here:

http://web.inter.nl.net/hcc/J.Steunebrink/k6plus.htm#bios

Thanks! - been searching for hours. Apparently I suck at googleing.

As soon as I finish setting it up, I'll run my usual batch of tests on it. If anyone is interested in performance numbers (and other random details) ask, and I'll post them here (together with screenshots). I'll put it up against a Baby-AT VIA MVP3 board made by Lucky Tech (witch is quite stable and decently fast, but doesn't have official K6-III and K6-2+ support). I will also test DOS performance on them since most DOS games seem to run great on fast K6-2 CPUs. Tried Supaplex and Golden AXE - there is no perceived speed up or stuttering using 450 and 500MHz K6-2 CPU's toghether with a Matrox G100, 3DFX Voodoo Banshee, Voodoo 3 3500 TV, Riva Vanta or Riva 128zx. Anything faster then these will cause stuttering in Supaplex and a weird scroll bug in Warcraft and Warcraft 2 under DOS.

Might want to consider a K6+ chip if speed becomes a problem as you can change the multiplier on the fly using the setmul tool.

I did - unfortunately I only own K6-2's and a K6-III. I bought a K6 laptop last week, believing it has a K6-2+ cpu inside, but to my surprise in had some wierd 475MHz? low power CPU witch supports neither on-board L2 cache or on the fly multiplier switching. I found another broken K6 laptop and ordered it - it's a toshiba with a 500MHz CPU. Theoretically that should be a K6-2+.

The thing is, the Ali board is currently having issues - random Rundll32 crashes in windows - still have to sort that out before I begin testing.

UPDATE: Could not make the board run stable so I caved and flashed the bios you linked above. Worked like a charm, and now the MB correctly detects larger drives. Currently installing win98se on a 40gb maxtor drive. Bios update also added some new options.

You should consider replacing the larger 5 green electrolytic capacitors... Old capacitors aren't usually a problem on older boards as long as they were good quality to begin with. The Asus P5A is a good example. Asus used Rubycon/Sanyo parts on that board and they hold up well over the years. Your board seems to have some poor quality caps.

all the caps on the mb look good. None are leaking / popped / swollen. Right now I underclocked the CPU and swapped out the ram. If this doesn't work, I'll pull the board and replace all caps + VRMs. One of the VRMs gets a little hot, but not so hot that you can't touch it, so it should be fine. Will update later.

Windows throws "illegal operation" errors and / or BSODs when updating registry / system settings during setup with the 40gb HDD, but not with the CF card. It happens randomly - no set point when it will error. When using the CF to IDE adapter, it installs windows OK but hangs during any driver install.

Update: got it thought setup by downclocking the CPU. Now it hangs after loading the desktop icons and start bar. One or more of the VRMs and/or filter caps is faulty. Will look it over and replace defective components tomorrow afternoon.

Sometimes capacitors bloat, sometimes they dry up. When the electrolyte dries up, it can look completely normal on the outside. The only way to know for sure is to take it out of circuit and measure it. And even if it meets its spec for capacitance, it must still be tested for ESR. Your caps say "Low ESR" on them, so it seems to me that the board requires low ESR caps. I'd reckon they're way out of spec on ESR given their brand and age.

Which PSU are you using? It seems to me your PSU is also suspect based on the symptoms you're describing.

I'm using a run-of-the-mill 200W AT PSU made by Linkworld. It worked fine with other systems it was in. I've swaped the caps and one of the VRMs. No dice - it now hangs when loading windows.

Guess it's time to swap the PSU.

Below are some pictures of the cap swap. Replaced the board's caps with two 6.3v 1000uf panasonic caps and two 6.3v 1500uf rubicons. 3 of the 4 caps I replaced tested good. The 4th one was within tolerance.

What about the last cap near the ISA slot? I think it's part of the VRM. Looks like the two tall caps are Panasonic FR series and the two short ones are Panasonic FC. FR might qualify as low ESR, but FC is closer to a general purpose capacitor in terms of ESR. It might be adequate here, but if you can get a brand and series off the old cap, you can reference the specs in their datasheet.

Also, when you say "tested good", was the ESR within spec?

Linkworld uses Fuhjyyu brand capacitors (Garbage, and almost certainly out of spec in a PSU of that age). I'd be curious to see a picture of the innards.

I only tested capacity, I don't have equipment to test ESR interference... The one near the ISA slots I wapped when I first got my hands on the board - it was bulging. I think it's a samxon.

I'll test with a new PSU and post pictures of the Linkworld's innards later tonight.

UPDATE:

OK - so i swapped PSU's with a known working quality Seventeam 250W AT unit - same issue.

After doing some research on socket 7 power circuitry, I replaced (again) the capacitor next to the ISA slots with a 1500uf cap over the 1000uf one that the board originally came with, and ADDED a 3rd 1000uf capacitor (a panasonic one) next to the two green samxons under the CPU socket.

As you can see in the first two pictures, the board doesn't come with a 3rd cap and there was no evidence that a 3rd capacitor was ever installed there. I decided to add it, and replace the 1k uf ISA one with a 1,5k uf one after looking at pictures of similar boards by other manufacturers - Lucky Tech Ali Alladin baby AT board for example - witch has a very similar layout but uses more capacitors.

.

Good job...

Samxon has reliable series such as GT, RS, GD, GC, GA, but the green Samxon caps are probably "GF" series, which are a defective series and should be replaced on sight.

Adding the third Panasonic cap probably picked up the slack for the two failed Samxon caps, but because it's now tasked with excess ripple, it might get stressed. What I would do is to go ahead and replace both of them with two of the same Panasonic FC 1500uF caps.

I do have 3 new panasonic 1,5k uf caps at my disposal - the long ones. Do you think it's worth replacing those tree caps (2 green samxons adn 1 black panasonic 1000uf) with 1500uf panasonic caps? Otherwise I'll have to order new good low ESR 1000uf ones on monday...

Can you confirm the series of the Samxon caps? It is printed on the cap.

Are the three 1500uF Panasonic capacitors "FR" series? I would definitely replace the green Samxon caps. My only concern is that the repetitive desoldering and soldering might weaken the pads and something might go wrong. So be careful.

Panasonic caps are indeed FR series. It's written on them. The green caps are GF series.

Definitely get rid of the Samxon GF.

FR is a much better replacement than FC here, becaue the original caps said "Low ESR" on them, and FR is suited for this here. But the short FC you already put in may be perfectly adequate here becaue we're not talking about a very demanding VRM here like on Pentium 4. Heck, this isn't even Pentium 3, this is only SS7, so as long as the caps match or exceed the spec of the originals, you should be fine.

Ok, after a week of testing I figured out two major issues with the board:

1. CPU voltage settings printed on the PCB don't correspond to actual voltage, or those in the MB's manual. I've been running my CPU's at either 2.5v or 1.9v. Board won't post at 2.1 and it's unstable at 2,2v and 2.3v so 2.4v is the closest I can run a K6-2 at. Settings in the manual don't correspond to actual Vcore either (measured with a multimeter at the back of the socket).

2. While it's nice and stable at 2.4V, even with my K6-2 450 OCed, windows will BSOD as soon as I install vga drivers with ANY agp card regardless of bios settings and driver version (yes I did install the ALi AGP chipset drivers first). Running any PCI card + Voodoo 2 yields good stability and performance, but I wanted to test out the AGP implementation.

I think the board is either CRAP or I have no ideea how to configure it correctly.

What os are u using?

Often with Super Socket 7 boards I treat the AGP port as a 66 MHz PCI slot and thus I do not install any AGP drivers.
I imagine that this saves some headache in many cases.