Here's an idea... remove GUS and see if SB16 gives you both, music and sound effects. If it doesn't, it's probably faulty.

A'ite, seems the SB EMU from the GUS drivers was hogging irq5. I've changed the SB16 irq to 10 now and it's working perfectly.
Just played a few tunes in FT2 and played a few levels of Tyrian, and sound was as flawless as such things get on this old hardware.

Great, glad you got it working 😀

I figured I'd post my Speedsys results. Don't know if they are what they should be, but I'm sure somebody would know.

I'm wondering if I'd gain any noticeable performance by getting an Intel DX4 instead.

Looks about right to me. I generally score about 38-39 in speedsys with my intel dx4-100 overdrive

it doesnt look like your l2 cache was tested. Tried cachechk?

Yeah, I was wondering about that. I got the distinct feeling that there was something funny with the memory benchmarks there, but I've done multiple runs and they all come out the same.
I dunno if I have to run my cache at WT mode, seeing as I got 64MB ram and only 256KB cache. Maybe I'll try to run the system with 32MB ram instead, as 64 is a bit overkill for a 486.
Running cachechk as I post this.

EDIT: Oops, finished just after I clicked "Submit". It looks like my L2 cache is disabled somehow, as cachechk says "This machine seems to have one cache!?" and reports only 8KB

LunarG wrote:

I figured I'd post my Speedsys results. Don't know if they are what they should be, but I'm sure somebody would know.
I'm wondering if I'd gain any noticeable performance by getting an Intel DX4 instead.

Refer to the ultimate 486 benchmark comparison.

The Cyrix DX4-100 has 8KB L1 cache, whereas the Intel DX4-100 has 16KB L1 cache.

The Intel DX4-100 scores 42.4 in Speedsys, whereas the Cyrix DX4-100 scores 40.3.

On average, the Intel DX4-100-WB is about 15% faster performance-wise than the Cyrix DX4-100.

EDIT: It looks like your L2 cache is disabled or non-functional.

Same results from Cachechk when L2 is set to WT. Does anyone have any good ideas as to what may be causing this?

Are the cache chips fake? Do they haves traces and are connected? or is it something simple like a jumper setting/bios setting or they just need to be extracted and re-seated? Have a good look at the pins and reseat them carefully.

rgart wrote:

Are the cache chips fake? Do they haves traces and are connected? or is it something simple like a jumper setting/bios setting or they just need to be extracted and re-seated? Have a good look at the pins and reseat them carefully.

Cache chips are labeled "W24512AK-15" on the top line and below 96191, aside from one that says 96121.
I've got a picture of another motherboard just line mine, which has the same type of cache chips, except they all say 96121.
Google gives me hits on datasheets for Winbond SRAM chips of the apropriate size, which makes me think that the cache chips should be alright, assuming they haven't been killed during shipping.
Would the motherboard still report the correct cache size if they chips are dead, or is that simply jumper related?

Tag chip is labeled "W24257AK-15" - 95481.

I dont trust what is reported at post. Its jumper related.

I think a good indicator of a dead cache chip is heat. If it gets hot its no good. Touch the top surface.

is L2 cache (external) enabled in bios?

Do you get errors when you run HIMEM?

Warm, but I wouldn't say hot. According to the BIOS settings it should be enabled. But I'll check it again. Who knows, I may have missed something.
Wondering about just buying some new cache chips. I assume it's alright to use a different brand, as long as they are same spec chips?
No errors from HIMEM.

Yeah I just picked up a bunch of 12, 15 and 20ns cache on ebay today for $10

theres still some on ebay from that ukrainian dude I think.

From what I have read on vogons: look for a company logo on the chips, correct font and part numbers and Im guessing anything from 10ns to 25ns will be fine ( usually the last number on the top surface)

I wouldnt rush to buy just yet. Are the jumpers set right? Did you try reseating the chips? Do you have any other cache to test? I have read on vogons sometimes with some boards you should check the traces on the motherboard because the cache sockets are not even connected.

How about this: drop a super standard cpu in that motherboard. Something from intel. A dx2-66 perhaps or anything u have laying around. Some boards really dont like cyrix and AMD. Once you do that run cachechk and see if you have L2. I have a 486 board that does all kinds of strange things when I use a 5x86 cpu (cyrix or amd)

rgart wrote:

Yeah I just picked up a bunch of 12, 15 and 20ns cache on ebay today for $10 […]
Show full quote

Yeah I just picked up a bunch of 12, 15 and 20ns cache on ebay today for $10

theres still some on ebay from that ukrainian dude I think.

From what I have read on vogons: look for a company logo on the chips, correct font and part numbers and Im guessing anything from 10ns to 25ns will be fine ( usually the last number on the top surface)

I wouldnt rush to buy just yet. Are the jumpers set right? Did you try reseating the chips? Do you have any other cache to test? I have read on vogons sometimes with some boards you should check the traces on the motherboard because the cache sockets are not even connected.

I don't have any other cache chips to test unfortunately, and as there seems to be absolutely no documentation for this motherboard online, I'm kind of in "trial and error" territory as far as jumper settings go, which is why I'm a bit apprehensive as to changing any of the jumper settings. There is a Winbond company logo on my chips, which makes me think they are genuine. I could try to reseat the chips I guess, but I don't have a dip removal tool, so I'd have to be very careful doing that. I may just have to do that though, if nothing else works.

There is nothing silk screened on the board about 256k or 512k cache? Damn

Does cachechk need to be run from a clean boot (i.e. no config.sys or autoexec.bat) to work properly?

I always run it without anything loaded and at a dos prompt. Try a couple of older cachechk revisions too.

my money is on a simple jumper problem 😀

Usually if the wrong cache jumper settings are set, you'll get HIMEM errors. Normally, when the cache is not seated properly, you'll also get HIMEM errors. Either the cache is fake (though it looks genuine), or it is somehow disabled, via a jumper, or in the BIOS. I see what looks like a 256 KB for cache during POST, this usually implies that the jumpers are set correctly. If you've confirmed that there are indeed traces leading from the cache sockets to at least somewhere on the board, then look for a setting to enable/disable L2 cache. This is usually found in the BIOS settings.

From what I can guess, you're passing HIMEM because the cache is disabled or non-existent.

All sort of weird problems can arise, anyhow... for example, I upgraded a Shuttle HOT-419 to 512 kb single-banked cache, jumpered it accordingly, yet the ONLY program that sees and can test the L2 cache is Speedsys. CTCM, Cachechk and the likes ALL report NO cache present...