VOGONS

Search

Common searches

motherboard has issues with sound-cards / bus speed confusion?

Topic actions

First post, by Yuuker

Posted on 
User metadata
Rank Newbie
Rank
Newbie

I recently aquired a Zida 4DPS / Tomato board and have been having a peculiar issue with it. It seems all my sound cards have been producing garbage noise (or no sound at all) with any cpu going faster then 66mhz. The two sound cards i have on hand are a PAS16-LMSI and a Generic YMF7-OPL3SAX card.

Prior to my testing, both cpus have been inccorectly inserted (a cringe-worthy mistake that i seriously regret) and powered on a few times briefly on accident.

They seemed to work fine after that but i never got to testing the sound prior and this garbage / distoreted sound ONLY happens when i use those cpus. The minute i put it my dx2-66 *OR* disable cache to slow things down the sound cards magically work as intended.

Here's something i partially noticed. My Zida board doesn't have any jumpers for cpu speed. It only has jumpers for bus speed, and while the only online manual i could easily stumble across lists diffrent jumper configurations for diffrent bus-speeds, trying to set them to match that of the cpu's as listed in the manual, results in an inncorect cpu-main speed being detected. Its only if i close -ALL- the bus-speed jumpers that the motherboard seems to...auto detect the right setting for the cpu? Eitherway, doing it that way will get it to list the cpu at its proper speed (and i assume the isa/pci bus too).

So maybe its a isa-bus issue? I tried playing with the iSA-Bus-Clock option in the bios, but it did nothing. Although to be fair, i don't fully know how it all works.

So now i'm at a stand point where either the cpu's got scrambled, the motherboard got partially fried. Its a bus-issue im overlooking the fix for, Or maybe it was just like that all along / my cards happen to be speed sensitive.

Last edited by Yuuker on 2018-03-23, 23:00. Edited 2 times in total.

Reply 1 of 5, by dionb

Posted on 
User metadata
Rank l33t++
Rank
l33t++

This sounds a lot like a bus speed issue, which means that those "CPU speed" jumpers are also doing all kinds of bus stuff. It also means that somewhere between the documentation you found and what you're doing on the board, something is working differently. At a guess, I'd say it's deriving the ISA clock from the PCI bus or CPU host clock with a divider, and that that divider should be 4 (33/4=8.14MHz) but instead it's maybe on 3 (intended for 25MHz, but at 33MHz that gives you 11MHz) or lower.

As for how to fix... well, if documentation fails, you're left with trial and error. Change the jumper settings one by one, and look for a setting that does *not* seem to change the CPU speed. Once you've found that, test the sound card. Or better, find a diagnostic tool that shows you ISA bus sped.

Reply 3 of 5, by Yuuker

Posted on 
User metadata
Rank Newbie
Rank
Newbie
dionb wrote:

This sounds a lot like a bus speed issue

I couldn't find an isa-bus speed diagnostics program. Although if anyone knows one, please send it my way 😎

The only options i have for ISA-Bus-Clock are:
1/4 PCLK, 1/3 PCLK, and 7.159MHZ (<--- this makes things worse)

Playing around with the only related jumpers just produced the motherboard displaying an inccorect cpu-main speed, unless i had ALL the bus jumpers covered (even if the only manual i can find shows diffrent bus jumper layouts for diffrent speeds)

Iv compiled a mini chart of the results i have so far with my 5x86 at 1/4 PCLK (Maybe i should try down-clocking it to 100mhz?)
===============
Duke 3D Setup: VERY bad garbage noise
Alien Carnage: Garbage noise and music plays strangely. -SOMETIMES seems to fix itself partially once in game.
Jazz Jackrabbit: Like Alien-carnage.
Inertia Player: Works fine.
Duke 3D in-game: Some distorted sound. -Seems to mostly fix itself once a level has loaded.
Doom 1+2: Like duke 3d, -ocassionaly produced distorted sound.
ROTT: Like duke 3d, -ocassionaly produced distorted sound.
=================
Very mixed results. Iv included a picture of my current ISA-bus setup in the bios.
Once again, disabling cache to slow the cpu got everything working 100%.

Attachments

  • biosmenu.jpg
    Filename
    biosmenu.jpg
    File size
    511.91 KiB
    Views
    605 views
    File license
    Fair use/fair dealing exception

Reply 4 of 5, by SarahWalker

Posted on 
User metadata
Rank Member
Rank
Member

It could be that the faster CPUs are using a write back cache, and the board isn't configured properly for this. The manual I found had a bunch of CPU type jumpers in addition to the bus speed jumpers - are these all set correctly? What CPUs are you actually using with this board?

Reply 5 of 5, by Yuuker

Posted on 
User metadata
Rank Newbie
Rank
Newbie
SarahWalker wrote:

It could be that the faster CPUs are using a write back cache, and the board isn't configured properly for this. The manual I found had a bunch of CPU type jumpers in addition to the bus speed jumpers - are these all set correctly? What CPUs are you actually using with this board?

I'm running a AMD 5x86@133mhz. I'm assuming this (https://www.rom.by/files/4dps0210.pdf) is the manual everyone is finding?

I have my cpu-type jumpers ("CPU SELECT") set as described in the manual for an AMD 5x86.

Now the "CPU-Frequency-select" portion of the manual is where things get strange. These should be for the bus speed i understand.
This is where my 4DPS board doesn't seem to follow the manual. If i configure the jumpers for 33mhz as it shows in the manual, the cpu will report its mainspeed as something other than 133mhz.

Here's a chart i created with the results of my testing.
===============
bus jumpers configured for 33mhz: 5x86 reports its mainspeed at 80mhz. -still has distorted / garbage audio-
bus jumpers configured for 25mhz: 5x86 reports its mainspeed at 100mhz. -still has distorted / garbage audio-
bus jumpers configured for 40mhz: 5x86 reports its mainspeed at 160mhz. (A proper overclock?). -isn't stable, couldn't test-
bus jumpers configured for 50mhz: -didn't want to boot, couldn't test-
all bus jumpers closed/covered: 5x86 reports mainspeed at 133mhz as it should: -still has distorted / garbage audio-
===============
Iv included an example of how my board needs to be configured for the cpu speed to be detected properly. Notice how that not at all shown in the manual.

Maybe i'm still overlooking something, but this does seem a bit strange.

Attachments

  • 133mhzsetup.png
    Filename
    133mhzsetup.png
    File size
    836 Bytes
    Views
    579 views
    File comment
    This is how my CPU-frequency-select jumpers need to be setup inorder for the full 133mhz to be detected properly.
    File license
    Fair use/fair dealing exception
Go to top of page Go to top of page
Back to General Old Hardware
Close window
Close window