VOGONS


First post, by ReeseRiverson

User metadata
Rank Member
Rank
Member

It's been a long time since I've really worked on any of my retro gaming projects, especially since I've been busy playing Splatoon more than anything these days. The chassis is the same one I used back in 2014 f0r this old build project:
Leaping Lab rats! Another build progress of some kind by Reese!

I've stopped using that ever since it would never detect a keyboard anymore, which since then I did remove all the hardware and decided to re-assemble the case more properly.

This build did lead to some rather complex situations, which all but one have been resolved. I'll explain more on that later.

1022202227c.jpg

To start off with, the AOpen Vi15G VLB motherboard, I originally got this motherboard a while back due to the inspiration of LGR's woodgrain 486 build, plus I never had an actual VLB based 486 system running. The chassis I'm using did originally have a VLB based board in it, but one day it decided to simply quit working years back. I still have the motherboard and would like to attempt to repair it sometime. That'll be another thread once I reach that point.

(Further details provided within the spoiler tag. TL:DR is: Complete vs incomplete jumper setting documentation, resolving reboot issues due to using correct CPU jumper setting, links, etc.)

AOpen Vi15G Jumper Documentation Details

Though I would like to point out that the documentation I found online for the Vi15G on the typical sites were not complete. Specifically elhvb and stason.

I know at least two versions of this motherboard exists, the 5V only version, which is the one I have, and then the 3.45V version, which is what LGR has. This in itself really isn't a deal breaker, let alone an issue in the grand scheme of things. However, this does mean the documentation also is missing a lot of jumper settings for the 3.45V version. Which also is applicable to the 5V version when it comes to using CPUs beyond a 486DX2*.

(Asterisk on that to state that you must use a CPU that supports 5V, an Overdrive chip, or an Interposer. Since the latter two have a voltage regulator.)

That said, I wanted to make sure to document this because I've searched all over Vogons and other internet sources and I could not find any accuracy in settings to match my configuration specifically. The issues caused by this resulted in the system to not be capable of rebooting via CTRL + ALT + DEL nor by software issued reboot command. This was resolved by using the correct jumper settings for my 486DX4 CPU listed by AOpen's actual documentation hosted on their FTP:
ftp://ftp.aopen.com/pub/tech/jumper/vi15g/default.htm

To preserve these settings in the event that anything could happen to an ftp site these days, I did save the documentation and made a screenshot for anyone to reference this:
121773427_10157974952254067_4534284164360744216_n.jpg

The CPU I have installed is the Intel 486 Overdrive. (DX4ODP100 - Sz957).

So I definitely hope this will help anyone who happens to have the same motherboard as this, either version they may have. I've already given the settings I've found to LGR as well.

System Specifications:
  • Motherboard
    AOpen Vi15G (5v Only, no 3.45v support.)
  • CPU
    Intel 486 Overdrive: DX4 100MHz (DX4ODP100 - SZ957)
  • MEMORY
    2 x 16MB 60ns FPM RAM (32MB total)
  • CACHE
    9 x W24257AK15 (32K x 8 SRAM) chips. (One used as TAG)
  • STORAGE
    Syba SD-CF-IDE-DI + Transcend 2GB Compact Flash Card
  • I/O, FDD, & HDD Controller
    Some kind of Promise VLB I/O FDD HDD controller, no BIOS onboard like a lot of other PROMISE controllers.
  • 2nd IDE
    Promise EIDE (2nd IDE provided for the CD-ROM)
  • GRAPHICS
    S3 805 (86C805-P) VLB VGA card, 2MB total memory.
  • SOUND
    Creative Vibra 16 CT2260 (For now.)
    MIDIMAN MPU401

This chassis ended up being my favorite retro computer case ever since I got it, which was roughly back in 2005, if I remember correctly. At one point, between then and that 2014 build thread is when the MHz display either was removed or not connected. I found it and installed it at a later date. It used to have a couple of 3.5mm audio jacks that ran through a small opening near the serial port by the PSU, connected internally to a 3-pin connector for the front of the chassis where the amplifier is located. At some point those disappeared, and I hooked the speakers up directly to an AWE32's MB_PRO header to feed audio from after re-pinning it. Now I'm using a bracket and had adapted the 3.5mm jack from the bracket to have a clean external connection since the current audio card doesn't provide any headers on it.

Audio LED VU Meters in Action:
https://youtu.be/2rnXY3tvLlo

It was a bit of a lengthy process to diagnose issues I had with the system, where I wasn't sure what the deal was/what things were related to. The details in the above spoiler have corrected the reboot issues. The issue that's left is The Adventures of Willy Beamish, the CD-ROM edition (Not the Floppy Disk version). I'll explain below:

Weird Willy Beamish CD-ROM edition glitches!

This particular issue isn't related to issues where the speed issue/game lockups can happen, though I do have that patch installed. This is like a weird graphical bug where it's like... missing graphics, or graphics isn't updating properly. There's missing credits text in the intro, some missing animations entirely, and some background updating issues.

Recording:
https://www.youtube.com/watch?v=ikW8VbCjEow

To point out a few:

  • At 1:57, you can see the background isn't updating like it should, it ends up being where Willy skateboards on the incorrect background before crashing into the school, where you see a small square of that update.
  • At 3:03 you can see a scene common between the CD-ROM and Floppy Disk version of this game. Willy daydreaming the Nintari Championship. You can see this scene glitching out graphically. (The Floppy Disk version works perfectly on this same machine.)
  • At 8:05 you can see another background issue. Again, no such issue on the Floppy Disk version on this systme.

Attempted Fixes/Tests:

  • I've tried running the CD-ROM version directly off the Compact Flash card and the CD-ROM itself.
  • I've tried bypassing the Autoexec.bat and System.cfg, I've tried running off a MS-DOS boot disk and loading the game off the CF and CD-ROM. Even a different CF card.
  • I've tried the version from GOG.
  • I've tried with multiple CPUs. Intel 486DX2, the currently installed Intel 486DX4 OD chip, and even the Intel Pentium OD chip.
  • I've tried different RAM combinations.
  • I've tried with L1, L2, and L1 + L2 cache disabling combinations.
  • I've tried a few different VGA cards. The currently installed S3 805 VLB card, an ATI Mach 32 VLB card, and a Western Digital ISA card...
  • I've even tried different I/O, FDD, HDD controller combinations between VLB/ISA, with/without the Promise EIDE I was using for the CD-ROM.
  • I've tried optimizing the BIOS settings, no changes here.
  • I've tried different Sound Cards.
  • I went as far to even try a different PSU. No change.
  • With and without the audio driver hack to allow for Roland MT32 in the CD-ROM version. (Especially when running from CD-ROM directly.)

The only component left is basically the motherboard itself...

These issues do not happen on my other 486DX4 system, which has a PCI board. Nor does it happen on my Gateway 2000's EISA based 486 system. I did replicate this issue on another 486 PCI based motherboard though. All of this with the same exact CD-ROM game disc. DOSBox runs the game without issues as well. I'm extremely baffled by this.

Does anyone have any idea what's going on? The only theory one of my friends can think of is motherboard chipset compatibility at this point.

As for other games, Descent works beautifully well, both I and II. Duke Nukem 3D works well after figuring out the issues I had were related to the audio card using high sample rate + MPU401.

Write issues I had apparently were never HDD controller related, but either a bad ribbon cable or too long of a ribbon cable for the CF to IDE adapter. Since I used a direct connect adapter, things have been working rather well.

One of the neat utilities I've found out about, called The Draw, I used to help make my own custom 'main/welcome' screen. I also found out about LaunchBox for MS-DOS here on Vogons and ended up liking it. This was a fun little project:
https://youtu.be/NKT4XjD8sgU

More Pictures

1009201917.jpg
1022202222.jpg
1022202223.jpg
1022202223c.jpg
1022202224.jpg
1022202224a.jpg

Plans for this machine:

  • Replace the aging capacitors on the PSU for obvious reasons.
  • Get the Orpheus Soundcard, I really would like the features it offers.
  • Get some shorter IDE cables so I can have the CF to IDE brackets installed on the back for easier access.

Attachments

Reply 4 of 9, by ReeseRiverson

User metadata
Rank Member
Rank
Member
chrismeyer6 wrote on 2020-10-27, 00:50:

I just watched the video on the VU meters and the sound quality with those built in speakers is quite good.

Yeah, the speakers built into this case easily best some of the computer speakers I've heard/had in the past. Compared to our Acer back in the 90s, the speakers that came with it didn't have any oomph, especially with bass. These have a little bit. At least for what they can do, being open into the chassis.

They can't touch my Altec Lansing speakers I have for my Gateway 2000, but that wouldn't be a very fair comparison. 🤣

Pierre32 wrote on 2020-10-27, 01:22:

Excellent build in an excellent case. Love the videos.

Thank you, I'm still tweaking the build with some. I remembered I had my SoundBlaster 16 stored in another chassis, so I swapped the Vibra16 with that.

I am curious about what the amplifier consists of, I may try to look into it further to see what transistors were used. Or if it was a semi-packaged amplifier module. The speakers have a 4ohm impedance on them.

Reply 5 of 9, by Intel486dx33

User metadata
Rank l33t
Rank
l33t

For DOS build and BEST audio and DOS game play back you want

at least :
16gb ram
SSD or CF card
32 bit video S3 or Cirrus Logic 5428 or higher
Sound Blaster 16 or compatible with Yamaha OPL3
32bit data transfer
32x CDROM drive or higher.

2x and 4x CDROM drives Seek allot.

Reply 6 of 9, by leileilol

User metadata
Rank l33t++
Rank
l33t++

FWIW, the same Willy CD issues are reproducible on PCem even on 386SX 16MHz+SB1.5+DOS5+1XCD configs that would be period appropriate for the release. Wouldn't doubt if there's some old 1993 usenet posts complaining about this...

apsosig.png
long live PCem

Reply 8 of 9, by ReeseRiverson

User metadata
Rank Member
Rank
Member

I've ran a couple of tests on the system, here are the screenshots of CacheCHK and the VGA test:
486%20CACHECHK.png
VGA%20FPS.png

leileilol wrote on 2020-10-28, 07:46:

FWIW, the same Willy CD issues are reproducible on PCem even on 386SX 16MHz+SB1.5+DOS5+1XCD configs that would be period appropriate for the release. Wouldn't doubt if there's some old 1993 usenet posts complaining about this...

How interesting, while I still am curious to the specifics on why, I also am relieved it's not limited to just my system here. Meaning there's definitely more of a weird issue with the game rather than the system.

Thank you for pointing this out.

Reply 9 of 9, by kinetix

User metadata
Rank Member
Rank
Member

hi, friend. I know this is an old thread, but it's the only place I've found to ask.

I have that same Acer/AOpen VI15G motherboard (actually, I have two), the revision with a voltage regulator and support for 3.3 volts. It is part of my two old PCs, stored about 16 years ago and that nostalgia has made me want to put them together again (the other is a pentium).

I remember that I had it with an Am5x86 installed, or so I think, but since I wasn't sure I changed the configuration (I forgot to write down the previous one, a pity, and important!). In the documentation it mentions a "Misc 5x86", but with that the 5x86 did not work. Then I read that it is the one corresponding to Cyrix´s 5x86, and that it was best to use the configuration for the Intel 486dx4 . however with this configuration I measured 5 volts in the socket!! I tried with the configuration of the AMD DX2 and Dx4 and if I got 3.45 volts. The bad thing is that this was after testing the 5x86 with the intel configuration, so I'm afraid I fried it. but I have another 80Mhz Amd 486dx2 and a 33mhz DX (3.3v). I tried with the DX2 anyway, the board does not boot, nor give beeps, anything

two questions

Do you have any idea why with the configuration for the intel dx4 it is giving 5 volts?

Could you share the configuration of the jumpers you are using?