VOGONS


First post, by ReeseRiverson

User metadata
Rank Member
Rank
Member

I figured I best post all these in one thread instead of making one for each, so I shall begin with a motherboard I am going to put back into a case, once I get a plan for the battery taken care of.

Does anyone know anything about this motherboard? It has a Texas Instruments 486 CPU, and it runs at 33MHz. About all I know, and it works just fine. Seems comparable to my Packard Bell 300SX Intel 386SX 16Mhz system in performance. This is just basing off how Bubble Ghost runs and my judgement, whether I am right or not. 🤣 - Though the battery did leak, and left most damage as you see there, but it didn't damage anything but where the positive part would go. I was wondering if I could solder wires into where the battery originally was soldered into, and tie that to a dual double A battery box, equipped with rechargable double A batteries for the 3 volt operation, since I already have these items, or if I should just make a connector and hook onto the External battery connector there, and not deal with a rechargable set?

Here it is:
IMG_2495-800px.jpg

Also, I have another board I'd really like to fix, if possible. It had ran well for me once upon a time, but it ended up not booting, but just giving out beep codes. What I hear from the speaker, are two high beeps close together, and then eight regular beeps, and then silence. Though according to what I read, there was 1 long, and 8 short beeps, which stand for display failure... but that can't be since I know the VGA cards I have are good, and I even tried one that I know for sure works. Here is a picture of that board, and it has no onboard CMOS battery, and I know it never seemed to have saved any settings for long when it did work.

IMG_2496-800px.jpg

Then finally, my last question for tonight, is this sound card... I have not tested it, but I am quite curious about it, and wondered if anyone may know anything about it? The largest chip seems to be a Texas Instruments DSP. With the number TMS320M520PQL on it.

IMG_2497-800px.jpg

Reply 1 of 39, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

That first board looks just like this one:

M396F motherboard

It's very good. Small and compact, not many parts, very stable and a super fast POSTer 😀

It would be awesome if you could benchmark that TI chip.

If you look close, there is a battery header 😀 Just use that with your battery pack.

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 2 of 39, by ReeseRiverson

User metadata
Rank Member
Rank
Member

Huh, they certainly do look the same exact thing, apart from different CPUs. I'm going to also guess a model number hides under the white sticker, on mine, too. 😀

Well, I'd be happy to benchmark it, but I never did this before, so I don't know what software to even look for. Hehe

Do you know which pin is hot and which is ground? Since our boards are identical? 😀

-Edit-

Oh I see you had a benchmark done on yours, I could just look that up too. 😀

-Edit 2 -

Got the sticker pulled back:

IMG_2498-800px.jpg

Which brings this up: http://stason.org/TULARC/pc/motherboards/F/FO … ml#.UsO5C_RDvXo

Seems we got the same model, just a different version. 😁

Reply 3 of 39, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

Yes very nice!

I'm really curious to see what you system can achieve. It should be quite a bit faster than mine. Could turn out to be an awesome board for a retro gaming DOS machine!

Wait I will attach some benchmarks for you to try...

EDIT: I wonder where these boards come from. They come with all sorts of CPUs, but all soldered. Very clean, minimal design as well.

Attachments

  • Filename
    Benching time.zip
    File size
    653.5 KiB
    Downloads
    54 downloads
    File license
    Fair use/fair dealing exception

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 4 of 39, by pyrogx

User metadata
Rank Member
Rank
Member

The sound card is a special card for the IBM EduQuest computer (should be this card: http://th99.bl4ckb0x.de/i/I-L/52907.htm). Its not ISA but a proprietary bus, so it is useless without an EduQuest system.

Reply 5 of 39, by sliderider

User metadata
Rank l33t++
Rank
l33t++

That Texas Instruments board should be a lot faster in CPU tests than a 386SX-16. That was the whole point of the 486SLC/DLC lines, to improve CPU performance in 386SX/DX based systems over what came with them. I can't believe that it doesn't run any faster.

Reply 6 of 39, by keropi

User metadata
Rank l33t++
Rank
l33t++
sliderider wrote:

That Texas Instruments board should be a lot faster in CPU tests than a 386SX-16. That was the whole point of the 486SLC/DLC lines, to improve CPU performance in 386SX/DX based systems over what came with them. I can't believe that it doesn't run any faster.

Maybe the internal 1kb of cache is not initialised ... this is done either via BIOS or a tsr for 486slc cpus.
I have upgraded my 386sx/20 IBM machine with a 486slc2 clip-on cpu, unless that 1kb gets initialized then the 486slc2 is not faster than the 386sx cpu...

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 7 of 39, by ReeseRiverson

User metadata
Rank Member
Rank
Member
pyrogx wrote:

The sound card is a special card for the IBM EduQuest computer (should be this card: http://th99.bl4ckb0x.de/i/I-L/52907.htm). Its not ISA but a proprietary bus, so it is useless without an EduQuest system.

Ah, Thank You very much, that is quite a bit of interesting history there! 😀

Mau1wurf1977 wrote:
Yes very nice! […]
Show full quote

Yes very nice!

I'm really curious to see what you system can achieve. It should be quite a bit faster than mine. Could turn out to be an awesome board for a retro gaming DOS machine!

Wait I will attach some benchmarks for you to try...

EDIT: I wonder where these boards come from. They come with all sorts of CPUs, but all soldered. Very clean, minimal design as well.

Alright, I ran two of the benchmarks. Plus, I ran it before and after I enabled the Cyrix Cache in the Bios!

First, basic system information:
IMG_2500-1000px.jpg

Now, with Cyrix Cache disabled.
IMG_2503-1000px.jpg
IMG_2504-1000px.jpg

Now, with Cyrix Cache enabled.
IMG_2505-1000px.jpg
IMG_2506-1000px.jpg

This does improve the game speed for Jazz Jackrabbit, however it is still a tad slower than it should be. Obviously the Intel 486DX 33MHz still beats this, but hey, some games, this would do well still. Since it's in between my 386SX and my 486DX4 system. So I'm happy as it is. 😁

keropi wrote:
sliderider wrote:

That Texas Instruments board should be a lot faster in CPU tests than a 386SX-16. That was the whole point of the 486SLC/DLC lines, to improve CPU performance in 386SX/DX based systems over what came with them. I can't believe that it doesn't run any faster.

Maybe the internal 1kb of cache is not initialised ... this is done either via BIOS or a tsr for 486slc cpus.
I have upgraded my 386sx/20 IBM machine with a 486slc2 clip-on cpu, unless that 1kb gets initialized then the 486slc2 is not faster than the 386sx cpu...

And yeah, it was due to the cache not being enabled. 🤣

Reply 8 of 39, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

Very nice, thanks for testing!

Well the CPU score is better, but the 3dbench score should be higher.

Is this 3dbench, and not 3dbench2?

What graphics card do you have? Fast ones are Cirrus Logic, TSENG ET4000 and WD chip cards.

ReeseRiverson wrote:

Do you know which pin is hot and which is ground? Since our boards are identical? 😀
In the thread I linked earlier I have attached the PDF manual 😀

It has all the jumpers and also the polarity of the external battery header.

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 9 of 39, by kixs

User metadata
Rank l33t
Rank
l33t

I used to have a Cyrix 486slc-33MHz on a similar board in 1993/4. It was great. In my experience, with ET4000, it was about 10-15% faster than a 386DX-40MHz (I still have some benchmarks - I was a benchmark freak even back then 😀 ).

Bye

PS:
I'm actually looking for one of these SLC boards... no luck yet.

Requests are also possible... /msg kixs

Reply 10 of 39, by ReeseRiverson

User metadata
Rank Member
Rank
Member
Mau1wurf1977 wrote:
Very nice, thanks for testing! […]
Show full quote

Very nice, thanks for testing!

Well the CPU score is better, but the 3dbench score should be higher.

Is this 3dbench, and not 3dbench2?

What graphics card do you have? Fast ones are Cirrus Logic, TSENG ET4000 and WD chip cards.

ReeseRiverson wrote:

Do you know which pin is hot and which is ground? Since our boards are identical? 😀
In the thread I linked earlier I have attached the PDF manual 😀

It has all the jumpers and also the polarity of the external battery header.

3dbench 2 is what I had ran, and the graphics card in this system is one with a Trident TVGA8900C chip and 1meg of video RAM. I also am not sure, but some odd graphical lines tend to flash on some games, and I can't run it more than 256 color in 800 x 600 in windows. Not sure if it is an odd issue with the card or not. Either case, here's the picture of the card:

IMG_2507-800px.jpg

You think this GPU is holding me back on this system then?

Reply 11 of 39, by kixs

User metadata
Rank l33t
Rank
l33t

Yes it does (a bit).

I did a few benchmark a while ago with 386DX-40.

3DBench 1.0:
Trident 8900D 1MB: 14.2
Tseng ET4000AX 1MB: 15.3

The same is in "real world" F1GP game (default install settings, no sound, Adelaide quick race. Press "O" for CPU usage before green lights):
Trident 8900D 1MB: 233-241% (CPU usage)
Tseng ET4000AX 1MB: 216-225%

After messing with the BIOS and setting ISA devider to 1:3 (but tested it with only ET4000) it was: 16.9 in 3DBench and 191-200% in F1GP.

I also tested it with 5x86/133:

3DBench 1.0:
Trident 8900D 1MB: 24.3
Tseng ET4000AX 1MB: 29.4
S3 325 2MB PCI: 62.5

F1GP game:
Trident 8900D 1MB: 108-116% (CPU usage)
Tseng ET4000AX 1MB: 91%
S3 325 2MB PCI: 50%

bye

Requests are also possible... /msg kixs

Reply 12 of 39, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

3dbench2. Got it.

Just be aware that the two version produce slightly different results.

Well it's pretty much a fast 386DX-33 with an ET4000.

If you get a faster graphics card your score will likely improve a bit. My 386DX-40 machines get a score of 16 - 17.

Either way it's a nice system for sure 😀

I know this board also comes with an SX-40 MHz. Would love to get my hands on one and see what that can do.

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 13 of 39, by ReeseRiverson

User metadata
Rank Member
Rank
Member

Thank you for the numbers, kixs, I wonder how the 386DX-40 fairs against this thing under the benchmark Sysinfo has.

Mau1wurf1977 wrote:
3dbench2. Got it. […]
Show full quote

3dbench2. Got it.

Just be aware that the two version produce slightly different results.

Well it's pretty much a fast 386DX-33 with an ET4000.

If you get a faster graphics card your score will likely improve a bit. My 386DX-40 machines get a score of 16 - 17.

Either way it's a nice system for sure 😀

I know this board also comes with an SX-40 MHz. Would love to get my hands on one and see what that can do.

Well, I didn't expect mine to be a speed demon, haha, but it is good to know it's still a step up from my 386SX Packard Bell, for when I need it. Also learning of the Cyrix Cache in the BIOS does add a nice boost.

Good news, is my dual double A battery box connected to the external battery header works well so far. Haven't had to reset the BIOS settings. Though these two batteries need to be charged up, they are only spitting out 2.5 volts for the pair at the moment. (Using rechargable ones.)

Also, you got me curious, so I decided to run the benchmarks on my 486DX4 system.

IMG_2508-1000px.jpg
IMG_2511-1000px.jpg
IMG_2512-1000px.jpg

Reply 14 of 39, by kixs

User metadata
Rank l33t
Rank
l33t

Here you are. Above 386DX-40 in Norton Sysinfo: 41.3

I know this board also comes with an SX-40 MHz. Would love to get my hands on one and see what that can do.

I did benchmark a 386SX-33, 8MB, ET4000AX 1MB (386DX-40 results in brackets)

3DBench 1.0: 10.3 [16.9]
Checkit 3.0:
Drystones: 7366 [10523]
Whetstones: 132K (no FPU), [3077.5K with IIT FPU]
F1GP: 308% [191-200%]
Norton Sysinfo: 26.7 [41.3]

Adding benchmarks for Cyrix 486DLC-40, 8MB, CL-5429 2MB ISA:

3DBench 1.0: 23.2
Checkit 3.0:
Drystones: 15785
Whetstones: 3854.5K with IIT FPU
F1GP: 150%
Norton Sysinfo: 65.6 [42.3]

Requests are also possible... /msg kixs

Reply 15 of 39, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

Nice and thanks for sharing!

That DLC performs like a FAST 486SX-25 MHz with PCI graphics:

http://www.youtube.com/watch?v=C1kWxxrs2ww&feature=youtu.be

The video also shows scores for PCPBENCH/VGAMODE and DOOM -TIMEDEMO DEMO3

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel

Reply 16 of 39, by kixs

User metadata
Rank l33t
Rank
l33t

DOOM -TIMEDEMO DEMO3

486DLC-40 = 5878 realtics == 12.70 FPS
386DX-40 = 7816 realtics == 9.55 FPS

The DLC board has actually two VLB ports, but I can't find the benchmark results with VLB graphics card (CL-5428 1MB). From memory the 3DBench 1.0 score was around 2FPS faster.

PCPlayer VGAmode: I only began using it recently. The 386DX-40 scores 4.3 and I haven't tested the DLC yet.

Requests are also possible... /msg kixs

Reply 17 of 39, by Mau1wurf1977

User metadata
Rank l33t++
Rank
l33t++

You can submit your scores here: Phil's Ultimate VGA Benchmark Database Project

My website with reviews, demos, drivers, tutorials and more...
My YouTube channel