VOGONS


3 (+3 more) retro battle stations

Topic actions

Reply 720 of 2108, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie

@chadti99
I am on win95 osr2 (or dos6.22/win3.11 - depends on which CF card I pick or what I want to test/check).
Do you get distorted sound in DOS ? Can check win98 as well - maybe it is something specific there.

How I get 2111 at 180mhz ?
By being very very Very Very VERY VERY patient with finding the right chips combination. I mentioned in previous posts that the UUD board is extremely picky about L2 cache chips if tightest BIOS timings are desired. With that in place it turns into one of the fastest socket 3 mobos, but the path to there practically rules it out for pretty much everyone. I can point you to previous discussion with Feipoa in the thread where prior to that I was on the opinion that the board is flaky/compromised in that regard, but his comments motivated me to experiment. It is possible.
12ns chips don't ensure tight timings. 256kb here happens with 15ns umc chips. 10124kb is with mixture of 15 and 10ns issi chips.

@Feipoa
I remember checking these threads.
How you tell if it is X or P related bios ?
Didn't observe any side effects other than slightly lower frame rate in some DOS tests which made me back off from it.

retro bits and bytes

Reply 721 of 2108, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie
feipoa wrote on 2021-11-24, 10:03:

I suppose they aren't great when compared with a voodoo2 in GLQuake at 800x600, but 640x480 at 19.4 fps is certainly better than software mode. If it looks better than software mode, I can see some interest in this product. Can the card go to 800x600? And at 320x200 - 30.4 fps. Isn't the record for software mode at this resolution on an Am5x86 only 20.6 fps?

It is certainly better than raw vga card.

retro bits and bytes

Reply 722 of 2108, by Chadti99

User metadata
Rank Oldbie
Rank
Oldbie
feipoa wrote on 2021-11-24, 10:03:
pshipkov wrote on 2021-11-22, 04:10:

These numbers are dismal.
The rest of the details you added only make it sound worse.
Thanks for spending the time on this.

I suppose they aren't great when compared with a voodoo2 in GLQuake at 800x600, but 640x480 at 19.4 fps is certainly better than software mode. If it looks better than software mode, I can see some interest in this product. Can the card go to 800x600? And at 320x200 - 30.4 fps. Isn't the record for software mode at this resolution on an Am5x86 only 20.6 fps?

Here are the supported resolutions, at the closest resolution to 800x600, which I think is 768x568, it still hits 19.8 fps and 856x480 at 21fps on a POD100.

Attachments

Reply 723 of 2108, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie

Soo, i am going through some POD100 tests and noticed that Pc-Chips M919 (Amptron DX-9700) has no listed jumpers for P24T/POD CPU.
All online databases mirror the same DX-9700 manual.
Anyone ?

retro bits and bytes

Reply 724 of 2108, by feipoa

User metadata
Rank l33t++
Rank
l33t++
pshipkov wrote on 2021-11-24, 16:10:
@Feipoa I remember checking these threads. How you tell if it is X or P related bios ? Didn't observe any side effects other tha […]
Show full quote

@Feipoa
I remember checking these threads.
How you tell if it is X or P related bios ?
Didn't observe any side effects other than slightly lower frame rate in some DOS tests which made me back off from it.

Whether it works properly or not. That's all I know.

I guess this means you don't have the BIOS from a -P VLB card that is 2.02 or later? Perhaps a BIOS from a -P PCI card will work. A user named mpe said he had the VLB card with 2.02, but sold or traded it without saving the BIOS. Elusive little beast isn't it...

pshipkov wrote on 2021-11-25, 03:01:

Soo, i am going through some POD100 tests and noticed that Pc-Chips M919 (Amptron DX-9700) has no listed jumpers for P24T/POD CPU.
All online databases mirror the same DX-9700 manual.
Anyone ?

I never found any specs for the P24T on the m919, probably because they knew it wouldn't work in L1:WB mode. Please share working jumper settings if you find them. So far, P24T works fine in L1:WT mode, even if the BIOS claims it is in WB mode. Use chkcpu to see what mode L1 is in.

Plan your life wisely, you'll be dead before you know it.

Reply 725 of 2108, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie

The 2.09 BIOS is from a PCI card with P chip on it (need to double check if still have it - it was Diamond card).
It works just fine on the S3 Trio64 Diamond Stealth DRAM VBL card with P chip.

Your second message is unclear.
First you said that you don't know the jumper settings, but then it sounds like you are able to run POD in the mobo, just with L1 in WT mode.
L1 WB works on this board with Am5x86 processor.

retro bits and bytes

Reply 726 of 2108, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Do you have a photo of the PCI card with -P and BIOS v2.09 on it and the Diamond VLB card you've tested with this BIOS successfully? It sounds like you have personally tested it on your Diamond DRAM-T with -P chip, correct? Did you try to perform a soft reset in a) DOS, and b) Win95 while the card was inserted in the VL/I-486SV2GX4 using Am5x86-160? If the answer is yes to both of those, then there may be some other hardware issue going on with my particular setup and v2.09 BIOS.

If you managed to read those other threads of mine, you'll notice that one other brand of BIOS worked fine on my Diamond card with soft-reset and the sought after VESA/LFB modes, but using it is inelegant (mixed brands). The original Diamond v2.01 BIOS doesn't have the soft-reset issue but is missing the VESA modes in the EPROM. v2.09 yielded soft-reset issues, but has the desired VESA modes. I'd like to track down what makes your setup successful in this regard. Because of photographic evidence, I know v2.02 came with the -P card, so it is very likely to work on my card. Whether it has the desired VESA modes is left to be determined.

Regarding my second message, I will try to insert more words. I did not find jumper settings for the P24T on the m919 v3.4 motherboard anywhere online or in the official printed manual that comes with this motherboard. I used the P24D settings for the P24T while inserted in the m919 motherboard. The P24T CPU will not work in WB mode properly with Am5x86, Cx5x86, nor P24D settings while in the m919. If I recall, the BIOS may indicate that L1 is in WB mode, but chkcpu and benchmark results demonstrate otherwise. If you find some magic jumper settings which let the P24T work on the m919 v3.4 in L1:write-back mode, please share them.

Plan your life wisely, you'll be dead before you know it.

Reply 727 of 2108, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie

Pci card was modest looking diamond something. I think I still have it somewhere.
There is a picture of the vlb card in the first post of this thread.
Cannot remember about win9x on Asus vli sv2gx4, but dos (and win3.x) - for sure. Soft reset is fine.
If I remember your vli based rig is chock loaded with hardware. Maybe something there ?

Thanks for the clarification for the second question.
Will see where things will go with this m919 guy.

retro bits and bytes

Reply 728 of 2108, by feipoa

User metadata
Rank l33t++
Rank
l33t++
pshipkov wrote on 2021-11-25, 05:24:

Pci card was modest looking diamond something. I think I still have it somewhere.

If you don't mind, please share a photo and re-send me the BIOS. I will check again.

pshipkov wrote on 2021-11-25, 05:24:

There is a picture of the vlb card in the first post of this thread.

Thanks, I'll check.

pshipkov wrote on 2021-11-25, 05:24:

Cannot remember about win9x on Asus vli sv2gx4, but dos (and win3.x) - for sure. Soft reset is fine.
If I remember your vli based rig is chock loaded with hardware. Maybe something there ?

Possibly, but doesn't explain why the original 2.01 BIOS works fine and a Hercules BIOS works fine - all with the same supporting hardware. My money is on v2.02 BIOS working. Also, I thought I had tested v2.09 BIOS before you came around, so it is possible I am using a different BIOS, so if you send yours again, I'll do a double take.

Plan your life wisely, you'll be dead before you know it.

Reply 729 of 2108, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie

Will share a picture of the other card soon.
Pretty sure i bought it from here. Looked exactly the same but with empty memory sockets.
But also, check this link as well. P model with 2.06 BIOS.

Hard to tell what your problem may be.
Do you have a real use for the VESA modes in DOS, or going for completeness ?

Attachments

retro bits and bytes

Reply 730 of 2108, by feipoa

User metadata
Rank l33t++
Rank
l33t++

I'd need to retrace my steps to recall exactly why I wanted the absent VESA modes; I suspect it was a benchmark or game which needed it, and running univbe sucked up too much conventional memory for, I think, wolfy3d, and it got messy. Turns this on, turn this off, edit, reboot, time goes by, and I forget the steps, etc. It is not elegant. Completeness is also a factor, but wasn't what set me on the journey.

yeah a 2.06 BIOS would be good to try as well, but not for $92.

Thanks, I'll try your BIOS once I see your PCI photo. So you definatley dumped this BIOS yourself - didn't download it from somewhere?

Plan your life wisely, you'll be dead before you know it.

Reply 731 of 2108, by mpe

User metadata
Rank Oldbie
Rank
Oldbie

This one is currently one slightly cheaper on a bidding platform https://i.ebayimg.com/images/g/GP0AAOSwGOZfnGnf/s-l1600.jpg

-P 2.02 PCI

There is no change in ROM contents between PCI and VLB BIOS versions (confirmed).

But it any case no BIOS would add VBE 2.0 support, so VESA TSRs are always superior.

Blog|NexGen 586|S4

Reply 732 of 2108, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie
feipoa wrote on 2021-11-25, 07:57:

I'd need to retrace my steps to recall exactly why I wanted the absent VESA modes; I suspect it was a benchmark or game which needed it, and running univbe sucked up too much conventional memory for, I think, wolfy3d, and it got messy. Turns this on, turn this off, edit, reboot, time goes by, and I forget the steps, etc. It is not elegant. Completeness is also a factor, but wasn't what set me on the journey.

yeah a 2.06 BIOS would be good to try as well, but not for $92.

Thanks, I'll try your BIOS once I see your PCI photo. So you definatley dumped this BIOS yourself - didn't download it from somewhere?

I see.
I kind of never had use for higher-resolution VESA modes.
Didn't happen for video games - 486 hardware cant do higher resolution interactive graphics.
The few apps that utilized higher resolutions in DOS at the time provided their own video drivers, or required vendor drivers which take care of business.

The bios file i attached in my previous post is a copy of what's in the EPROM in the picture here.

Attachments

  • 1.jpg
    Filename
    1.jpg
    File size
    81.22 KiB
    Views
    834 views
    File license
    Public domain

retro bits and bytes

Reply 734 of 2108, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Thanks for the photo! Very interesting, that is assuming one of the previous owners didn't swap the EEPROM chips on this card. I'll have to test the v2.09 BIOS again.

Plan your life wisely, you'll be dead before you know it.

Reply 735 of 2108, by pshipkov

User metadata
Rank Oldbie
Rank
Oldbie

This crossed my mind.
While the card has scratches everywhere i remember looking more carefully under the EPROM chip and there are scratches on the plastic socket that are typical when sharp tool is inserted multiple times for chip removal.
Impossible to say if it came with that bios chip out of the factory or it was changed throughout the years since then.

retro bits and bytes

Reply 737 of 2108, by feipoa

User metadata
Rank l33t++
Rank
l33t++
pshipkov wrote on 2021-11-26, 06:18:

This crossed my mind.
While the card has scratches everywhere i remember looking more carefully under the EPROM chip and there are scratches on the plastic socket that are typical when sharp tool is inserted multiple times for chip removal.
Impossible to say if it came with that bios chip out of the factory or it was changed throughout the years since then.

Tough to say. If I had to guess (e.g. because there was a gun to my head), I'd be leaning towards the BIOS chip was swapped.

Last edited by feipoa on 2021-11-27, 14:15. Edited 1 time in total.

Plan your life wisely, you'll be dead before you know it.

Reply 738 of 2108, by feipoa

User metadata
Rank l33t++
Rank
l33t++
Chadti99 wrote on 2021-11-27, 12:17:

Does anyone know which resistor controls the 4v CPU voltage on the LS486-D ? I’d like to try and apply Feipoa’s voltage mod to this board.

I'm not sure what voltages you are after, but I just soldered a 180 K-ohm resistor in parallel with R16 to achieve 3.675 V on the 4V jumper. My notes indicate that putting 220K in parallel with R16 would yield 3.75 V.

Plan your life wisely, you'll be dead before you know it.

Reply 739 of 2108, by Chadti99

User metadata
Rank Oldbie
Rank
Oldbie
feipoa wrote on 2021-11-27, 14:14:
Chadti99 wrote on 2021-11-27, 12:17:

Does anyone know which resistor controls the 4v CPU voltage on the LS486-D ? I’d like to try and apply Feipoa’s voltage mod to this board.

I'm not sure what voltages you are after, but I just soldered a 180 K-ohm resistor in parallel with R16 to achieve 3.675 V on the 4V jumper. My notes indicate that putting 220K in parallel with R16 would yield 3.75 V.

Yes thank you, wanting to test between 3.45 and 4. Was going to try the variable resistor.