VOGONS


Tyan s1590s Graphics Cards problems

Topic actions

Reply 21 of 26, by pentiumspeed

User metadata
Rank l33t
Rank
l33t

Exactly about that chipset is poorly done is what I research first as I wanted to have working, easy to find parts, relatively speaking.

Cheers,

Great Northern aka Canada.

Reply 22 of 26, by shamino

User metadata
Rank l33t
Rank
l33t
Marmes wrote on 2020-07-28, 16:57:

I measured on mosfet , like you said. 1.8v, no post with any kind of graphics cards. Jumpers are all ok. I don't have a matrox g400, it may work or not.
I need to get one.

1.8V won't POST even if you don't overclock the CPU? This is at 400MHz?

I'm not suggesting that you get a G400, I'm just saying that since it works then it seems like a G200 should also work. So I think the problem is something more than the cards not being compatible.
If the cards are not even showing the BIOS text screens then that seems like a more fundamental problem than the quirkiness of VIA's AGP implementation.

But if the cards do POST and just misbehave when running accelerated drivers, then it could be a configuration/setup issue with AGP drivers, etc.
When you install AGP drivers, pick "Safe" mode install. This puts it in AGP 1X. This is a lot more reliable. There's no reason to run AGP 2X on this board, it does nothing for performance and just creates problems.

Reply 23 of 26, by shamino

User metadata
Rank l33t
Rank
l33t

But if you want a suggestion for a cheap card that I'm sure works, the Geforce2 MX is one. Those are dirt common cards and they run well.
I ran that on my system for years back then, never had any problems once I got it configured right.
You have to force it into AGP 1X mode though. AGP 2X will produce graphic glitches. I used older VIA drivers that had the "Safe" mode option during install.

Reply 24 of 26, by brichter1978

User metadata
Rank Newbie
Rank
Newbie

This an old thread, but hoping someone might have an idea. I also have the Tyan s1590s. I was debating moving to a newer ATX PS and notice the jumpers are barred into the AT position.

Instead of jumper pins it is a small bar solder across with both closed, which equals AT power.

Has anyone seen this? Pic attached.

Thanks,
Brian

Attachments

Reply 25 of 26, by PC Hoarder Patrol

User metadata
Rank l33t
Rank
l33t
brichter1978 wrote on 2024-01-18, 16:43:
This an old thread, but hoping someone might have an idea. I also have the Tyan s1590s. I was debating moving to a newer ATX PS […]
Show full quote

This an old thread, but hoping someone might have an idea. I also have the Tyan s1590s. I was debating moving to a newer ATX PS and notice the jumpers are barred into the AT position.

Instead of jumper pins it is a small bar solder across with both closed, which equals AT power.

Has anyone seen this? Pic attached.

Thanks,
Brian

Welcome to Vogons 😀

Quote from the manual (looks like you have the latter!)...

**Note: JP7 and JP14 are provided on
some S1590 boards. If your board does
not have JP7 and JP14, please disregard
the above chart - the correct settings
are already built-in to the motherboard.

.and version with jumpers.

Tyan S1590 AT_ATX JPs.jpg
Filename
Tyan S1590 AT_ATX JPs.jpg
File size
235.76 KiB
Views
89 views
File license
Fair use/fair dealing exception

Reply 26 of 26, by shamino

User metadata
Rank l33t
Rank
l33t
brichter1978 wrote on 2024-01-18, 16:43:

I also have the Tyan s1590s. I was debating moving to a newer ATX PS and notice the jumpers are barred into the AT position.
Instead of jumper pins it is a small bar solder across with both closed, which equals AT power.
Has anyone seen this? Pic attached.

I think my board has those same soldered jumpers.
I don't have it in front of me but if I remember correctly, those jumpers connect the onboard 3.3V linear regulator to the 3.3V rail on the board. That onboard regulator is there because AT power supplies don't supply 3.3V.
Since the jumpers are soldered on this revision, it means that the onboard regulator stays connected and an ATX power supply's 3.3V rail will be connected in parallel with it.
I've read of parallel power sources being a bad practice, but they did it anyway. Despite the soldered jumpers, the ATX 3.3V pins *are* connected on this board (I've confirmed this on Rev C).
If you want to use AGP video cards, then I do recommend using an ATX PSU because it provides a 3.3V rail, which takes the load off the onboard regulator. AGP cards draw from 3.3V so some of them can be risky when using an AT PSU.
Back when I had AT power, I ran this board with a Geforce2 MX, which is low power and won't overload the onboard regulator. Nowadays I have it powered with an ATX PSU, it works fine that way and can probably handle much more power hungry cards now (I haven't tried any yet).

The odd thing with this board is that even though it supports ATX, it needs an 8 slot case. ATX cases only have 7 slots, so if you use an ATX case you lose the AGP slot. So if you do want to run AGP cards on this board, the best setup seems to be an AT case (which has 8 slots) but with an ATX power supply. You'd have to use the reset button as the power switch.