VOGONS


Radeon 7500 mobile on pci card.

Topic actions

Reply 20 of 32, by darry

User metadata
Rank l33t++
Rank
l33t++
Sphere478 wrote on 2024-07-11, 20:44:
Demolition-Man wrote on 2024-07-10, 16:12:

I only had the Radeon 7000M in the ThinClient for comparison. Good compatibility under DOS, no problems with corresponding Win 98 3D games. No unusual driver was necessary either. On the other hand, this self-assembled 7500 PCI from 2011, according to the sticker, is practically unusable. I have no idea why you would need it in a suitable PC.

It’s a conversation piece. I would like to get it working though 😀

Best of luck. If you can't source or modify drivers to get it working in Windows, Linux might be something to try.

Reply 21 of 32, by Demolition-Man

User metadata
Rank Member
Rank
Member

I can't solder. So I have to get rid of the card as quickly as possible.Thanks again for the warning. xD

Reply 22 of 32, by momaka

User metadata
Rank Oldbie
Rank
Oldbie
Demolition-Man wrote on 2024-07-13, 09:40:

I can't solder.

So take it as an opportunity to learn?? 😀
Or you've tried and know this just isn't for you (I get it, if that's the case.)
Either way, try not to trash / junk the card, even if you don't want to try fixing it. Perhaps someone else might want it to give it a try. Most of my ebay cards were intentional non-working purchases (and typically for cheap). From the ones that were with bad caps, most worked fine after. The three that didn't work were 2x EVGA GeForce 7600 GT's (PCI-E version) and one XFX GeForce 8600 GT.

Reply 23 of 32, by Demolition-Man

User metadata
Rank Member
Rank
Member

I meant sell as long as it is still possible.

So take it as an opportunity to learn?? 😀

Medical cause, sorry.

Reply 24 of 32, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

There was a sticker between the gpu and the heatsink it said 2011 on it. The paste looked pretty fresh. The card had dust on it, it had been used for sure.

Here is the gpu:

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 25 of 32, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

As for the caps.

They are totally unacceptable lmao.

Also note the dual smd/thru hole mounting scheme.

I think these all need to be swapped out for smd to avoid the case of the cap possibly touching the pads.

Seems like there should be a old laptop driver somewhere we can use for this?

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 26 of 32, by darry

User metadata
Rank l33t++
Rank
l33t++
Sphere478 wrote on 2024-07-16, 02:50:
As for the caps. […]
Show full quote

As for the caps.

They are totally unacceptable lmao.

Also note the dual smd/thru hole mounting scheme.

I think these all need to be swapped out for smd to avoid the case of the cap possibly touching the pads.

Seems like there should be a old laptop driver somewhere we can use for this?

Maybe it's the lighting and/or angle, but some of these caps look more bloated than Baron Vladimir Harkonnen after a few weeks of a zero fiber diet.

When you get a chance, please share the PCI device and vendor IDs.

Reply 27 of 32, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

Oh that’s a good idea.

Yes, the capacitors have already launched an attack on house Atreides.

The electrolyte must flow.

Do you think any of those resistors are DEV ID config?

So what’s the final word on this chip being Radeon desktop 7500 silicon or not? If same thing, change Device id and call it done or force driver install. If not….. is there something else that shares this silicon?

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 28 of 32, by AaronS

User metadata
Rank Member
Rank
Member

Hi, I know this is a bit of a topic bump but I think these are from the same company/factory. I picked up a joblot of what were labelled as "ATI 9000 64M" and a single "ATI 9200 128M", they're all dated from around 2009 and they all look exactly like the Mobility 7500 you have (except the fanless 9200), I suspected they were mobility chips before I bought them because I've seen these now and then on ebay, but thought I'd give them a chance.

The attachment Image1.png is no longer available

Right off the bat none of them will post on my system (its a modern Asus M5A97 AM3+ UEFI motherboard with 2 PCI slots, AMD 9xx chipset, so they aren't bridged PCI slots). I thought this was strange because I have successfully booted using a Nvidia Quadro FX600 PCI and also a Nvidia Geforce 6200 PCI, I tried resetting the bios and even removing the CMOS battery but nothing. Then I remembered reading somewhere on here about certain PCI cards (I can't remember which, maybe these ones) where they couldn't get the system to post UNLESS there was an AGP slot present (even old OEM boards that only have PCI slots etc.), I'm not sure if this is the case as I don't have an AGP motherboard to test but since these Mobility cards were used in laptops and were usually AGP 4X it sort of makes sense I suppose.

Anyway, using a modern PCIe GPU (or even the PCI Quadro FX600 or 6200) alongside these cards allows my system to post and then boot into Windows (in this initial case; XP SP3), and the cards show up in Everest, HWinfo etc.

The attachment Info.PNG is no longer available

Now comes the dreaded drivers part, even with desktop Radeons of this era the drivers can be a nightmare, but I spent a few hours testing. Firstly the Radeon Mobility 9000; I went straight in with Omega drivers from philscomputerlab (cat_7.12_ati_omega_xp2k_48442.exe), the card was detected and installed fine, after a reboot the Mobility 9000 now becomes the primary display and the PCIe output switches off. All seems to work fine, allowing me to switch to 32 bit color and 1024x768 in Display, until I tried to launch 3DMark2001, "No Direct3D detected", checked dxdiag, same thing, No Direct3D enabled etc. Uninstalled completely and then downloaded the official drivers (again, thanks phil) "wxp-w2k-radeon-7-90-030605m-009437c-efg.exe" and "control-panel-7-90-030605m-009437c-efg.exe" (.NET Framework 2.0 is needed first) extracted the drivers and because these don't include the Mobility cards (usually left up to OEMs) I force installed "RADEON 9000 SERIES" (I know they're not exactly the same as the desktop variant) in Device Manager (making sure to use the right Video Controller: 1002:4C66), and then installed the Control Panel, rebooted and all seems good, no Direct3D error message in dxdiag and 3DMark2001 launches and runs... up to the nature scene, then the PC force rebooted itself. I checked a bunch of options in ATI Control Panel and it would always crash on this one part, I had a feeling it was the vertex shader and I tested the vertex shader specifically in 3DMark2001 and yep, same thing there, reboots the system. I then went into 3DMark2001 settings and changed the "Pure Hardware T&L" setting to "Software T&L" this time the benchmark completely fully but the performance was a bit lower.

Certain apps such as fraps, rivatuner, GPU-Z that "probe the hardware" also causes the reboot. I tried using different drivers including ones that don't use a vertex shader such as RADEON 7000 SERIES but still, the reboot thing happened with them all. 3DMark2000 completed fine however.

At this point I concluded that these aren't going to play nicely with DirectX 8 games, despite being a DirectX 8.1 card.

I found a driver that works afterall, will update post with more info soon!

I wasn't able to get them working under Windows 98SE or ME at all on this system, you absolutely need a legacy system (perhaps with a AGP slot) in order to use it as a primary video card from the get go. I get a black text box when Windows 98 is booted (using the PCIe GPU as primary), and on the Radeons output it says something like "Driver is installed, configure in Display panel", but unable to make it work as a secondary card, perhaps with a second PCIe Radeon or something.

As for the Mobility 9200 128M card, not sure but I can't get any driver to work at all for it under 98, ME or XP. Everest, HWinfo etc. detect it as M9+ and it has an ID of 1002:5C63 but otherwise, no luck at all. Stats in Everest are of course innaccurate until I am able to install a driver for it.

The attachment Untitled.png is no longer available

Overall pretty disappointing, I would have better luck with a legacy system (with AGP slot if that is what is needed) but even after that, they also seem very finnicky to actually get working, and when they do they are still iffy. I did test some OEM Mobility 9000 drivers I found and changed the Device IDs accordingly but they didn't seem to work after a reboot.

Reply 30 of 32, by swaaye

User metadata
Rank l33t++
Rank
l33t++

Just try the Driver Heaven Mobility Modder / Mobility Mod Tool to convert a desktop driver release into a Mobility compatible package.

https://www.vogonsdrivers.com/index.php?catid=22

After this does its thing you can run the ATI setup program and it recognizes the GPU.

I used this all the time back when I had Mobility Radeon notebooks and it was the only way to get a driver update.

Reply 32 of 32, by Grem Five

User metadata
Rank Oldbie
Rank
Oldbie
swaaye wrote on Yesterday, 22:32:

Just try the Driver Heaven Mobility Modder / Mobility Mod Tool to convert a desktop driver release into a Mobility compatible package.

I never got around to trying that tool but I linked it in this thread 2 years ago. Re: Radeon 7500 mobile on pci card.