VOGONS

Common searches


First post, by egbertjan

User metadata
Rank Member
Rank
Member

I have a FIC VA-502 motherboard with an Intel Pentium 233 CPU. The board has a VIA VT82C580VPX chipset. I am trying to install an IDE driver by VIA, version 2.3.18, made for the chipset I just mentioned. I used a 120 GB SSD (and later a 30 GB SSD) but under DOS 6.22 the computer hanged after having installed the IDE driver. You can see it on the picture I added.

20231003_161953.jpg
Filename
20231003_161953.jpg
File size
1.68 MiB
Views
1188 views
File license
Fair use/fair dealing exception

The version of this driver for Windows 3.11 did btw very well, the 32 bit disk and file acces and DMA were enabled.

I also have an Asus board with an ALI chipset. The IDE driver provided by ALI worked very well, under DOS as well as under Windows 3.11. The result of the driver is a big increase of the speed. The 32 bit disk and file acces and DMA were enabled.

Does anybody know of my problem with the VIA IDE driver? Is there a better version. Can I do something with other parameters in config.sys? Probably something in the BIOS?

Reply 2 of 7, by egbertjan

User metadata
Rank Member
Rank
Member

The bios of the FIC VA502 can easily handle 120GB hard drives and SSDs. I have a custom bios on it. Otherwise it won't go beyond 30 GB. I know that DOS can handle 2 GB per logical drive. I have divided it into 4 parties of 2 GB.

Reply 3 of 7, by megatron-uk

User metadata
Rank Oldbie
Rank
Oldbie

Have you tried alternative (U)DMA drivers that are distributed with FreeDOS? I believe they are called xdma.sys and uide.sys, I recall them supporting various Intel and Via chipsets.

My collection database and technical wiki:
https://www.target-earth.net

Reply 7 of 7, by sgray

User metadata
Rank Newbie
Rank
Newbie

I suggest testing with a smaller IDE hard drive from that time period. Remove all the variables and newer hardware. Although you may have a custom bios, the driver is from the 90s, talking to a chipset that was designed in the 90s, that expects to talk to a drive from the 90s. Unpredictable things happen when we mix-in hardware that's 20 years newer. Even though it "should" work, that isn't always the case. I would re-test with, say, a ~10gb IDE hard drive and see if it still hangs. Although it shouldn't be required on that system, i've had some Pentium-class systems where Dynamic Drive Overlay fixed weird compatibility issues. If it flies with that, then likely the size is an issue somewhere along the line of code between the driver and chipset. I don't think it would be anything in the signalling between the controller and drive, but then again, it's possible. Back then, we had a LOT of chipsets and hardware on the market and everything was changing rapidly. Some system designs were more clean and standardized, while others were a mess.

I know there's a strong drive to put huge SSD drives in old systems and I get that storage is cool, but in my own testing I have found that weird things happen when hardware from different generations is mixed together like that. For my collection and in my play lab, I do use SSDs. For pre-Pentium systems, i've standardized on Memtech AT3515 1GB IDE/PATA solid state drives and my images include the DDO layer because...it just works. The Memtechs look and act just like the old 1.2GB 3.5" drives of the time and all my testing has showed the systems don't know the difference. On Pentium and newer systems, given that 8-20GB was common at the time, I have had luck with SATA SSD modules, combined with *some* IDE-SATA adapters (mainly the red Startech ones). The Transcend 32GB 2.5" IDE SSD drives are great......if you can find them. They are usually quite pricey though, compared to my 16GB Apacer modules for $1ea in bulk.

But yeah, if you haven't already, I would remove all variables and the SSD, try again and see what happens. Please post the result as I am writing some master install media for my museum currently, and I am about to tackle the task of matching-up the best old drivers for various chipsets, to bundle with the media.