Could be the CF/IDE, have you tried with just a regular small real IDE HD only on Primary ? Another cheat/solution would be a DIR C:\ (and then maybe a CLS if you don't want to see the Dir listing) added to end of your Autoexec. bat.

I had a 500MB Seagate HDD as primary, and the CF adapter as secondary, but the HDD died on me and why I went the two CF adapters way 😀 so no, I don't have an easy way to test, and my setup is working fine. It's actually not a problem per se, it's the smallest of inconveniences and working around it is easy enough (your suggestion being the most obvious). I'm really just curious and interested in trying to understand the why! 😀 thanks though!

There are some other topics here about some odd effects of CF/IDE on certain older hardware. Your thought about the secondary IDE bus not fully initialized at start up could be true. Only way to tell is swap out some parts or try same parts on different mobo to elminate BIOS issue but that is a lot of work if it all works great except that one quirk. Some of the early VLB controllers actually have that type issue when no HD is on second channel but only a cdrom, is an IDE-ATAPI to IDE bus issue afaik and some included a special driver to help initialize the secondary controller.

is the CD and CF both on the same cable?

have you tried another driver for the config.sys, like oakcdrom.sys?

I second that. Also try to not to use devicehigh=, but device=.
From what I remember, vide-cdd.sys doesn't like to be "high" all the time. It gets stoned quickly, I suppose.

Horun wrote on 2020-08-16, 21:34:

Only way to tell is swap out some parts or try same parts on different mobo to elminate BIOS issue but that is a lot of work if it all works great except that one quirk.

Yeah, when the HDD died and I replaced it with another CF adapter, I did actually try different setups like 1 CF + CD-ROM on first ribbon, and the other CF alone on the second ribbon, but BIOS wouldn't detect stuff correctly. The current setup works fine except that one quirk as you perfectly put it, and my curiosity is not worth the effort to meticulously try every combination again I'm afraid!

Warlord wrote on 2020-08-17, 00:26:

is the CD and CF both on the same cable?

It's in my first post: Two CF adapters on primary IDE, CD-ROM alone on secondary IDE.

Cobra42898 wrote on 2020-08-17, 02:01:

have you tried another driver for the config.sys, like oakcdrom.sys?

Not yet, will try!

Jo22 wrote on 2020-08-17, 02:05:

I second that. Also try to not to use devicehigh=, but device=.
From what I remember, vide-cdd.sys doesn't like to be "high" all the time. It gets stoned quickly, I suppose.

Haha 😁 Thanks, will try too.

Q. Why cannot I do a directory (dir) on a good ISO-9660 format disc on my CD-ROM drive although it is ready?
A. Your CD-ROM drive might work on PIO mode zero only. When your PC system is configured in a faster IDE interface mode, this CD-ROM drive might get this kind of communication problem. The solution will be that you have to turn down the interface speed between the CD-ROM drive and your PC system through BIOS or interface card setting. After the configuration is changed, you need to reboot the system.

Q. I have an enhanced IDE card with two ports, but I cannot install the CD-ROM drive on the secondary port properly.
A. The secondary port on some enhanced IDE card (eg. CMD0640x) is left un-opened when the PC system is initially powered on. Under such a condition, the CD-ROM device driver will not act stably when this CD-ROM drive installed. To resolve this problem, you can install the device driver bundled with your enhanced IDE card to open the secondary port.

Very interesting...found this two questions in the VIDE-CDD readme file, particularly the second one. I need to identify my VLB I/O card

Lots of good information on a link I found in a very old thread : http://www.mdgx.com/newtip1.htm#CDROM4
And the driver package itself that contains the readme: http://www.mdgx.com/files/VIDE-CDD.ZIP

well CD drives are ATAPI and I didn't realize you were using a VLB IDE and not all of those support ATAPI. Instead conect the CD rom to a sound card that has an IDE ATAPI header.

Here's my IO card.

From searching around it seems like it's not some obscure brand or model, but couldn't find any precise information about the jumper settings and potential drivers. Nothing in TH99 or stason.org.
An eBay listing with a .1 revision of the same board that looks exactly the same gives these alternative names, but cannot find anything either:

UMC 8287D01 (=SST-2946X =TK82872JD01.1)

I did find this: https://archive.org/details/umc8672
It uses the same UM8672F chip, and with a pretty similar jumper config. I may just try the DOS driver and see what happens.
Looking for chip information I found this: UM8672F / Function: VLB 2 EIDE ports sw/hw programmable upto PIO4 / Used: PT-627B, MI300.

As expected I've already spent way too much time on what is basically a non-issue, but hey it's part of the fun with retro hardware 😀

Warlord wrote on 2020-08-17, 09:01:

well CD drives are ATAPI and I didn't realize you were using a VLB IDE and not all of those support ATAPI. Instead conect the CD rom to a sound card that has an IDE ATAPI header.

Hey don't get me wrong but do you carefully read posts before answering to them? Really no offense, just find it amusing. After missing the only relevant details of my setup that were in the first post, now you're telling me to connect it to a soundcard when I made it perfectly clear all works fine apart from a little quirk that requires me to 'dir' once on C: to 'wake' the CD-ROM up after boot, so I don't think it's an ATAPI problem... or is it me who completely missed your point? 😉

Boohyaka wrote on 2020-08-17, 11:14:

Here's my IO card. io.JPG […]
Show full quote

Here's my IO card.
io.JPG

From searching around it seems like it's not some obscure brand or model, but couldn't find any precise information about the jumper settings and potential drivers. Nothing in TH99 or stason.org.
An eBay listing with a .1 revision of the same board that looks exactly the same gives these alternative names, but cannot find anything either:
UMC 8287D01 (=SST-2946X =TK82872JD01.1)

I did find this: https://archive.org/details/umc8672
It uses the same UM8672F chip, and with a pretty similar jumper config. I may just try the DOS driver and see what happens.
Looking for chip information I found this: UM8672F / Function: VLB 2 EIDE ports sw/hw programmable upto PIO4 / Used: PT-627B, MI300.

As expected I've already spent way too much time on what is basically a non-issue, but hey it's part of the fun with retro hardware 😀

Nice that it has the jumpers silk screened. I have a UMC based IDE card also but have not tried CF or Cdrom on it yet, is sort of a spare that I do know works very well with one HD, floppy and mouse (that was as far as tested).
The FCC ID returned nothing yet,,,,

Boohyaka wrote on 2020-08-17, 11:14:

Here's my IO card. […]
Show full quote

Here's my IO card.

io.JPG

From searching around it seems like it's not some obscure brand or model, but couldn't find any precise information about the jumper settings and potential drivers. Nothing in TH99 or stason.org.
An eBay listing with a .1 revision of the same board that looks exactly the same gives these alternative names, but cannot find anything either:

UMC 8287D01 (=SST-2946X =TK82872JD01.1)

I did find this: https://archive.org/details/umc8672
It uses the same UM8672F chip, and with a pretty similar jumper config. I may just try the DOS driver and see what happens.
Looking for chip information I found this: UM8672F / Function: VLB 2 EIDE ports sw/hw programmable upto PIO4 / Used: PT-627B, MI300.

As expected I've already spent way too much time on what is basically a non-issue, but hey it's part of the fun with retro hardware 😀

Are you using UM8672.SYS? If so perhaps try reversing the order in which you load it vs. the CD-ROM driver.
If you aren't using it, perhaps try. It may speed up the card as a bonus. However, I can tell from experience that um8672.sys ignores CD-ROM drives (I have hard drive an cd-rom on one cable) yet they work anyway. It's possible to speed up cd-rom access by consulting the Linux umc8672.c source to see how to program speeds on this card with your own assembly program.

Thanks man! I could only dream I'd be able to look at a Linux c source to write my own assembly program 🤣

In any case:
- DEVICE or DEVICEHIGH, no difference
- UM8672.SYS installed or not, no difference (but I installed it, may help with speed as you said!)
- UM8672.SYS before or after CD driver, no difference
- Switched VIDE-CDD.sys with TEAC_CDI.sys : bingo! with the added benefit that it feels faster, and allows for a /S speed switch. As my 4x drive recently died I've had to install a 52x drive I had laying around, so it's nice to be able to slow it down for less noise.

For curiosity's sake I did try all the other options with the TEAC driver (device/devicehigh, UM8672.SYS etc..) and no difference, it works in all cases, so bottom line it must really be a VIDE-CDD driver problem with my config.

Good news is I did find the manual for a very similar IO card that explains all jumpers positions, yay! It is there: http://computer-museum.net/docs/FEIOL-UC3Cusersmanual.pdf

Would anyone be able to explain the two SPEED jumpers though?

Mine is currently set to Option 1 (1-2 1-2), my cpu is a DX33 so I guess I should be running Option 3 (2-3 1-2) instead? Options 2 and 4 I really don't understand them as the explanation is mixing speed and feature with an "or"... What is their purpose and when should they be used?

Thanks!