Still has exclamation points. Only thing that was out of place during install is that the instructions says to remove all IDE controller devices, but it would only let me removed one of the three. "Standard Duel PCI IDE Controller". The primary and secondary won't let you remove them.

Damn! I was hoping that would work, oh well.

@rasteri

I don't suppose you'd be willing to bring out your Wee86 build with the serial port to do some debugging of this?
The Windows 98 DDK has a debug build of the ESDI_506.PDR driver with .SYM symbols that could help identify what is causing this issue if you're interested.

@rasteri

This is the debugging version of the ESDI_506.PDR driver with SYM symbols.

Duffman wrote on 2022-10-18, 05:05:

I don't suppose you'd be willing to bring out your Wee86 build with the serial port to do some debugging of this?

Unfortunately I don't think the vortex86SX in the wee86 supports booting from SD cards.

I think I have another 86DX board somewhere, though... I'll have a look

Has anyone tried my modded VGA driver? The .inf file in the MOD folder?

It might give us OpenGL/Direct acceleration in Windows 98 assuming it works.

@rasteri @Nachtrave

OK, My next IDE mod driver attempt for the WeeCee.

I've gotten a better understanding of how 95/98/ME INF files differ from INF files in2000/XP, among other things, so I do hope this one works. Please try it.

@rasteri @nachtrave

Do you guys also have a working LAN driver on Windows 98? or should I try and port the XPe driver for it?

Duffman wrote on 2022-10-20, 05:35:

Do you guys also have a working LAN driver on Windows 98? or should I try and port the XPe driver for it?

Existing LAN driver seems to work fine.

Sorry I've not had a chance to test your other stuff, I will this weekend.

@rasteri

Ah, I missed the 98 LAN driver being there on the page you linked.
I must have scrolled past it as I originally only saw the XP embedded version.

Also, can you confirm wether or not the single USB port on the WeeCee is 2.0? 1.1? or 1.0?

Duffman wrote on 2022-10-20, 12:28:

Also, can you confirm wether or not the single USB port on the WeeCee is 2.0? 1.1? or 1.0?

It's USB2, but I'd be surprised if the traces that I've routed can do high-speed.

@rasteri

Cool,
I'll run a benchmark on that port once I have my WeeCee.

I think the highest bandwidth I've seen on a USB 2.0 port in windows is about 25MB/s - I'll be pleasantly surprised if it reaches that.

Hate to have to rehash old posts in this thread - but on SN11 I'm making for Duffman here, I cannot for the life of me get this Audio chip to initialize. I've already replaced the crystal chip (twice now) as well as the EEPROM. I've tried everything I can, but I just cannot establish CONTROL connection to the chip.

I know @640kisenough! has mentioned the difficulty with the CS4237B, with it getting into a stuck state somehow. I think I might need some of that magic. I've tested the board now to the point I know that the right connections should be getting made. I am starting to think that it is less an electronics issue and more a software issue, alas. Everything I have tested on board shows it to be working, but the Crystal chip is just refusing to talk. Help?

Update: Solved, was a bridged pin on SOM header.

NachtRave wrote on 2022-10-21, 11:27:

Hate to have to rehash old posts in this thread - but on SN11 I'm making for Duffman here, I cannot for the life of me get this Audio chip to initialize. I've already replaced the crystal chip (twice now) as well as the EEPROM. I've tried everything I can, but I just cannot establish CONTROL connection to the chip.

I know @640kisenough! has mentioned the difficulty with the CS4237B, with it getting into a stuck state somehow. I think I might need some of that magic. I've tested the board now to the point I know that the right connections should be getting made. I am starting to think that it is less an electronics issue and more a software issue, alas. Everything I have tested on board shows it to be working, but the Crystal chip is just refusing to talk. Help?

If the software error didn't exist for the previous 10 weecees it probably isn't a software issue now (assuming you're using the same software). Almost certainly a soldering error, or maybe the PCB itself has a broken trace or something (it happens)

Try another SoM on it? Harvest the components and put them on a new board?

I'm by no means an expert on the CS4237B like e.g. 640K!enough, but on my cards, at least on the first programming of the eeprom, I had to always force the port setting for the control connection in the utility, otherwise it wouldn't detect it.
If that's what you're already doing, then please ignore 😀

What I personally would try then: removing the eeprom and see if the card is detected (by e.g. unisound) in its default state then - or: putting an already programmed eeprom on the card / chip in question and see if that helps. After that, I'd probably check basic things like the crystal, reset, datalines, etc. with an oscilloscope, to see if there's some life in it - and if that's leading nowhere, it's probably really a job for the magicians of the CS4237B or the PCB / SoM / soldering in other areas is to be questioned for defects.

@rasteri

have you had the chance to test my most recent IDE driver mod?

Duffman wrote on 2022-10-22, 11:26:

have you had the chance to test my most recent IDE driver mod?

Yep. Still no worky (Code 10, this device is either not present etc).

The Display drivers don't work either, I'm limited to 640x480x256. Honestly though I'm pretty sure the Volari Z9 doesn't have 3D accelleration, it's a chip for servers etc...

@rasteri

Can you post some screenshots of the RDC storage controller in device manager? One of it showing the error, one of the Driver File Details, and one of the resource allocations (if any)?

Does Device Manager show any error message with my driver on the Volari Z9? If not then try running a game with some 3D acceleration?

@rasteri @Nachtrave

OK, for this one I modified the INF to change the device class from SCSIAdapter to HDC as well as adding IDE Channel IDs.

For this one, first remove any existing driver installed earlier for the RDC Storage Controller and then restart windows.

After the restart, when windows prompts to install a driver for the IDE driver point it to this one.

Fingers crossed - hopefully this one works!

EDIT: Some Promising news! I Successfully managed to get the ReactOS UNIATA driver to load in a VM! - I'll upload it here shortly for you guys to try out.

UNIATA Working in VM.

UNIATA98 Attached.

UNIATA98 - INSTRUCTIONS

1ST. GO INTO DEVICE MANAGER, EXPAND HARD DISK CONTROLLERS AND REMOVE "Standard Dual Channel PCI IDE Controller"
WHEN PROMPTED TO RESTART - DO NOT RESTART!

2ND. PLACE UPDATED NTKERN.VXD IN C:\WINDOWS\SYSTEM

3RD. RENAME MSHDC.INF IN C:\WINDOWS\INF TO MSHDC.ORG TO STOP IT INTERFERRING WITH UNIATA98.INF

4TH. NOW YOU CAN RESTART - WHEN PROMPTED TO INSTALL IDE CONTROLLER POINT TO UNIATA98.INF

MJay99 wrote on 2022-10-21, 11:51:

I'm by no means an expert on the CS4237B like e.g. 640K!enough, but on my cards, at least on the first programming of the eeprom, I had to always force the port setting for the control connection in the utility, otherwise it wouldn't detect it.
If that's what you're already doing, then please ignore 😀

Yup, I've definitely heard of people having to do that. Have tried that at least a dozen times, for both 538, 120, and others. Nothing seemingly has so far worked.

Update: Solved, need to run unisound after fixrdp to re-init the board.

MJay99 wrote on 2022-10-21, 11:51:

What I personally would try then: removing the eeprom and see if the card is detected (by e.g. unisound) in its default state then - or: putting an already programmed eeprom on the card / chip in question and see if that helps. After that, I'd probably check basic things like the crystal, reset, datalines, etc. with an oscilloscope, to see if there's some life in it - and if that's leading nowhere, it's probably really a job for the magicians of the CS4237B or the PCB / SoM / soldering in other areas is to be questioned for defects.

Update: Solved, was a tiny bridged connection on the SOM header.

On another note - Did you all see how much our SN9 build went for? 1k USD. It's a good 650-700$ to make one, more if you don't have some of the more specific little things like the right sized hex standoffs. I've been able to make costs with this guy and from here on out it's all "profit" (read as: rent money).