we don't do anything special with the joystick on the MK8330 - it is connected like the reference design dictates and it's actually the only way to do this connection...
IIRC 10 pins on the 8330 are the gameport pins and then the chip interfaces that to the ISA bus

could it be some conflict on port 201h that is the gameport address?

keropi wrote on 2023-04-10, 16:49:

we don't do anything special with the joystick on the MK8330 - it is connected like the reference design dictates and it's actually the only way to do this connection... 10 pins on the 8330 are the gameport pins and then the chip interfaces that to the ISA bus

could it be some conflict on port 201h that is the gameport address?

I don't think it's the Gameport because I had the same issue with Microsoft Sidewinder Pro USB (v3.02 software/drivers) and Gravis Gamepad Pro (USB & Gameport using old Gravis drivers and Gravis Gaming Experience later versions). I used version 2.00 CMI drivers, but I didn't test 1.64 or others.

I guess the port 201h is used also with other soundcards? If yes, gameport is ok on my setup with Creative CT4170 at least.

Wasn't there an old problem of unsifficient power to the gameport with the old CMI8330 cards, but i think this issue has been fixed with the MK8330?

201h is used by pretty much every gameport ever

Is there an onboard sound chip that's conflicting on the joystick port? If there is, see if it can be disabled in the BIOS.

this old power issue thing does not affect MK8330
I assume when you calibrate the joystick in some game/utility all look fine and responsive?

I tested wolf3d and alqadim on my 386DX/40

wolf3d indeed slows down music when you enable joystick
alqadim on the other hand works fine for me, I can select joystick , control the game, save - quit and load again with joystick controls being fine

I think wolf3d has some issue on certain combos, here is a person having the same issue with a CT1600: Wolf3D Joystick menu music slow down Question..
maybe it's a forgotten issue with it?

Hi Keropi.

Indeed, i have now Al Qadim working fine with a simple joystick instead of the Gravis GamePad Pro which might need a setup software i don't have. The Wolf3D issue seems unrelated to the soundcard as you have found.
Enjoying Al Qadim now!

The Gravis GamePad Pro is a digital device with more buttons than a traditional joystick. I believe it requires additional software to work.

Kahenraz wrote on 2023-04-16, 01:21:

The Gravis GamePad Pro is a digital device with more buttons than a traditional joystick. I believe it requires additional software to work.

The USB version doesn't need any software (unless for a Mac). Windows 98 recognizes all of the buttons out of the box. Star Wars Rogue Squadron 3D for example (which I have tested), worked great.

I'd have to double check on the Game Port version as I always install the GRIP software, which goes along with it.

I was assuming the gameport version, yeah.

I also want to point out that a lot of the cheap modern USB gamepads from China work fine with Windows 98. An NES and PS1 style controller I tested worked perfectly. I also hadsl success with a wireless Xbox style controller.

So if for some reason your USB Gravis isn't working, there are lots of other options.

Came across this project and might purchase one during the next in-stock batch if I have money whenever that will be, to replace the CT1750 in my 386SX-25 machine. but I had a thought: Aside from the PCMIDI add-on board, is anything else expected to release for the ISA expansion header? Specifically, one could put an XT-IDE on there, or alternatively a Lo-Tech ISA ROM board, thus freeing up an ISA slot in some systems that need a BIOS extension for whatever reason. in my case I need to use a 3C509B to load an XT-IDE bootrom so I can use more modern storage solutions, as my BIOS doesn't allow for anything but the predefined Type 1-47 and no custom definition. As the PCMIDI add-on is of no use to me, the addition of something else like this would be heavily ideal for a case like mine where I'm dealing with only two ISA slots available.

With something like either of those two solutions I wouldn't need to use the network card in that slot and could instead disable the onboard video and drop in a much faster graphics card, as would be my intention moving forward. I'm actively searching for ways to do that, and if that means ponying up some international currency, I'll do it. Honestly even a board that right-angles the expansion header into an 8-bit ISA slot and some smoothing capacitors would be helpful towards that... even sounds like something I could accomplish designing, if all I'm going to do is shove an EEPROM in there. gives me a reason to learn PCB layout software. Would also need to design something to support the add-on board, but that shouldn't be too hard either.

Sudos wrote on 2023-06-04, 16:31:

[...]
Aside from the PCMIDI add-on board, is anything else expected to release for the ISA expansion header?
[...]

the header is only wired for PCMIDI , it's not a 1:1 ISA header

keropi wrote on 2023-06-04, 16:42:

Sudos wrote on 2023-06-04, 16:31:

[...]
Aside from the PCMIDI add-on board, is anything else expected to release for the ISA expansion header?
[...]

the header is only wired for PCMIDI , it's not a 1:1 ISA header

oh oof. might be something to consider if a board rev is ever done, especially if the header is a mirrored ISA port even with the missing connections. I'd be plenty interested in that plus the aforementioned nonexistent right angle adapter to shove a boot ROM card on just to save on a slot, and I'm sure I'm not alone, but I'm definitely in a niche category.

a new batch of MK8330 cards is made, most likely it's the last ever batch and the plan is to retire the project afterwards (still offering support and service if needed)
so in case anyone is interested now it would be the time to get one

Get them before they are gone folks because they are A-MA-ZING.

appiah4 wrote on 2023-06-16, 18:43:

Get them before they are gone folks because they are A-MA-ZING.

And they are going...going...GONE!

I have a strange issue with the gameport on my MK8330. Wing Commander 1 and The Lion King as well as several Joystick test tools work fine, but when I start Tie Fighter or X-Wing CD-ROM collection it either doesn't recognise the Joystick or sometimes it works but I can only use it in a certain direction. Setup is a 386 DX 40 on a 4386-VC-HD mainboard with a TVGA 9000B graphics-card, MS Sidewinder 3D Pro Joystick and 8MB RAM. Is my 386 maybe too slow for some games to recognise the Joystick correctly in combination with the MK8330? The same setup but with a CPS Soundblaster 2.5 worked fine with the Sidewinder in Tie Fighter. I also have an ISA I/O card that has a gameport which I deactivated via its appropriate jumper (although that didn't stop the gameport of the Soundblaster 2.5 from working).

Not such a big deal since X-Wing CD-Collection and Tie Fighter are a bit too demanding for the PC anyway but got me curious.

[Edit]Just a heads-up: After upgrading the PC to a 486DX2-66 and 20MB of RAM, X-Wing and Tie-Fighter still didn't recognize the Joystick or only very rarely, sometimes after playing another game, but without any pattern I could make out. I've now installed an old Game Card with two Joystick ports and the MS Sidewinder 3D pro now works fine. I guess it may be an incompatibility of the MS Sidewinder that could be rather picky when it comes to gameports (it's even mentioned in the provided readme that it doesn't work with certain soundblaster models).

Btw - is there any way to disable the onboard gameport of the MK8330? The setup utility doesn't give an option to do so from what I could tell.

Death Rally 1.1 CMI8330 patch made by Jurgen added to the MK8330 website: http://pcmidi.eu/mk8330.html
it patches the game sound driver that has issues with SB16 mode and CMI8330 chip , more info in MKRALLY.TXT

thanks to Jurgen for sharing this !!!

Thanks for the patch, works like a charm!

At first I set DMA for 16-bit, instead of 8-bit, so the game started but I didn't have any sound. When I changed DMA to 8-bit (DMA 1 instead of 5 for my setup) the sound worked. Makes sense since only SB16 and up use 16-bit DMA and the patch disables the SB16 path.