VOGONS


Reply 101 of 159, by feipoa

User metadata
Rank l33t++
Rank
l33t++
Rio444 wrote:
Show quote
feipoa wrote:

It might be sufficient to test the motherboard without needing the emulator card. Simply boot with your motherboard and disable COM1 in the BIOS. Then boot into Windows 95 to confirm that COM1 has vanished. Then emulator card should run on COM1 without any problem.

OK, I'll try it.

feipoa wrote:

Did you confirm if the DOS and Windows drivers would work as-is if they used IRQ's 2, 12, or 7?

No, I haven’t done it yet. I need more time for this.
Today I tried to find a motherboard with WinAMI BIOS. But all boards with any AMI BIOS turned out to be 486 or older, without built-in COM ports. :depressed:

Nearly all 486 PCI motherboards have built-in COM ports.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 102 of 159, by Rio444

User metadata
Rank Member
Rank
Member
feipoa wrote:

Nearly all 486 PCI motherboards have built-in COM ports.

Yes, and at the same time, almost all of them have Award BIOS.

I searched the Internet and found out that there are very few boards with WinBIOS and built-in ports.
Here, at Vogons, there is even a topic about motherboards with WinBIOS https://www.vogons.org/viewtopic.php?t=52562&p=564673
But I note that no one can guarantee that the reason is WinBIOS. This is just an assumption.
I have already sent out more than a dozen adapters. August is the vacation season, now is the beginning of school. I think that 2-3 weeks will pass and we will get much more test results.
I hope you find the time and empty HDD and install fresh Win95 on it. This will at least give confidence that the problem is software, or vice versa, hardware.

Reply 103 of 159, by feipoa

User metadata
Rank l33t++
Rank
l33t++
Rio444 wrote:
Show quote
feipoa wrote:

Nearly all 486 PCI motherboards have built-in COM ports.

Yes, and at the same time, almost all of them have Award BIOS.

I searched the Internet and found out that there are very few boards with WinBIOS and built-in ports.
Here, at Vogons, there is even a topic about motherboards with WinBIOS https://www.vogons.org/viewtopic.php?t=52562&p=564673
But I note that no one can guarantee that the reason is WinBIOS. This is just an assumption.
I have already sent out more than a dozen adapters. August is the vacation season, now is the beginning of school. I think that 2-3 weeks will pass and we will get much more test results.
I hope you find the time and empty HDD and install fresh Win95 on it. This will at least give confidence that the problem is software, or vice versa, hardware.

I used a clone of the NexGen hard drive and run it on my IBM 5x86c system, which has an AWARD BIOS. When I disable COM1/COM2 on the AWARD/5x86c system and boot into Windows95, COM1/COM2 disappear from device mananger, but the LPT port got re-detected at boot. This indicates to me that the issue isn't my Windows installation. The problem is that Windows is re-detecting COM ports which may only be partially disabled, or Windows is unable to probe the BIOS to see that they were set as disabled.

There are several items in queue before I can start testing this in other AWARD and WINBIOS boards.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 104 of 159, by Rio444

User metadata
Rank Member
Rank
Member
feipoa wrote:

I used a clone of the NexGen hard drive and run it on my IBM 5x86c system, which has an AWARD BIOS. When I disable COM1/COM2 on the AWARD/5x86c system and boot into Windows95, COM1/COM2 disappear from device mananger, but the LPT port got re-detected at boot. This indicates to me that the issue isn't my Windows installation. The problem is that Windows is re-detecting COM ports which may only be partially disabled, or Windows is unable to probe the BIOS to see that they were set as disabled.

There are several items in queue before I can start testing this in other AWARD and WINBIOS boards.

The presence of another COM-port on the same IRQ does not interfere with the adapter. I showed it above.
If another COM port has the same address (3F8 for COM1), then the adapter will not work guaranteed. This is also checked more than once. Including this is written in this thread.
Perhaps the reason is the incorrect detection of the COM port with Windows. But this has nothing to do with the real address and IRQ of the COM-port or the adapter.
I was recently advised, try to disable PnP OS option in BIOS.
Maybe it will help.

Reply 105 of 159, by feipoa

User metadata
Rank l33t++
Rank
l33t++

I tried another AWARD 486 board and it was able to turn off the COM ports. When I finish up testing in the other threads, I'll pull out a few WINBIOS boards and see if they can properly disable the COM ports.

Although you mention that two COM ports can share the same IRQ, there may be some bug in the NexGen's BIOS, RS232, or SuperI/O which doesn't work well with this sharing. In which case, if my two other WinBIOS boards also do not disable COM ports properly via BIOS, and do not allow for sharing of an IRQ between the emulator card and an unused COM port, then I will perform a clean install of Win95c on one of these motherboards, even though I think there is almost zero possibility that the issue is with the Windows installation.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 107 of 159, by SirNickity

User metadata
Rank Oldbie
Rank
Oldbie

Interesting read so far. I have a WinBIOS board, but as usual, it's a VLB 486 with no on-board serial. I've always thought about integrated I/O being a Pentium-era thing. I guess late (PCI) 486 might have it as well, but I don't have any of those. That's a rare bird, I would think, that you would have no PS2, onboard serial, a WinBIOS, and no jumpers to disable the onboard peripherals. Some of those criteria tend not to overlap.

I would agree that a conflict is more likely to occur because of sharing the base address. IRQs (usually) only fire when an event of note takes place. An unused serial port isn't going to generate those events, so "sharing" that IRQ with a dormant port should not be problematic. But, having two devices on the same bus address would cause contention. It seems likely that Win95 might have special awareness of that underlying chipset (and/or a lack of awareness of the BIOS) and is programming the chipset irrespective of the desired BIOS configuration. Not quite sure how you would easily circumvent that behavior, and so disabling the port in WDM before allowing a hardware conflict seems like the only realistic workaround.

Reply 108 of 159, by Rio444

User metadata
Rank Member
Rank
Member

There is an assumption that this situation is caused ISA PnP. BIOS disables the port, but leaves it possible to reinitialize COM-port with Windows using PnP function of the chipset.

I re-read the topic. And I did not figure it out, in this situation Feipoa tried to remove the disabled COM1 and restart the computer? Does it arise again as COM1?
file.php?id=67656&mode=view

Reply 109 of 159, by feipoa

User metadata
Rank l33t++
Rank
l33t++

When I wrote “root” in the photo it should be “rebooted”. To answer your question, the system hung at reboot. So I cannot view device manager.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 110 of 159, by Rio444

User metadata
Rank Member
Rank
Member

That is, the system crashed during reboot after removing the disabled port COM1 in the list of devices?
And if you reboot it in safe mode? What ports are displayed?

Reply 111 of 159, by feipoa

User metadata
Rank l33t++
Rank
l33t++
Rio444 wrote:

That is, the system crashed during reboot after removing the disabled port COM1 in the list of devices?

Yes.

Rio444 wrote:

And if you reboot it in safe mode? What ports are displayed?

Safe Mode also hangs.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 113 of 159, by feipoa

User metadata
Rank l33t++
Rank
l33t++

"Boot with PnP OS" is enabled.

Too much time has passed and I do not recall if I tried "PnP OS" in the disabled position.

I vaguely recall needing to leave this feature enabled to get my IBM ISA ethernet card working properly. The NexGen system was difficult to get running properly so I don't want to test this option now. I can look at this again when I have time to test my other WinBIOS motherboards w/integrated RS-232. The HOT-433 and M919 were/are very popular boards with WinBIOS and integrated RS-232, so I'm sure more test results will emerge. I don't have time right now to test these other motherboards.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 114 of 159, by csoren

User metadata
Rank Newbie
Rank
Newbie

I am happy to report the card works out of the box with the NuXT. I only changed the jumper to COM2 so it wouldn't conflict with the onboard serial port. I loaded CuteMouse and then tested it with CheckIt included with the NuXT. Flawless! Awesome job Rio444!

IMG_20190921_155636.jpg
Filename
IMG_20190921_155636.jpg
File size
260.96 KiB
Views
256 views
File license
Fair use/fair dealing exception
IMG_20190921_155654.jpg
Filename
IMG_20190921_155654.jpg
File size
215.72 KiB
Views
256 views
File license
Fair use/fair dealing exception

Reply 115 of 159, by Rio444

User metadata
Rank Member
Rank
Member
csoren wrote:

Thank you very much!
Please, if possible, test the adapter in different programs and games.
One of the buyers found that the adapter with XT requires the selection of the driver and the settings of the adapter. Perhaps the reason is poor compatibility with the adapter of his hardware or mouse.

Reply 117 of 159, by feipoa

User metadata
Rank l33t++
Rank
l33t++

I attempted to get my PC Chips M919 and my Shuttle HOT-433 v1 motherboards working today as they both have the WINBIOS. Unfortunately, both boards, after sitting for years in storage, aren't working very well. They are very flakely. Both are having issues reading floppy drives, they hang on boot up, and the HOT-433 board refuses to work with my KVM for even the keyboard. I've never had the best experience with these two boards, but at one point, I was at least able to boot Windows with them. I'll see if I have other WINBIOS boards to test. I should also have a v4 Shuttle HOT-433.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486

Reply 118 of 159, by feipoa

User metadata
Rank l33t++
Rank
l33t++

Rio444, I was able to test your ps/2-to-serial emulator card with another AMI WinBIOS motherboard, the Shuttle HOT-433 v4 with BIOS version 433AUS33. This motherboard does NOT exhibit all the same problems that my NexGen AMI WinBIOS board did.

I set in the BIOS COM1 to disabled. Serial emulator card is NOT plugged in yet. I boot Windows 95c with a serial mouse on COM2 and open the Device Manager. I still see COM1 listed. If I delete COM1 and reboot, Windows finds COM1 again (still no emulator card installed yet). So this symptom is the same as on my NexGen motherboard.

The difference is that when I install the serial emulator card and boot Windows (BIOS COM1 still disabled), it boots fine. With the NexGen board, it would hang because of the two COM1 ports. With the HOT-433 motherboard, I boot Windows with the emulator card installed, I still only see one COM1 port. No need to disable the motherboard's COM1 as it doesn't show up in the Device Manager list.

But as in the previous paragraph, if I remove the emulator card and leave COM1 disabled in the BIOS, Windows 95 Device Manager still sees the motherboard's COM1.

So it appears as if the HOT-433 motherboard's BIOS does a better job at disabling COM1, but Windows 95 still sees it, though seems to ignore it when the emulator card is installed. Interesting behaviour.

HOT_Shuttle_433-v4_PS2_emu_test_motherboard.jpg
Filename
HOT_Shuttle_433-v4_PS2_emu_test_motherboard.jpg
File size
411.35 KiB
Views
157 views
File license
Fair use/fair dealing exception
HOT_Shuttle_433-v4_PS2_emu_test_WinBIOS.jpg
Filename
HOT_Shuttle_433-v4_PS2_emu_test_WinBIOS.jpg
File size
308.91 KiB
Views
157 views
File license
Fair use/fair dealing exception
HOT_Shuttle_433-v4_PS2_emu_test_DevMgr.jpg
Filename
HOT_Shuttle_433-v4_PS2_emu_test_DevMgr.jpg
File size
405.84 KiB
Views
157 views
File license
Fair use/fair dealing exception
Last edited by feipoa on 2019-09-26, 10:20. Edited 1 time in total.

Ultimate 486 Benchmark | Ultimate 686 Benchmark | Cyrix 5x86 Enhancements | 486 Overkill Graphics | Worlds Fastest 486