I'm not quite sure I understand the need for the epoxy glue. I can see it to reduce stress on the PCB solder pads which are soldered to pins? Afraid that the insertion force will pry off the solder pad?

On my module, I tested it over and over again on my 386 board with AWARD 4.20 BIOS, but I could not get it working. So I don't know if my module is at issue, or this particular motherboard or BIOS has issues. I suppose I could test my module out on an AMI BIOS board and test your TSR.

So is the software a TSR or a driver? As far as I know drivers take up some memory, while TSRs do not.

Anonymous Coward wrote:

So is the software a TSR or a driver? As far as I know drivers take up some memory, while TSRs do not.

Any DOS driver is a TSR.
When you run a driver, it makes necessary checks, adjustments and, if success, leaves in memory a resident part - TSR.

feipoa wrote:

I'm not quite sure I understand the need for the epoxy glue. I can see it to reduce stress on the PCB solder pads which are soldered to pins? Afraid that the insertion force will pry off the solder pad?.

Yes, you are right. In addition, I am afraid of the cleavage of the SMD elements. I myself often pull out the KBC by inserting a thin screwdriver to its bottom.

feipoa wrote:

On my module, I tested it over and over again on my 386 board with AWARD 4.20 BIOS, but I could not get it working. So I don't know if my module is at issue, or this particular motherboard or BIOS has issues. I suppose I could test my module out on an AMI BIOS board and test your TSR.

Did you test my TSR?
I do not think that the problem is in the BIOS.
True, I did not check it with the so old Award bios.

Anonymous Coward wrote:

Rio444, did you test the turbo functions with this module installed? With my original prototype, I noticed that it interfered with the operation of CTRL-ALT +/- which toggles turbo on some boards. From what I have seen, this problem can likely be fixed with a copy of AMIBCP.

I just checked. Everything works great.
Below are the screenshots of the Norton System Information 6.01.
You can see the mouse cursor.
The motherboard is the same as in the photo above. It is M326 V.5.2.

Rio444, how thick is the PCB substrate and what material is it made from? I personally do not have any transparent epoxy, so it would be beneficial if you cover the surface with epoxy yourself. I can epoxy the PCB to my own KBC.

What material are the traces made of? Is it Aluminium?

Do you have a Holtek KBC to test with your PCB?

Sorry, I haven't tested your TSR yet. In fact, I haven't done any work on the KBC since I posted the photos of my module in 2015. I'm currently working on something else at the moment, but will test it out with my module in the coming days.

Did you note which AMI BIOS revision dates you tested your design with? Do you have a board with an AWARD 4.20 to test?

feipoa, It is ordinary single sided copper clad board with thickness 0,5мм. The tracks are copper. They are covered with tin-lead solder. This is the cost of simplified home technology.
I have three new VT82C42N left. Perhaps I can find Amikey-2 on one of the old faulty socket 7 motherboards. I met them quite often. If for you Holtek KBC is important, I will try to find it.
But I need to make another PCB. I did the first copy for a very long time. I hope the next will be faster. (In fact, it was the second sample, the first one was so terrible that I did not want to show it.)

I have several different models 386 boards, similar to the photo above. I chose the one because it already had a socket under the KBC. I also checked two 286 boards. They too are with KBC sockets. Unfortunately all these boards are with AMI bios.
One 486th board, which I tested, had Award 4.51 bios. I had to unsolder the KBC chip and to solder the socket. I have another 486th board with soldered KBC and AMI bios. I have not tested it yet.
Maybe I forgot about some boards. I will look for them. So now I only have one motherboard with Award 4.51 bios and some with AMI bios.

Rio444, I have pulled my module out of the box and will setup a test bed to test it with a motherboard containing an AMI BIOS. I will use your TSR. I assume I boot up DOS and run the *.com file at the DOS prompt? Or does it need to be loaded before everything else?

Did you test your mouse in Win311 or Win9x?

The Holtek is not necessary - any KBC which supports PS/2 mouse is fine.

feipoa, you must run ps2suppc.com before run mouse driver. I repeat link to ps2suppc.com https://yadi.sk/d/5ge32bgv3RoUHs (click "Скачать" in the right upper corner of the page).
I highly recommend CuteMouse driver rev.2.1 http://cutemouse.sourceforge.net/
It is very tiny.
If you mouse driver running with autoexec.bat, put string "ps2suppc.com" before string, which run the mouse driver.
Or just run ps2suppc.com and after it run a mouse driver.

I tested the module and the TSR in Win311 or Win9x. They work fine.

I assume you also tried DOS?

I've been using, I think, a DOS driver from Logitech, but it may be from MS. I don't recall at the moment; I'm not at the computer. It contains a little DOS test program to check movement, left, middle, and right mouse function. The DOS driver auto detects serial or PS/2.

I did not see Скачать in the upper right hand corner, but I did see the word "Download". Perhaps the website or the browser is doing some auto translation.

feipoa wrote:

I did not see Скачать in the upper right hand corner, but I did see the word "Download". Perhaps the website or the browser is doing some auto translation.

Probably so. It is perfectly!

feipoa wrote:

I assume you also tried DOS?

Oh sure!

I pulled my PS/2 KBC mouse adapter out of the storage bin today and tested it on 3 different motherboards. Unfortunately, it looks as if something has happened to it between now and when I used it last because all 3 motherboards show "Keyboard interface error". In the past, I was able to use the keyboard with my adapter. Rio444, any chance I can get an adapter from you for testing?

Of course there is a chance.
But I still can not promise any definite deadline.
Too many other things that I'm busy with.
As soon as the situation changes, I will contact you.

The same thing happened to my original prototype. I guess breadboard just sucks.

breadboard or solder board? The photo you provided early in this thread shows a solder board.

I might remove my KBC and make sure that is still working. Otherwise, the inverter IC might have died, or there is a severed connection.

I removed the KBC from my module and the KBC works fine alone in the socket, so something must be messed up on my module or that module's thinner and shorter leads aren't making contact with the motherboard's socket.

EDIT: I think I found the issue with my adapter module, however, fixing it will take some time. I'll try to get it fixed within a week and start testing the TSR.

OK. So the last time I used my adapter, I must have been moved a wire around because I found one wired in the wrong place. After correcting it, my PS/2 mouse adapter works fine now.

I have tested my adapter on only one computer thus far, a 386 based on the SiS Rabbit chipset. I loaded Rio444's driver, then ran mouse.exe and test.exe and everything seems to work fine, at least in DOS. I'll test it on other computers sometime this week, including in Windows 3.11.

Fantastic job on writing the driver file Rio!

Oh, and I did test my mouse adapter on a 386 board based on the VIA 495 chipset. I only tested the AWARD 4.20 BIOS mod and, unfortunately, the result is the same as it was some years ago - the PS/2 mouse did not get detected. I still need to test Rio's driver on this board though.

I am pleased to report that I am now able to use the PS/2 mouse attached to my KVM within Windows 3.11 and that there isn't any lag or ghosting of the mouse cursor. The issue I had with all serial to PS/2 mouse adapters was the issue of lag and ghosting when used behind my KVM.

I have still only tested one motherboard, the SiS Rabbit-based one. Unfortunately, I am no longer able to use the Logitech Mouseware v7.21 driver with my PS/2 mouse when using the KBC mod, neither in DOS, nor in W311. The DOS driver just doesn't find the mouse, even if you specify the PS/2 flag. In Windows 3.11, the mouse cursor disappears on boot. I must use the DOS mouse.com driver, which I think is either from Win3.11 or DOS 6.x, and I must use the Win3.11 mouse driver "Microsoft, or IBM PS/2" within Windows 3.11. The DOS and Windows Logitech drivers worked fine with my Logitech serial mouse and, if I recall correctly, with my PS/2 mice when connected to a 486 motherboard.

Lastly, does anyone know of a program for Windows 3.11 which can set the PS/2 mouse rate? I use PS2Rate v0.5 in Windows 95 to make the tracking smoother by setting the rate to 80 Hz. I think the default is 40 Hz. The version of PS2Rate that I have doesn't work with Windows 3.11.

Update1: Rio444's driver does not work with my AWARD 4.20 BIOS.

Update2: Mr BIOS for UMC 481/482 contains BIOS support for PS/2 mice. Adapter works in this board without needing Rio444's mouse driver. Upon adding the KBC module, the boot screen said "PS/2 mouse configuration change"

Update3: I am able to use the Logitech 7.20 drivers with MR BIOS for UMC 481/482 in DOS and Windows 3.11.

Update4: I tried the adapter in another UMC 481/482-based motherboard, an MSI MS-3131. Works just as well as in the previous UMC 481/482-based board.

Update5: Tried the adapter with the same MSI MS-3131 board, yet I used the board's original AMI BIOS. The TSR works, but as before, I could not use the Logitech 7.20 drivers.

Update6: Works with SiS460-based motherboard and AMI BIOS. Same as above.

Update7: Works with CHIPS-351/355/356-based motherboard and AMI BIOS. Same as above.

feipoa wrote:

Update1: Rio444's driver does not work with my AWARD 4.20 BIOS.

What does the TSR say after run?
There should be strings 'PS/2 mouse support is already enabled.' or 'PS/2 mouse is not found.' or 'PS/2 mouse support is enabled!'

Could you view the contents of memory at 0040:000Eh using MS-DOS utility DEBUG?
Run Debug and type on the keyboard 'D 40:E' Enter.
And write here first four byte.
Do it before run the TSR and after.

Here is the information you requested for the AWARD 4.20-based system with non-modified BIOS.

This is what the TSR says:

This is what my mouse driver says:

This is what the mouse test program says (mouse does not move, buttons do not work, computer is frozen):

This is what Debug D 40:E looks like (computer frozen? I cannot type anything again):