I am beginning to think that the ps2 mouse header on this board is dead, and I have no other options for testing.

As I mentioned earlier, the ps2 to serial adaptors only work with dual mode mice that usually have a switch to select between serial and ps2 operation. They will not work with
more modern ps2 optical or ball mice.
Sorry if this is a stupid question, but I assume You have turned on the ps2 mouse function in the bios?

Have a read through this thread, might be something there that helps.
Think my mouse port died on my 4dps :( Options?

Concerning your cache issue, it is very important that you set TAG BITS to 8 when using L2 cache in write-through mode and set TAG BITS to 7 when using L2 in write-back mode.
Revision 1.50 BIOS has caused me problems. I suggest using my modded BIOS for this board.

There was no PS/2 mouse header standard. You need to measure out where each pin on your ps/2 DIN goes to on the header. The pin assignment printed in the Tomato manual is correct.

Another general observation with this board is that the I measured the 3.3V setting on this board at 3.35 V. This is too low for some faster CPUs, like the Cyrix 5x86-100.

This board prefers FPM RAM.

Imperious wrote:

As I mentioned earlier, the ps2 to serial adaptors only work with dual mode mice that usually have a switch to select between se […]
Show full quote

As I mentioned earlier, the ps2 to serial adaptors only work with dual mode mice that usually have a switch to select between serial and ps2 operation. They will not work with
more modern ps2 optical or ball mice.
Sorry if this is a stupid question, but I assume You have turned on the ps2 mouse function in the bios?

Have a read through this thread, might be something there that helps.
Think my mouse port died on my 4dps :( Options?

Thank you, Imperious, I missed the mention of the dual mode mouse issue. While I have worked with many computer systems from 1993 onward, 99% of them had standard ps/2 support (physical connector, rather than a motherboard header.) I believe that out of the myriad of mice I have ever owned, I never came across a dual mode mouse. I did not see a ps/2 option in the bios. I have been looking at serial optical mouse options on ebay, just in case I fail to resolve the PS/2 issue.

feipoa wrote:

Concerning your cache issue, it is very important that you set TAG BITS to 8 when using L2 cache in write-through mode and set T […]
Show full quote

Concerning your cache issue, it is very important that you set TAG BITS to 8 when using L2 cache in write-through mode and set TAG BITS to 7 when using L2 in write-back mode.
Revision 1.50 BIOS has caused me problems. I suggest using my modded BIOS for this board.

There was no PS/2 mouse header standard. You need to measure out where each pin on your ps/2 DIN goes to on the header. The pin assignment printed in the Tomato manual is correct.

Another general observation with this board is that the I measured the 3.3V setting on this board at 3.35 V. This is too low for some faster CPUs, like the Cyrix 5x86-100.

This board prefers FPM RAM.

Thank you for your reply! I will check my bios settings concerning the the cache. I am using a 486DX2, so I wasn't sure of the settings, though when I first tested the board with the now faulty 486 CPU, I think that the settings in question were enabled. My previous 486 board was an ALR Evolution X which didn't have all of the jumpers that the Tomato board uses (plus it had the dual PS/2 ports,) and it was rather "plug and play" for the time. I don't remember such options in the bios, but I could be wrong as it has been a while. The last 486 I had before the ALR was from around 1994! That board simply stopped posting, and I haven't been able to identify the issue.

I am trying to think of ways to measure out which wire is going to which pin on the PS/2 connector I have. I may have to strip it to look directly to wear each colored wire is going. Another option is to replace the 5-pin connector with individual housings for each wire and just experiment through trial and error. I assume that the red wired is the 5V, and the green wire is the ground. I did not get anywhere with swapping the blue and yellow wires.

It was suggested that I replace my bios chip with another, since I have not been able to flash my bios. While I have built many a PC and can usually figure out issues, I have never pulled chips from a board, nor would I know how to go about replacing my bios with a blank chip and putting a new bios on it (though I will be researching this.)

EDIT: according to MSD, IRQ 12 is reserved.

I do not recommend randomly swapping the the wire colours of the ps/2 mouse header around. You may damage your keyboard controller chip or your mouse. Get yourself a $5-$10 multimeter. If its a $5 multimeter, it probably doesn't have the 'beep' setting, as I call it, so you can just use the resistance setting to see if two points are the same node by seeing that they have a resistance lower than approx. 1 ohm. For a little more money, get a multi-meter that has a continuity setting, which will beep at you when the two probes are at the same node in a circuit. Use this 'beep' feature or the resistance feature to see where the wires on your ps/2 mouse header are relocating to. It sounds to me like you haven't determined the right combination.

The assumption that red is Vcc is not good. I've seen these ps/2 header cables with colour schemes all over the board. I'm looking at one now where green is Vcc (5V). On another one, orange is Vcc and red is Gnd. Just to drive this point in further, I picked up a third manual with my notes written in it and on that board, Vcc is yellow and red is Data.

It sounds like you have a UV eraseable BIOS chip. There are online services which will send you a flasheable BIOS and flash your BIOS ROM for you. I recall reading some people on this forum used such a service. The other option you can do is to hot-swap the BIOS chips while the board is on to write-in a BIOS image to a EEPROM chip (non-UV) on a different motherboard. I'm not an advocate of that method and feel that buying a BIOS writing aparatus the preferable course, e.g. a Wellon VP-390. Such devices are also nice for testing SRAM cache modules and writing code to various PLL, microcontrollers, etc.

IRQ 12 is reserved for the PS/2 mouse.

I have a multimeter somewhere, I will see if I can hunt it down. Looking at the manual, I am assuming that the mouse pin closest to the keyboard port is pin one. The thing is, I flipped the connector to make sure I didn't have it connected backwards, and the optical mouse lights up regardless of the way it is connected to the header. I think that, in this case, it is either the blue wire or the red wire that is the 5v. I know that this may not matter without knowing what the other wires are.

I will think further on the bios issue. It is definitely a UV eraseable. I removed the sticker and saw the round window. I never pulled one from a board, and, like everything else on this tiny board, there isn't a lot of room to work with. That chip precariously close to the first ISA slot. If I can resolve the mouse issue, I may stay with the 1.5 bios, unless I run into any other issues.

I did order an inexpensive optical serial mouse to use in the interim.

Thank you feipoa and Imperious for all of your help, suggestions, and information. I found my multimeter, and I was able to determine where the wires belong, and now CTMOUSE is detecting my ps/2 optical mouse. Now I can move forward with the project as now everything seems to be in working order.

Glad you got it working. Without a working PS/2 port, this motherboard isn't that interesting. What is your target CPU?

I'll keep a 486DX2-66 in there. This build is one of three for nostalgia, running games up to about 1994, along with WFW 3.11. It will be installed in a custom wooden box, soon (another reason why I love the small size.) It also has an MPU-IPC-T, and an ancient 10mb network card, though I may throw a 10/100 PCI card in there if I find one in my parts box that had dos drivers. It also is running with 16mb/ram.

The second build is a P-II system with a Voodoo2 (both run ATI Mach 64 cards for 2d, and have AWE32's with daughter boards.) The third build is an Athlon XP 3200+ WinXP box. I built quite of few of those back in the day, for myself and customers.

I'm going to eventually play with my old ALR Evolution X board, soon. Other than the size of the board, it was by far the best socket 3 board i ever worked with. I never see them on ebay. Like the Tomato, it supported DX4 CPUs. The best part was the dual ps/2 ports. The only drawback was that it didn't have a secondary IDE port, so i used the port on my AWE32 for the cdrom.

Great news You fixed all the issues. Now You can actually enjoy using it.

My 1st PC was a Intel 486 dx2-66, and even though I have a couple of AMD 5x86 133's, Somehow the 66mhz chip seems more right being in there.

If you are wanting to stick with an Intel DX2-66 chip, how about using one with the write-back L1 cache instead of the slower write-through L1 cache? These should be marked SX955 and were not nearly as common as the write-through chips. Some SX955 were gold capped some were black capped.

I will check to see which 486 chip I have, especially if write-back helps with Win98 (though I may throw Win95 on there.) This is primarily for MS-DOS and WFW 3.11, though sometimes it is easier to run Win9x just to move files around more quickly. The pins on my CF reader bend easily and I had to perform micro-surgery to straighten them, so I would rather not remove them from the CF-IDE card to move bulk files.

While this is a nostalgia box, I also cheated a little and put MS-DOS 7.10 on there, for convenience. Dual 4gb FAT32 cards without partitions is great! Now I need to hunt down a PCI 10/100 NIC with DOS drivers.

Anything with the RTL8139 chipset will work great!

I'll look in my parts bin to see if I have one!

I agree with BSA Starfire. Pull the motherboard back out and retest outside the case. It's possible there is an alignment issue in the case that is causing cards to not seat fully, or maybe a grounding issue somewhere. It's also possible that standoffs are not providing enough support such that the motherboard is flexed into a position that is causing loss of current through the board.

edit: sorry, I missed the part where you got it working! disregard 😀