Don't lose hope yet. Remove all cache SRAM, KB controller ICs from MB and retest it with a POST card, make sure that BIOS chip makes a good contact in its socket, also, measure RESET pin on CPU socket, while the MB is powered on and has a CPU installed.

@Tiido, RDOSUMB on SY-25P2 detects it to be with SiS-486 chipset. As the only difference between the N and P variant is VRM, i assume it's the same with SY-25N.

I think i did test it without any keyboard controller while trying to read out something from debug card. Only way to get any different result from the board was when i took out the BIOS rom. But later i confirmed that the culprit was not the BIOS, as i was able to read it easily and also burned fresh roms for the board and nothing like that helped...

I'll try the keyboard controller from the 25P board to see, what happens...

GigAHerZ wrote:

* All keyboard leds are on

Have seen this happen when the picofuse for the keyboard port has blown in a strange way, turning into a high resistor instead of open circuit.

Resurrection... of the thread. 🙁

Unfortunately with other keyboard controller, the behaviour is still same - all leds are lit up.

I also went through all cache chips. I now have TL866II Plus, which can also test SRAMs. All chips are good.
No beeps, debug card doesn't show any numbers/codes (only garbage value right from the beginning, if shutdown and powered on immidiately)
On debug card CLK is on, IRDY is off, FRAME is on, RESET is on at the beginning for brief moment, then off.

The KB fuse is okay, showing 0 ohms with multimeter.

@Tiido, yes, it's Sis471 board. (Y)

So if you have any other ideas to check, let me know! 😀

Is there any corrosion around the battery area? If so, please post a detailed pic.

There is a tiny bit greenish thing around as i haven't done a proper cleaning to it yet. But in general it looks healthy with no broken traces.
I did replace the battery when i received it to a brand new NiMH battery.

Current jumper settings are for 486SX2-50 that i have laying around so easily accessible for testing.

I have two such boards, both needed some repair near the battery. Keyboard LEDs staying on provide an important clue about the location of the problem - the keyboard controller. One or more of the smaller traces could be broken. Follow each of them, make a "map" of all the connections from the KBC, trace each visually, and with the multimeter check if all of them really connect to their endpoints.
There's no guarantee this will fix the board (well, if you find broken traces near the KBC it surely might), as there could be other problems too... Some broken via, bad solder joints, dead chipset or glue logic.

GigAHerZ wrote on 2020-01-04, 00:21:

There is a tiny bit greenish thing around as i haven't done a proper cleaning to it yet. But in general it looks healthy with no broken traces.
I did replace the battery when i received it to a brand new NiMH battery.

Current jumper settings are for 486SX2-50 that i have laying around so easily accessible for testing.

Hi there. Resurrecting this topic because I had a similar issue. Did you end up fixing your board? I recently purchased a non-working Soyo 025P2 from ebay and was able to fix it easily. The problem was a couple of missing factory-set jumpers. I can see from your provided picture that JP17 and JP27 are set different to mine. Please try setting these jumpers as per my picture below and see if that makes a difference.

Also you mentioned that you have BIOS version J1 (1995) for this board. I am currently on BIOS G2 (1994) and would like to try version J1. Where can I get this please? Thanks.

Hi Gulzyee!

The bios i got from my board is uploaded here: Re: Resurrecting an Socket 3 motherboard (Soyo SY-25N)
Other bios'es are also uploaded into this forum thread here.

I'll check the jumpers when i have a chance and will report back. Thanks!

I've got a fully working 25n if you need any pictures of it

Though, some time has passed, the board has not resurrected.
I've checked all the jumpers, tried different settings and i have 025P2 board side-by-side with it, yet this dead board stays dead - keyboard lights stay on, debug card shows nothing, etc. 🙁

Possibly only worth to salvage parts... 🙁

GigAHerZ wrote on 2020-12-28, 17:52:

Though, some time has passed, the board has not resurrected.
I've checked all the jumpers, tried different settings and i have 025P2 board side-by-side with it, yet this dead board stays dead - keyboard lights stay on, debug card shows nothing, etc. 🙁

Possibly only worth to salvage parts... 🙁

I recently started trying to resurrect my own. I started by clipping off the battery, then I had no codes on POST card. Reseated BIOS chip, then started getting codes. Stopped at a memory code. I had some 30 pin SIMMS so slapped them in there and got a POST screen. Tried again with 72 pin and nothing. I figured out that the "SIMM jumper configuration" had to be all set to "configuration 2" (2-3 JP35 /JP36/JP37). I believe configuration 2 is for 72 pin SIMM and configuration 1 is 30 pin SIMM.

I'll be honest though and mention that I don't remember if the keyboard light just stayt constant when I first started all this.

When I'm not selling any out of situations like yours and we're talking about 150 boards fixed to date, I usually start over.

remove cache chip !

- with the tester

1 checks the continuity between pin 1 and pin 39 of the kbd controller and the keyboard connector,
2 check the data bus between the kbd chip and the eprom should be in parallel

- with the oscilloscope by turning the board on and off at each check

1 - check the presence of the clock and the frequency on all the quartz
2 - check the various signals on the isa connector

- with your finger
press the two chips of the chipset alternately and turn on the board.
so check the lack of welds which happens quite broken.

your bios on x86Box

dataino.it
—
Today at 11:09 PM
!bios
Attachment file type: unknown
original_025n2.BIN
64.00 KB
86BotBOT
—
Today at 11:09 PM

Vendor: Award
Version: v4.50G
String: 09/26/94-SIS-85C471B/E/G-2C4I9S23-00
Sign-on: REV .G2.

86BotBOT
—
Today at 11:09 PM

Type: Award
Date: September 26 1994 (09/26/94)
Identifier: SIS-85C471B/E/G
Core version: v4.5x (2)
Bus: VLB + ISA (C)
CPU: 486 (4)
Chipset: SiS 85C471B/E/G (4I9)
Manufacturer: Soyo (S2)
Project: 3-00

Hello! I have Soyo SY-S25p2 and the post problem was the factory jumper that was missing. After I install jumper, mobo start to give post codes but stopped to number 20. After several attempts I have got picture one time and saw in screen bios rev G2 and"unknown processor at 33MHz" ( I used dx2 at 66)!!! After I reboot it kept stopping at number 20 (post card) with no screen! I also got no sign of kb leds flashing. Do u believe that bios version is wrong? Because G2 is for 025p/M/N and not for 25p2.....

Hi keeper5511,

All boards from the SY-025M/N/P series share the same BIOS, so your Rev G2 BIOS is the correct BIOS for the 025P2 as well.

But there must be something very wrong with this board or the CPU, if you get an “Unknown” indication for the DX2 processor. Also POST code 20 is not a valid code for an Award v4.50G BIOS. The normal POST code sequence in this BIOS is from 03 to 1E and then from 30 to 63, with some specific Bx and Cx codes in between.

Is your 486DX2-66 from Cyrix? This CPU needs different jumper settings than the DX2s from AMD/Intel. Check also if your CPU needs 3.3V or 5V and check if the voltage jumpers are set accordingly.

Let us know what develops.
Cheers, Jan

My CPU is INTEL dx2 at 66Mhz (5V) and is full fuctional. So why in retro web for Sy-025p2 has diffirent bios???

https://theretroweb.com/motherboards/s/soyo-sy-025p2#bios

It also stuck sometimes at E6 error code

keeper5511 wrote on 2025-09-10, 21:36:

My CPU is INTEL dx2 at 66Mhz (5V) and is full fuctional. So why in retro web for Sy-025p2 has diffirent bios???

https://theretroweb.com/motherboards/s/soyo-sy-025p2#bios

It also stuck sometimes at E6 error code

The separate SY-025P2 page on TheRetroWeb is strange. The SY-025M/N/P series page already covers the SY-025P2 as well and the same Rev G2 and Rev G3 BIOSes are used on all variations of the 025M/025N/025P boards.
I sent a request to TRW to merge these to pages and their BIOSes.

Note that the figure behind the letter only indicates the amount of L2 cache that was on the board when it left the factory. So an SY-025P2 board has 256KB L2 cache and an SY-025P5 has 512KB L2 cache. For the rest, these 025P boards are identical.

The E6 POST code is also an unknown one for your BIOS. I don’t know what is going on but does your POST analyzer card have a button to scroll back through all recorded POST codes? If so, the list of codes may help to find the problem.
Also a photo of your present jumpers setting may help.

Jan

Thqnk you very much for your help! Jumbers have been checked multiple times allready. I have noticed one more thing. Keyboard IC controller seems to be different from retro web as I found motherboard in a scrapyard. It has NEC Japan chip that seems to have different datasheet from some others. I also dont get 3 leds flashing when I first power the motherboard (5V comes to the connector). Is there a possibility this IC generates all the problems?

In conclution: Post stop at 20, sometimes at E6 code. One time get picture (G2 Bios) for 2 seconds and sometimes I get monitor logo but no picture!!! Post screen wrote unknown processor at 33 and I have dx2 at 66MHz!!!

I have no ready explanation for the 33MHz CPU speed indication on your i486DX2-66. Possibly there is something wrong with the Chrontel clock generator or its 14.318MHz crystal next to the CPU socket, so that it outputs 16.5MHz instead of 33MHz.
Or the CPU is seriously slowed down somehow. On 486 CPUs, the BIOS measures the CPU speed via instruction timing.

About the strange keyboard controller, you can try a boot with this IC removed. The BIOS will then probably hang as well, but hopefully at a different point and with a meaningful POST code, so we know what works up to this point.
Also check for trace damage between the Keyboard connector and controller IC. These traces run along the CMOS battery and if the battery has leaked, corrosion will have eaten these traces.

Jan