goes crazy at [3D 1.Initialise keyboard] post code
responsible chip lower right corner https://pdf1.alldatasheet.com/datasheet-pdf/v … ND/W83627G.html
more cards more stable might just mean plugging one of the cards flexed the board slightly - try pushing this chip down with finger and then start computer (while pushing), or get a loupe + needle and go over every single pin

another option is more cards = more load = cleaner power, would need scope to check

does this same ps2 keyboard still work in another computer?

I'm 99% confident the keyboard works fine, it did when the bx master came back to life.
The super I\o was my guess (or let's say hope).
I tried to push it before and while turning on the machine, nothing changed.
I took a picture, I still think the damage is inside the chip: https://i.imgur.com/8pXjYKZ.jpg

Let's say the culprit is the adapter (since seriously, it was the ONLY peripheral in common between the Tusl2-c and the msi bx master).
I do wonder, is the ide port in any way connected to the Super i\o? Is the speaker connected to the super i\o?
I also noticed that the bios got corrupted (it was this morning) after I changed some settings (mainly disabled com, lpt and ide controllers).
So I need to find another WINBOND W83627GF AW, the similar models: W83627HF-AW or W83627HG-AW, would be ok anyway?

Sounds to me like the motherboard has caps that are bad. You shouldn't be hearing a buzz from the PC Speaker unless the power delivery is very dirty or you have something shorted from power to ground.

Could be bad caps. Could be a bad voltage regulator or diode.

Could also be chipset but I don't see how something on the IDE controller affecting the USB or PS2 ports unless something is shorted between them.

Also don't see how a CF to IDE adapter would cause that.
CF is basically just micro IDE. The power from the PSU is just to power the CF card and the LEDs.

There are a couple diodes on that adapter. You can always check to see if one of them has low or 0 resistance in both directions.

cyclone3d wrote on 2023-03-05, 19:07:

Sounds to me like the motherboard has caps that are bad. You shouldn't be hearing a buzz from the PC Speaker unless the power de […]
Show full quote

Sounds to me like the motherboard has caps that are bad. You shouldn't be hearing a buzz from the PC Speaker unless the power delivery is very dirty or you have something shorted from power to ground.

Could be bad caps. Could be a bad voltage regulator or diode.

Could also be chipset but I don't see how something on the IDE controller affecting the USB or PS2 ports unless something is shorted between them.

Also don't see how a CF to IDE adapter would cause that.
CF is basically just micro IDE. The power from the PSU is just to power the CF card and the LEDs.

There are a couple diodes on that adapter. You can always check to see if one of them has low or 0 resistance in both directions.

That would be right if the problem was only with the asus, but since it did the same damage to another motherboard, also the asus was working fine until 3 days ago, in 3 days can something so catastrophic happen inside a box?

I'm not sure those things are inductors:

I still measured them, they all have values ranging from k to m.

Which caps have gone bad in your opinion?
It could be an hypothesis considering that the buzz are connected to light blinking on the debug card.
I'm willing to take measures of things if I don't have to desolder, like measure voltages and other stuff if necessary.

Nemo1985 wrote on 2023-03-05, 18:25:

I'm 99% confident the keyboard works fine, it did when the bx master came back to life.

might be intermittent

Nemo1985 wrote on 2023-03-05, 18:25:

Let's say the culprit is the adapter (since seriously, it was the ONLY peripheral in common between the Tusl2-c and the msi bx master).

same keyboard both times?

Nemo1985 wrote on 2023-03-05, 18:25:

I do wonder, is the ide port in any way connected to the Super i\o?

no, entirely different bus (lpc = kinda serial ISA)

Nemo1985 wrote on 2023-03-05, 18:25:

Is the speaker connected to the super i\o?

yes, but beeping is most likely just bios letting you know about problem with keyboard buffer (full etc).

Nemo1985 wrote on 2023-03-05, 18:25:

I also noticed that the bios got corrupted (it was this morning) after I changed some settings (mainly disabled com, lpt and ide controllers).

cmos or bios?
cmos is not stored in this chip

Nemo1985 wrote on 2023-03-05, 18:25:

So I need to find another WINBOND W83627GF AW, the similar models: W83627HF-AW or W83627HG-AW, would be ok anyway?

afaik no, even one letter matters with those chips.

I still think its the keyboard 😀

Those rectangles on IDE-CF converter are diodes. In my experience Asus used rock solid caps and I havent seen Asus board in need of recap yet.

Older boards, including ASUS, during the capacitor plague had bad caps.

rasz_pl wrote on 2023-03-05, 20:00:

might be intermittent […]
Show full quote

Nemo1985 wrote on 2023-03-05, 18:25:

I'm 99% confident the keyboard works fine, it did when the bx master came back to life.

might be intermittent

Nemo1985 wrote on 2023-03-05, 18:25:

Let's say the culprit is the adapter (since seriously, it was the ONLY peripheral in common between the Tusl2-c and the msi bx master).

same keyboard both times?

Yes.
Uhm i'm puzzled, how can a keyboard damage a motherboard? I mean the bx master was working "fine" after I removed the adapter and still used the keyboard.
What test do you suggest?

rasz_pl wrote on 2023-03-05, 20:00:

no, entirely different bus (lpc = kinda serial ISA) […]
Show full quote

Nemo1985 wrote on 2023-03-05, 18:25:

I do wonder, is the ide port in any way connected to the Super i\o?

no, entirely different bus (lpc = kinda serial ISA)

Nemo1985 wrote on 2023-03-05, 18:25:

Is the speaker connected to the super i\o?

yes, but beeping is most likely just bios letting you know about problem with keyboard buffer (full etc).

Nemo1985 wrote on 2023-03-05, 18:25:

I also noticed that the bios got corrupted (it was this morning) after I changed some settings (mainly disabled com, lpt and ide controllers).

cmos or bios?
cmos is not stored in this chip

Bios chip, I took it off because it stopped booting and when I did a verify test with the TL866IIPLUS said that there were like 1500 difference between the stored image and the bios on my daily pc.
The cmos may also be corrupted but silly asus doesn't have the jumper for clear cmos and anyway without battery and psu connected should loose the settings in no time.

rasz_pl wrote on 2023-03-05, 20:00:

afaik no, even one letter matters with those chips.

I still think its the keyboard 😀

Those rectangles on IDE-CF converter are diodes. In my experience Asus used rock solid caps and I havent seen Asus board in need of recap yet.

Damn they nowhere can be found.
I agree, that era of caps on the asus were good, as cyclone said during the plague (p4, I saw several asus boards with leaking caps). While I do not agree with the keyboard, I mean after flash the bios that keyboard hasn't been connected to the pc, still...
How can I troubleshoot a killer keyboard without connecting it to a pc?

Nemo1985 wrote on 2023-03-05, 19:33:

I'm not sure those things are inductors: 002.jpg 001.jpg […]
Show full quote

I'm not sure those things are inductors:
002.jpg
001.jpg

I still measured them, they all have values ranging from k to m.

Those things are diodes, and they are there because it was cheaper than putting on an actual 3v3 voltage regulator.
The two diodes should be in series and they each have a voltage drop of 0.6~0.7V, combined 1.2~1.4V, which delivers a voltage of 5-1.4 = 3.6V
As a voltage dropper, diodes are the worst possible solution and should be avoided at all cost.

Luckily they should only be in use when the jumper for 3V3 is selected.
Though unluckily it seems that they somehow still connect to one pin on the main ide/ata connector from what I gather from the measurements you posted earlier.

killer keyboard ? might be as easy as failing insulation on the cable causing shorts .

How are the KBPWR and USBPWR jumpers set up ? Try to put them on the 5v rail if they are set up for 5V SB , the super I/O chip should get power also from the 5V SB rail , maybe you have a PSU with a weak 5VSB rail .

bogdanpaulb wrote on 2023-03-05, 20:23:

How are the KBPWR and USBPWR jumpers set up ? Try to put them on the 5v rail if they are set up for 5V SB , the super I/O chip should get power also from the 5V SB rail , maybe you have a PSU with a weak 5VSB rail .

They are in default position (which worked on the very same psu before the disaster).
kbpwr is in default position (enable).
usbpwr are all disable

Try with the kbpwr in the disable position .

bogdanpaulb wrote on 2023-03-05, 20:45:

Try with the kbpwr in the disable position .

Nothing changed unlucky

weedeewee wrote on 2023-03-05, 20:21:

Those things are diodes, and they are there because it was cheaper than putting on an actual 3v3 voltage regulator. The two dio […]
Show full quote

Those things are diodes, and they are there because it was cheaper than putting on an actual 3v3 voltage regulator.
The two diodes should be in series and they each have a voltage drop of 0.6~0.7V, combined 1.2~1.4V, which delivers a voltage of 5-1.4 = 3.6V
As a voltage dropper, diodes are the worst possible solution and should be avoided at all cost.

Luckily they should only be in use when the jumper for 3V3 is selected.
Though unluckily it seems that they somehow still connect to one pin on the main ide/ata connector from what I gather from the measurements you posted earlier.

killer keyboard ? might be as easy as failing insulation on the cable causing shorts .

The 5v option is selected so... A killer keyboard sounds very unlikely to me. Ok it's old, I've used it in many different configurations, but I didn't stretch or cut the cable, it's in perfect shape. The keyboard just miss the F key, that's the only defect other than being old, that being said with the multimeter I can measure the continuity between the pins or the resistance (or whatever is need).

Details ahead.
Following the advice of a fellow vogonians I conducted a visual and smell inspection.

No smell was noticed nor during the break event nor now, yeah I know when the stuff goes puff since some days ago I tried an hard drive who was in short, the smell was noticeable.

The visual inspection showed something very minor, imho, some pictures are as attachments others with links:

Just 2 noticeable things, which were present already when motherboard worked perfectly fine, I marked in red squares:
A cap near the ps2 ports has this weird thing that just seem dirt but it won't come off, all other caps are in perfect visual conditions, the coil near it has some black like burned "things":
https://i.imgur.com/XtFN6Ke.jpg

Some traces near the ide ports seems faded, imho is just the solder mask (they are not cut) also this was present when motherboard worked fine:
https://i.imgur.com/jQzeiq3.jpg
https://i.imgur.com/dZWmPJt.jpg
https://i.imgur.com/jBvvDw4.jpg

Bottom:

The only thing weird are those 2 little marks, they are barely noticeable but:
https://i.imgur.com/R4Ziloa.jpg

That's it, if you notice anything else from the picture I can take other details, my eyes are notorious to be tricky.

discolored enamel on the coil wire, might signify dried caps, might be nothing
but what I finally noticed is what the board is on, cartoon box is not a solid flat surface so every time you plug cards you stress the pcb and cracks might form
you can try measuring keyboard ps/2 port data clock pins for shorts (resistance to each other, ground, power)

rasz_pl wrote on 2023-03-06, 12:33:

discolored enamel on the coil wire, might signify dried caps, might be nothing
but what I finally noticed is what the board is on, cartoon box is not a solid flat surface so every time you plug cards you stress the pcb and cracks might form
you can try measuring keyboard ps/2 port data clock pins for shorts (resistance to each other, ground, power)

That's a point, which caps? The one near the coil wire?
The card board is full inside so it doesn't flex that much but it could be...
Measure... i'm sorry but about it i'm quite dumb, can you show me where I have to use the tips of the meter?

Nemo1985 wrote on 2023-03-06, 12:49:

That's a point, which caps? The one near the coil wire? The card board is full inside so it doesn't flex that much but it could […]
Show full quote

rasz_pl wrote on 2023-03-06, 12:33:

discolored enamel on the coil wire, might signify dried caps, might be nothing
but what I finally noticed is what the board is on, cartoon box is not a solid flat surface so every time you plug cards you stress the pcb and cracks might form
you can try measuring keyboard ps/2 port data clock pins for shorts (resistance to each other, ground, power)

That's a point, which caps? The one near the coil wire?
The card board is full inside so it doesn't flex that much but it could be...
Measure... i'm sorry but about it i'm quite dumb, can you show me where I have to use the tips of the meter?

These old boards are fragile so even light flexing can break solder joints, the issue is as soon as the pressure from the flexing is removed the joints go back to being closed.

Makes fault finding a painful process.

Nemo1985 wrote on 2023-03-06, 12:49:

That's a point, which caps? The one near the coil wire?

all behind the coil

Nemo1985 wrote on 2023-03-06, 12:49:

Measure... i'm sorry but about it i'm quite dumb, can you show me where I have to use the tips of the meter?

https://en.wikipedia.org/wiki/PS/2_port
pins 1 3 4 5
also pins 1 5 all the way to superIO chip pins

Might be the pic , but that trace looks interrupted by corrosion .

bogdanpaulb wrote on 2023-03-06, 14:25:

Might be the pic , but that trace looks interrupted by corrosion .

They certainly dont look right and since they look like they are going to and from the IDE sockets they might be worth checking for continuity.

did not read all , but did you change keyboard? seems a kb issue or at least kb input issue somewhere on the line of KB....