Replacing OST caps on Microstar KT6 Delta motherboard

[socketa]

Re: Replacing OST caps on Microstar KT6 Delta motherboard

Replaced ATTP1 thermal protection chip (had higher than normal resistance between VREF and ground pins), and the CPU fan now spins after pushing the power button; but don't have any BIOS activity (checked with a scope)
There is 3.3V on VCC and 3.3V on OE# pin,
OE pin is connected to R686 (4.7K ohms)
but no voltage on CE# pin, and no resistance to ground either, and can't find where it goes to (probed all of the nearby components).
5.1V on all pins of RN91
(i don't have a schematic for this version board - the schematic for the "KT6 Delta rev 100" has major differences from the ver 2.0 board that i have - even the BIOS chip has a different board identifier "U32")

Comparing to a similar BIOS chip on another board:
The chip on another board has 11Kohms resistance to ground on OE and CE pins
5.1V on Vcc OE# and CE#

Maybe there is an internal layer that is damaged at the area of the burned VRM SMD cap?

Am tempted to just run 3 jumpers from RN91 (located just under the CPU 4-pin socket) to VCC, ROMCS# and MEMR# in accordance with the schematic, and see what happens; but i think that it might destroy the BIOS.

https://www.badcaps.net/forum/attach...1&d=1662632095
https://www.badcaps.net/forum/attach...1&d=1662632291
https://www.badcaps.net/forum/attach...1&d=1662632095
https://www.badcaps.net/forum/attach...1&d=1662632095

[socketa]

Re: Replacing OST caps on Microstar KT6 Delta motherboard

I removed the BIOS chip and noticed that the chip enable pin was soldered to a pad that went nowhere.
Since CE is 0v (Low) i'm assuming that that's OK, and that it's not being utilized;
but the Output Enable pin doesn't go low, so no data is read (and no codes appear on the PCI analyser card, although the voltages, and CLK, LEDs are lit up.
"The read operation of the W49F002U is controlled by CE and OE, both of which have to be low for
the host to obtain data from the outputs."
Could that mean that the BIOS chip needs to be reprogrammed?
I'm also thinking that the 3.3V that is on the Vdd pin is OK, because it's directly connected to the 3.3V rail, even though the data sheet says that it's a 5V chip
Thoughts?

[socketa]

Re: Replacing OST caps on Microstar KT6 Delta motherboard

Well, i was looking at the wrong datasheet for the flash chip
It's the same package type, so i assumed that it was the same flash chip as in the schematic.
The flash chip is an SST49LF040
So i compared it with another exact same chip that's on a good board, [Standby power only/PSU turned on]

--------Faulty board--------Good board
OE---0V/3.3V---------------1V/3.3V
Vdd---0V/3.3V--------------1V/3.3V
WE---0V/3.3V--------- ----1V/Continuous signal from 0V to 3.3V
Clock signal is on both pins of LCLK, on both chips

Does it look like there is something wrong with flash chip?

[socketa]

Well, i finally reflashed the BIOS and put it back on the board.
Turned on the power ,and no response whatsoever.
The CPU VRM was running before i removed the chip, but there was no code on the PCI analyser card.
But, as i moved the board, it started up, and halted after the second analyzer code
So that was a good sign
Then i noticed that i could cut power by flexing the board
After some more flexing, it went through the entire POST sequence, and the BIOS screen appeared. Yay! 😀
So, after many years, it appears this board is finally fixed (well, close enough)
Looks like the original power surge took out that VRM SMD cap and also corrupted the BIOS

I had a look at the CPU socket solder joints on the backside of the board, they all appear to be ok
Reinserted the CPU, but the board still turns off (or on) when i flex it

[socketa]

Now it's fixed
After fixing this nearly soldered joint

[Agent24]

Congratulations!

You will get better solder joints on that IC if you wick the old solder off the pads after removing the old chip. If the pads are all flat, the IC will sit flat on the board as it's supposed to, and you can solder the new one much easier. Some good liquid flux helps a lot, too.

[socketa]

Yeah, it's about time,
I thought that there might have being a problem with the CMOS, since it didn't automatically identify the CPU correctly, but when i changed the FSB clock setting to what is was meant to be, then it identified it correctly
The only problem is that it's not seeing the SATA ports, despite being enabled (or disabled) in the CMOS
According to the CMOS setting, "PATA-IDE" and "SATA-IDE" are both controlled by the same VT8237 chip
The IDE port works OK, but there's no SATA in the first CMOS section "Standard CMOS Features" that shows the Hard drives, nor do they show up in Windows

[socketa]

Now it's all good.
I tried two SATA drives one of them was a DVD drive, and neither of them were detected
After a bit of frustration, decided to try a spare laptop drive, and it was recognized
All that is left to do is to replace the initial VRM SMD capacitor back on to the tiny bit of copper that remains

One thing that i found out is that i couldn't MS-DOS USB boot with later versions of Rufus
Which made me think, for a while, that there was a problem with the motherboard USB

[Agent24]

As I recall, one of those early VIA chipsets, possibly yours, won't detect any SATA device above 1.5g speeds, you have to use older drives or limit them with a jumper. I think it only applies to the "A" revision of the chipset. "B" version shouldn't have the problem.

[socketa]

Yep,
Putting the jumper on worked
Thanks