ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

[Biessea]

Hi there,

I'm a technician and a computer/retro computer lover.

I just bought a great Sabertooth motherboard (I really love the colours and the ASUS military scheme) to install a nice FX-8320 from AMD and make a gaming computer with a Geforce GTX 970.

For the first time I had problems with the system, but I solved and then correctly booting, I installed Windows 10 and then I started to overclock the system. I reached a stable configuration of 4,8GHZ with 1,475v.
Then I saved fan configuration, and other parameters, and saved to the overclock profile of ASUS bios tool.

But here we arrive at the problem: when I shut down the PC without disconnecting the cable from the power supply, the bios remember correctly the settings, when I decided to take apart the system and disconnect the cables, bios is like forgot all, like the EEPROM (Winbond W25Q64FVAIG) doesn't work as expected. Every time I boot after disconnecting the system from the power cable, it ask me to press F1 to recover bios setting.
Then I enter, I load the overclock profile and all goes nice.

So I decided to go deeper and try to find the problem to try to solve.

I take apart the bios chip (is not soldered, but it is inserted in a socket thanks God) and with the multimeter try to discover something. But the chip seems healty.
Then I reseat the bios chip on the socket and changed the battery with a new one. No improvements.
Then I put a wire from the positive of the battery to the BIOS 8th pin VCC to try to supply it without passing through the motherboard, assuming it has some broken track.
When I tested the VCC pin with a multimeter I can see that the voltage slowly decrease. Like the battery can mantain the correct power supply voltage.
Then I take out the bios from the motherboard socket and in that situation the VCC pin is exactly at the same voltage of the battery (around 2,95v.).

So, when I put the bios chip inside the socket, it's like that something goes wrong and that situation make the voltage fall down.

Now I put photos of the chip and the motherboard, I really hope someone can help me to solve.

I am worried that it's nothing to solve anyway, I think the motherboard socket of the bios is shorted, or some component is broke.

I really hope someone of you here can give me an idea.

PS: I just bought a bios chip new from an holland ebayer, for my 990FX REV.2 to make another try.

Thanks a lot guys,

Loris.

[Bozog]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

I'm certainly no expert on this but the bios chip is not powered by the battery, nor are your BIOS(CMOS) settings stored here. CMOS settings along with RTC functions are stored on the southbridge chip, it is this that is powered by the battery.

Now I hate to state the obvious, but have to tried replacing the battery, 2.95v is low. Failing that I would measure the current draw from the battery when the board is unpowered, not sure what it should be but I would expect less than 30 microamps.

[Bozog]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

Shorting the battery positive to BIOS pin 8 will power +3VSB from the battery and as you have found out the battery will not like that too much.

[Biessea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

Shorting the battery positive to BIOS pin 8 will power +3VSB from the battery and as you have found out the battery will not like that too much.

Ok I have understood.

But, if the BIOS datas is stored on the southbridge (SB950), what is the purpose of that chip that is inserted on that socket?

The Winbond chip I said.

Have you any ideas about what to do?

I have tried before to change the batteries with a new 3,1v. , but nothing changed.

[Bozog]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

The bios chip contains the bios firmware.

Do you have a microamp meter to measure the drain on the battery?

[Biessea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

The bios chip contains the bios firmware.

Do you have a microamp meter to measure the drain on the battery?

I have a Fluke multimeter, I don't think I have the current measurement.

Anyway I am a bit surprised, I always thought that the little bios chip (in my case the Winbond chip) is a EEPROM that store all the modify I do on a motherboard like voltage profiles, fan profiles, boot sequences and other.
I really didn't know that these datas is stored on the southbridge.

So what do you think can be the problem that every time I disconnect the power cord I lose all the settings?

[Biessea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

The bios chip contains the bios firmware.

Do you have a microamp meter to measure the drain on the battery?

Anyway, when I measure the voltage of battery without the power cable, it is stable to 2,95v. (it is not low for a computer battery, they say that it works correctly until 2,4-2,5v).

I measured that power drain from the battery only when i shorted the positive battery with the 8th pin (vcc) of the Winbond Bios Chip.

When I connect the power cable and regularily switch on the power supply I can measure 3,3v. On that 8th pin of the bios chip anyway. So it has to be supplied with power thar Winbond chip.

[piernov]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

SPI ROM does store BIOS settings too, however if an RTC reset happens, BIOS will also reset settings to default.
RTC is part of the southbridge.
RTC reset happens when the RTC/CMOS jumper is triggered, or when the southbridge loses power on the RTC power well.
RTC power rail is generated either from 3VSB (itself generated from 5VSB of the PSU) when the computer is plugged into mains, or from the coin cell battery when the computer is unplugged from mains. A dual diode is used to combine these 2 sources.

EDIT: SPI ROM is not powered from the RTC power rail, it never takes power from the coin cell battery. (it does not need to since it's a non-volatile memory)

[Biessea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

SPI ROM does store BIOS settings too, however if an RTC reset happens, BIOS will also reset settings to default.
RTC is part of the southbridge.
RTC reset happens when the RTC/CMOS jumper is triggered, or when the southbridge loses power on the RTC power well.
RTC power rail is generated either from 3VSB (itself generated from 5VSB of the PSU) when the computer is plugged into mains, or from the coin cell battery when the computer is unplugged from mains. A dual diode is used to combine these 2 sources.

EDIT: SPI ROM is not powered from the RTC power rail, it never takes power from the coin cell battery. (it does not need to since it's a non-volatile memory)

Thanks for the answer, clear as water.

Anyway now, what can I do to understand why my computer reset bios datas every time I disconnect the power cord?

What other check can I do?

[Biessea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

Anyway

"The BIOS settings are stored in the CMOS chip (which is kept powered up via the battery on the motherboard). That is why the BIOS is reset when you remove the battery and re-attach it. The same program runs, but the settings are defaulted."

I never heard before that the bios setting is stored on the southbridge.

Here they say that is it stored inside cmos memory.

[foshland]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

Hi, from my experience a 3V on battery is low and I have motherboards that will reset to defaults when its 3V or under.

Did you try to change the battery to a newer than 3.1V one? maybe that one goes under 3V when powered off

[RJARRRPCGP]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

But here we arrive at the problem: when I shut down the PC without disconnecting the cable from the power supply, the bios remember correctly the settings, when I decided to take apart the system and disconnect the cables, bios is like forgot all, like the EEPROM (Winbond W25Q64FVAIG) doesn't work as expected. Every time I boot after disconnecting the system from the power cable, it ask me to press F1 to recover bios setting.

I never witnessed this wacky-ness with a UEFI-BIOS, but this will happen with legacy AMI BIOSes on Asus motherboards when a UEFI-era video card is installed, such as installing a Radeon RX 580 on an Asus socket 1366 motherboard.

That BS happens every time I install my XFX Radeon RX 580 on my Asus P6T6 WS Revolution motherboard.

[SMDFlea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

Theres another chip above the winbond bios , BIOS AI1314-A2 .Thats for bios flashback , you can flash a bios from USB without a CPU installed.

[Biessea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

I never witnessed this wacky-ness with a UEFI-BIOS, but this will happen with legacy AMI BIOSes on Asus motherboards when a UEFI-era video card is installed, such as installing a Radeon RX 580 on an Asus socket 1366 motherboard.

That BS happens every time I install my XFX Radeon RX 580 on my Asus P6T6 WS Revolution motherboard.

Thanks for the reply, but this is not my problem; I was using a Nvidia 9800GT. So old video card and old bios.

[Biessea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

Theres another chip above the winbond bios , BIOS AI1314-A2 .Thats for bios flashback , you can flash a bios from USB without a CPU installed.

Do you think problems are that red circle you made?

Miss a resistance and a pin shorted?

[SMDFlea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

Do you think problems are that red circle you made?

Miss a resistance and a pin shorted?

I forgot to ask last night when i posted.Is the resistor damaged ? What is that brown thing thats circled

[Biessea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

I forgot to ask last night when i posted.Is the resistor damaged ? What is that brown thing thats circled

I don't think so.

I just made other photos after cleaning with a cotton.

Take a look.

[Biessea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

So, no other idea here.

I don't know how to proceed.

[SMDFlea]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

So, no other idea here.

I don't know how to proceed.

Try a bios recovery while you`re waiting for the new chip to arrive.If you can trace pin 8 of the bios or the + pin of the CMOS battery to the dual diode you can test that as well

[RJARRRPCGP]

Re: ASUS Sabertooth 990FX REV.2 bios chip problems, probably socket shorted?

Thanks for the reply, but this is not my problem; I was using a Nvidia 9800GT. So old video card and old bios.

In that case, then you need CSM, which is enabled by default, IIRC.

Because there's no way a video card that old, has GOP. (graphics output protocol, a standard for UEFI)