Asus mainboard repair

[sunfire]

Hello guys,

I received a non-working mainboard Asus Crosshair III Extreme which doesn't start and I'm trying to repair it for myself. The mainboard seems decent and I've got it for free. When connected to power the onboard RGB leds and the OLED display do their "dance" but the mainboard doesn't react to the power button. I traced the power button and the PSU start pin to the Nuvoton NCT6798D super io chip. Unfortunately I couldn't find any datasheet for this super io. One big question which I have is the voltage across the crystal oscillator: when measured (while the mainboard is connected to power) with my Fluke 189 the crystal voltage reads ~12mv AC in one direction and ~38mv AC in the other direction. Doesn't this seem low ? I would have expected a few hundreds of mv at least. Unfortunately I don't have an oscilloscope (it would have sped up the investigation significantly). The super io chip has correct Vcc (3.3v) but other than that I have no idea how to test it. If the crystal is indeed bad what frequency does it have ? It's marked "15G61", I would guess 15MHz ? The crystal oscillator size is ~ 3mm x 2mm. I'm attaching pics taken with my phone of the super io chip and the crystal oscillator.

Help please ?

Big thank you !

PS: Electronics is my hobby, it's not my main profession. I have enough tools to repair most electronic things but unfortunately no oscilloscope. I know my way around electronic things but I'm far away from EE level. Also, English is not my native language but I think you guys understand me.

[fzabkar]

You won't see the crystal's AC voltage on your DMM. That's because your DMM's frequency response is much too low.

Measure the DC voltage between each end of the crystal and ground. If the crystal is oscillating, the DC bias should be about half the supply rail.

[sunfire]

You won't see the crystal's AC voltage on your DMM. That's because your DMM's frequency response is much too low.

Measure the DC voltage between each end of the crystal and ground. If the crystal is oscillating, the DC bias should be about half the supply rail.

I just did, at one end the DC voltage is ~1.8v, at the other end the DC voltage is ~2.15v. So the crystal seems to be oscillating then. I will investigate the auxiliary components around the super io chip, if those are fine then I need to see where I can source a super io chip for this board. As far as I know these super io chips require programming (with a programmer which I don't have).

Thank you fzabkar !

[Le_Bassiste]

whoah! bottom pin row in upper picture has the most shitty solder job i've ever seen on any mobo. 
you sure that's an Asus and not an A-sus?

edit: might easily be reparable by just resoldering, using a ton of flux.

[sunfire]

It must have been bad lighting when I took the picture with my phone. it was close to midnight and the angle of the photo does indeed hint at bad soldering of the bottom row pins.
But fear not, all the pins are soldered ok, I already tested each one of them before I posted. The soldering job is also Asus quality, it must be my crappy camera of the phone, sorry.

[Rasz]

Asus Crosshair III Extreme

no such thing? its either III Formula or VIII Extreme

[thm_w]

Watch some of this guys videos: https://www.youtube.com/c/adamantcomputers/videos

Usually he sees corrupted BIOS. I would start investigating there, if you can see no physical damage to the board.

Check if it supports USB reflashing.

I assume you put a working CPU in it?

[sunfire]

Asus Crosshair III Extreme

no such thing? its either III Formula or VIII Extreme

Sorry, my bad, it's a Crosshair VIII Extreme

[sunfire]

Watch some of this guys videos: https://www.youtube.com/c/adamantcomputers/videos

Usually he sees corrupted BIOS. I would start investigating there, if you can see no physical damage to the board.

Check if it supports USB reflashing.

I assume you put a working CPU in it?

It does support USB flashing (Asus calls the feature "USB Bios Flashback"), however, I'm not suspecting the BIOS to be an issue at the moment.
I don't have any Ryzen CPUs on hand, and I'm beginning to believe that's the reason why the mainboard doesn't start at all. I've seen mainboards with chipset X570 reported to refuse to even power on without a CPU installed. This Asus mainboard may be one of these boards. When the power supply is turned on there is a "CPU status" led which immediately lights up on the mainboard, this may be a clue that a CPU is actually required for the mainboard to start.

I did investigate deeper the super io chip and i haven't found anything wrong with it up to now. It turns out the super io chip is expecting some signals from the PCH before it turns on the power supply, and these signals from PCH are missing at the super io chip. The next step was to have a closer look at the PCH. At first glance I thought the PCH core is shorted (its resistance was less than 1 ohm which at first I thought it's too low). But supplying the PCH with a current limited external power supply showed the PCH doesn't trip the current limit and it doesn't get warm (I tested with 1v, the PCH takes 2.5a, doesn't trip the current limit on the external power supply).

For the moment I put the repair on hold while I look for a Ryzen CPU which is supported on this mainboard.

Thank you for the hints guys !

[rob77]

Nuvoton NCT6798D super io chip. Unfortunately I couldn't find any datasheet for this super io.

https://www.nuvoton.com/resource-files/NCT6796D_Datasheet_V0_6.pdf

literally among the first hits from google.

[sunfire]

Nuvoton NCT6798D super io chip. Unfortunately I couldn't find any datasheet for this super io.

https://www.nuvoton.com/resource-files/NCT6796D_Datasheet_V0_6.pdf

literally among the first hits from google.

Hello Rob, I know that datasheet, that's how i was able to figure out there are some missing signals from PCH before the super io actually turns the PSU on. But that datasheet is not 100% correct for NCT6798D (there are a few different pins between the two). As Google told me it looks like NCT6798D is custom made for Asus.

I haven't found a datasheet for NCT6798D, but for the moment I don't even need it anymore. I don't think the super io is bad, I only need a Ryzen CPU to test the board. I did replace an N-FET transistor which had the source-drain junction with very low resistance ( ~ 45 ohms, which according to the datasheet it's the Rds ON resistance ), but no matter how much I tried i couldn't turn the N-FET transistor off. It was pulling the enable pin for the buck converter generating the PCH core voltage low, I'll see if this was actually the problem when I get my hands on a Ryzen CPU supported by the board.