Highly overvolted z77 extreme6

[Wingklip]

Hi guys, first post.

I bought a bent pin z77 Extreme 6 from a guy with one pin bent. I managed to fix that, but it seems that there was more to it than meets the eye. After ten attempts booting with the CPU in socket, I tried to reset bios, but at the end I found that my led lighting on the switches stopped working. The error codes were reading 10, 15, then shorted off. The last ones only showed 00, despite booting fine.

Looking over the board I found that one of the traces was actually burnt up completely between the ram slots. I had a look there and ran a bus wire to the burn points, and to my surprise, it actually booted. But at this point the bios codes and lights were no longer working at all, even before this fix.

However, much like another thread for the extreme 4, this is a case of 'touching transistor 9T16GH feels super hot' while the one right next to it is cold as a stone idle. Looking in the bios, and confirming with a multimeter, the immediate voltage readouts to ground are 13v, 3.5v, 5.5v, and in bios reads 14v, 4.5v, and 6v, with CPU on 2v vcore. Luckily it seems the Vcore voltage is only 1.3 or so on the non-k overclock setting measured from the inductors to ground. All the fanspeed sensors are bugged out to max as well. Very odd sight to behold.

While the audio and usb chips also get hot, I'm fairly certain that this overvolting would be causing such a thing. The cap next to that transistor also gets fairly hot. If anything, I'm taking a limb out and guessing that the transistor would be the biggest culprit here? It's the one right in-between the two above the Southbridge. 5 caps below them.

I haven't seen the ability for ALL the power rails to go that far out of spec at once (tried 2 working PSUs, antec and cougar 80+ new), so I'm assuming somewhere there is a out of spec chip to the voltage regulation circuit. Could it be caused by that one transistor (or maybe it's inline with it?)

Where should I begin getting test point readouts?

Any ideas before it fries itself from overvoltage? Thanks for the read and help :')

[dicky96]

Re: Highly overvolted z77 extreme6

If a capacitor next to the transistor is getting hot - and it is actually getting hot, not just because some very nearby component is also getting hot - then pretty much for sure that capacitor is short.

It's a bit hard for me to tell from your description. Is the hot capacitor extremely near to the hot transistor?


Also - are you sure your multimeter is actually reading voltages properly? What do you read on the CMOS battery with the PSU off?

[Wingklip]

Re: Highly overvolted z77 extreme6

CMOS battery is reading 3.2v, so I'd assume as expected, right? The transistor heats up first but then the capacitor gets hot after. In the second image it's the second cap from the left

[Wingklip]

Re: Highly overvolted z77 extreme6

Also yes, the cap is right next to it. Multimeter works 100%, used it since forever

[Wingklip]

Re: Highly overvolted z77 extreme6

New intel lads, got some of the readings off the bios. VERY out of spec, aka +1v on the 5v rail, +1.5 on 12v, +0.7 on 3.3v, holy heck.

Once again I have the sneaking sus that the transistor is the cause. It shares the same 2 bus in and outs with the transistor next to it... But only that transistor gets hot. The capacitor gets hot only several minutes after. I've got a macro of this now with a label, have a look

[dicky96]

Re: Highly overvolted z77 extreme6

But correct me if I am wrong - the 12V 5V and 3.3V come from the ATX PSU. But you say you have tried two different PSU with the same results? What do you measure from ground on the ATX connector to 3.3V 5V 12V on the same connector? And also on EATX12V connector.

[momaka]

Re: Highly overvolted z77 extreme6

The caps and MOSFETs you show in this picture appear to be all liner regulator circuits. So knowing that, I would suggest to remove the cap that gets hot and just replace it anyways. You don't need to use an expensive polymer capacitor here or even that great of a low-ESR electrolytic, because linear regulator circuits don't really care. Just match or use slightly higher capacity. If the MOSFET still gets hot, then it could be the MOSFET that is bad... though IMO, I suspect the issue would be more likely to be a component down the line that is powered from the voltage rail that this MOSFET generates. In that case, you'll have to find what component is causing that. The fact that your fan RPM monitors all ready at max value and voltage are off suggests the Vcc for the monitoring chip (SuperIO / LPC, most of the time) may be noisy... which again could be due to that hot cap being bad or a another failed component drawing excessive current and making the MOSFET in the linear regulator circuit not regulate properly.

On a different note, also check your multimeter batteries. Low batteries in a multimeter can make it read voltages higher than they are. 3.2V for the CMOS battery is not too high, but still a bit higher than normal (most of my CR2032 read 3.0 to 3.1V) So just as a sanity check, once you verify the batteries in your multimeter are good / new, check voltages directly form the PSU at the 24-pin ATX cable with the motherboard turned On. Then report back what voltages you get here, so we can see how far the motherboard is off. While at it, also check the voltage on the Drain and Source of that MOSFET that gets hot.

[Wingklip]

Re: Highly overvolted z77 extreme6

The caps and MOSFETs you show in this picture appear to be all liner regulator circuits. So knowing that, I would suggest to remove the cap that gets hot and just replace it anyways. You don't need to use an expensive polymer capacitor here or even that great of a low-ESR electrolytic, because linear regulator circuits don't really care. Just match or use slightly higher capacity. If the MOSFET still gets hot, then it could be the MOSFET that is bad... though IMO, I suspect the issue would be more likely to be a component down the line that is powered from the voltage rail that this MOSFET generates. In that case, you'll have to find what component is causing that. The fact that your fan RPM monitors all ready at max value and voltage are off suggests the Vcc for the monitoring chip (SuperIO / LPC, most of the time) may be noisy... which again could be due to that hot cap being bad or a another failed component drawing excessive current and making the MOSFET in the linear regulator circuit not regulate properly.

On a different note, also check your multimeter batteries. Low batteries in a multimeter can make it read voltages higher than they are. 3.2V for the CMOS battery is not too high, but still a bit higher than normal (most of my CR2032 read 3.0 to 3.1V) So just as a sanity check, once you verify the batteries in your multimeter are good / new, check voltages directly form the PSU at the 24-pin ATX cable with the motherboard turned On. Then report back what voltages you get here, so we can see how far the motherboard is off. While at it, also check the voltage on the Drain and Source of that MOSFET that gets hot.

No particularly clicking voltage comes to mind when I measure the MOSFET voltage, just that it's about a volt from tab to drain or gate something like that. But one thing to note is that it's parallel to another MOSFET. Shouldn't the MOSFETs share load with each other, to not have one run super hot? That's my take. I've just gotten new batteries and will check in the morning

PSA: solder a fat 50v 860uf cap onto your 12v CPU line to ground and overclock to the sun (Asus P877V)

[momaka]

Re: Highly overvolted z77 extreme6

when I measure the MOSFET voltage, just that it's about a volt from tab to drain or gate something like that.

Well, that's not exactly technical "info" we can use.

Going to need more solid measurements and write things down. On that note, DON'T check voltages between pins on the MOSFET. DO check voltages with respect to ground (that is, black multimeter lead connected to ground or chassis of the PSU.) Then just check voltage on the tab (Drain) of these MOSFETs and also their right-side leg, which should be the Source.

But one thing to note is that it's parallel to another MOSFET. Shouldn't the MOSFETs share load with each other, to not have one run super hot?

It depends on the circuit.
I can't quite 100% tell if they are in parallel or not. But if their Source pins are clearly tied together to the same trace/pad and their tabs are tied together to another trace, then they likely are in parallel... and if one is running very hot and the other is not, then the hot MOSFET is probably bad.

[Wingklip]

Re: Highly overvolted z77 extreme6

Well, that's not exactly technical "info" we can use.

Going to need more solid measurements and write things down. On that note, DON'T check voltages between pins on the MOSFET. DO check voltages with respect to ground (that is, black multimeter lead connected to ground or chassis of the PSU.) Then just check voltage on the tab (Drain) of these MOSFETs and also their right-side leg, which should be the Source.


It depends on the circuit.
I can't quite 100% tell if they are in parallel or not. But if their Source pins are clearly tied together to the same trace/pad and their tabs are tied together to another trace, then they likely are in parallel... and if one is running very hot and the other is not, then the hot MOSFET is probably bad.

Alright, back with some more intel. Been measuring the legs of the main motherboard connector and all seems within range now. Multimeter voltage was far too low.

It appears like some voltage is probably too low to the monitoring IC if this is like a multimeter in the motherboard. They are most definitely parallel MOSFETs, but the caps aren't. I'll measure more again tomorrow but rn work is rek. From tab to either pin is the same for both Mos, 0.5 or 1.5v + or -

If this is in fact the case is there a way to check that IC? Does anyone know a way to check intel z77 voltage/general monitoring IC? Or where it generally might be would help :

[dicky96]

Re: Highly overvolted z77 extreme6

Hi Wingklip
Nice to see your multimeter is reading voltages correctly now. You must understand, no matter what over voltage you think has been set in the BIOS, it can in no way ever affect the voltages coming out of the ATX power supply!

Regards the IC you are trying to check.

First, like Momaka asked, please connect the negative (black) lead of your multimeter to ground. An easy way to do that is poke it into the ATX power connector on any of the black wires. Now leave it there.

1. Power the board on and measure the voltage on each pin/tab of the hot components with the red lead of your multimeter.

2. Power the board off, set your meter to resistance, and measure the resistance from each pin/tab with the red lead of your multimeter.

If you can, post a photo of that part of the board with the voltage and resistance measurements shown - and also can you tell us the part number (markings) on that component

With a bit of luck that will tell us all something useful..