MSI 770T-C45 unstable

[alindumitru46]

Hi,
After long time I come back with the same old problems.
The mobo has a unstable behavior: with PCI-e videocards cannot boot (1 long 2 short beeps) or boots extremly rare. I put an very old PCI native videocard and now the mobo starts with fewer power-on issues.
I met some issues with all DDR slots used, however I use 1 o 2 blsck slots of RAM when I tested the mobo.
Like I said here and to the next post of @momaka: https://www.badcaps.net/forum/showpo...6&postcount=16, I changed the PSU with one very solid (Corsair VX450 powered by Seasonic platform), and I re-verified the entire number of mosfets.
There are 2 identical NIKOS P75ND2LDG mosfets (Q49 and Q47, in the DDR respectively chipset area) that give continuity, drawed in the photo.
There is my search for a replacement:http://alltransistors.com/mosfet/crs...TO220AB&page=0
Do you have additional advices?
Thanks.

[alindumitru46]

Re: MSI 770T-C45 unstable

Of course, great haste makes great waste...
The mosfets are NIKOS P75ND02LDG and the list of replacement is
http://alltransistors.com/mosfet/crs...007&caps=TO252

[cpt.charlie]

Re: MSI 770T-C45 unstable

I don't think those fets are shorted, when you have a short, the PSU will shut down immediately.
Check voltages and contacts at pci-e.

[alindumitru46]

Re: MSI 770T-C45 unstable

Before opening this discussion I washed carefully with isopropillic alcohol
all the ports, including PCI-e, by fine spraying followed by fine brushing.
With voltages... it's a little harder

[jazzie366]

Re: MSI 770T-C45 unstable

I don't think those fets are shorted, when you have a short, the PSU will shut down immediately.
Check voltages and contacts at pci-e.

This depends on how dramatic the short is. You can have a shorted mobo fet, if it's only a 6 ohm short, that'll draw a measly 2A from the PSU, the chip will light up red until the resistance rises and the mobo shuts down from a fault. However this could be a high resistance short, meaning it might only be dissipating 120mA of current, it'll get hot, but that's about it aside from mobo instability.

[alindumitru46]

Re: MSI 770T-C45 unstable

When the mobo boot, I can run a live Linux for hours...
I do not know anything about mosfets behavior in load, especially if they are out of specification.

[momaka]

Re: MSI 770T-C45 unstable

A MOSFET is either good or bad. When bad, the motherboard won't work at all. As such, intermittent issues are never caused by bad FETs.

You might want to move the video card around in the PCI-E slot. Sometimes a marginal connector will cause the video card to not get detected. Cleaning the slot may not always fix it. I have a finicky 939Dual-SATA2 that sometimes won't detect video cards on PCI-E. Fooled me quite a few times that I had dead GPUs when the motherboard was the issue the whole time.

The ATI 770 NB/chipset may also have BGA issues. I thought ATI made reliable chipsets, but apparently not.

[alindumitru46]

Re: MSI 770T-C45 unstable

I cleaned with a brush and isopropillic alcohol the mobo, already ...
Today I managed to power on the mobo on first attempt, and right now the prime5 runs for more than 1h and 30min and will run until it crashes...

LE: @momaka, I posted without seeing your message

[Agent24]

Re: MSI 770T-C45 unstable

I had intermittent boot problems on an MSI DKA790GX board, turned out the buck converter controllers for the NB and DRAM power supplies were bad.

[alindumitru46]

Re: MSI 770T-C45 unstable

Last night I finally decided to give a reboot and the board, of course, does not boot
Right now, after 12hours, the board runs fine .
@Agent24, can you be more explicit with these the buck converter controllers?
Where are they, how can I test / measure / check?
Thank you !

[Agent24]

Re: MSI 770T-C45 unstable

https://www.badcaps.net/forum/showthread.php?t=42510

They may be different for your board, they may not even be the problem. But if you can get the schematic you might be able to follow a similar troubleshooting approach.

[momaka]

Re: MSI 770T-C45 unstable

I had intermittent boot problems on an MSI DKA790GX board, turned out the buck converter controllers for the NB and DRAM power supplies were bad.

Reading through that thread again, I wouldn't call the issues there intermittent. More like you had a motherboard that did not boot at all and you got it to boot by replacing a bunch of parts.

When a motherboard boots and POSTs successfully only sometimes, that's an intermittent issue, as is the case with the O/Ps.

[cpt.charlie]

Re: MSI 770T-C45 unstable

Maybe a bad solder joint or even a BGA problem.

[Agent24]

Re: MSI 770T-C45 unstable

Reading through that thread again, I wouldn't call the issues there intermittent. More like you had a motherboard that did not boot at all and you got it to boot by replacing a bunch of parts.

When a motherboard boots and POSTs successfully only sometimes, that's an intermittent issue, as is the case with the O/Ps.

I seem to recall that before starting that thread and repair, the board might have worked again once or twice before dying completely. But I guess ultimately you are right. By the time I was seriously working on it, the whole thing was pretty dead.

Still, I can imagine it's plausible for a failing voltage rail to cause intermittent POST issues?

[Th3_uN1Qu3]

Re: MSI 770T-C45 unstable

I would go for failing northbridge BGA.

[alindumitru46]

Re: MSI 770T-C45 unstable

Trying a heating gun for "basic" reflow can be a solution ?!
Or is like a rubbing the wooden leg ?

[alindumitru46]

Re: MSI 770T-C45 unstable

But why Q49 and Q47 give continuity if there can be a northbridge issue?!

[Th3_uN1Qu3]

Re: MSI 770T-C45 unstable

Continuity as in a reading of zero, or continuity as in low voltage drop but enough to trigger the beeper? If it's the latter, it's not a short circuit, just the resistance of the circuit fed by those MOSFETs. If you really had a short the board wouldn't power up at all.

You can try using a heat gun and some flux, just be careful not to overheat it, and you should start by heating the underside of the board. A basic temperature probe is a very good idea. Lead-free reflow profiles you can find on the internet. Desktop boards tend to hold up better after reflow than laptop ones as the overall temperatures are lower and there is room for cooling improvements.

Or you can send it this way if you wish a pro to have a go at it. If you just want a reflow it's gonna be cheap, i promise.

[alindumitru46]

Re: MSI 770T-C45 unstable

Nice