H12SSL-I Stuck at "bmc initiating"

[RolloZ170]

Hello, I have the same problem with the H12SSL-I motherboard.

afaik only one or more of the chips U4,U5,U6 are damaged.
U4 looks different.

 

[techmart]

afaik only one or more of the chips U4,U5,U6 are damaged.
U4 looks different.

Yes, no output on the three rails. L7 coil has very low resistance(10 Ohms), is that okay?? Btw, thanks for the quick reply.

 

[RolloZ170]

Yes, no output on the three rails. L7 coil has very low resistance(10 Ohms), is that okay??

if a chip is broken it can build wrong bypass resistances. if one VRM gives no output, the following is disabled.
RMA or replace yourself. i doubt any coil is dead.

 

[techmart]

if a chip is broken it can build wrong bypass resistances. if one VRM gives no output, the following is disabled.
RMA or replace yourself. i doubt any coil is dead.

I will try to remove the coil tomorrow and let you know whether the low resistance is caused by the IC itself or from the BMC side.

 

[RolloZ170]

I will try to remove the coil tomorrow and let you know whether the low resistance is caused by the IC itself or from the BMC side.

don't damage more than need. once you solder you can't RMA the board. correct in circuit measurement of a coil gives often wrong value, and you need henry, not ohm.

 

[thunderbolt93]

happend to me a well... Rev 1.01 board

killed the VRMs while installing a GPU
all three regulators output 0V

fans just go full tilt, no video, board doesn't seem to boot (is it supposed to boot anyways, even without BMC?)

started a RMA (and also created a support ticket, apparently chances of RMAs getting accepted is higher if you have a support ticket)

 

[RolloZ170]

is it supposed to boot anyways, even without BMC?

no. BMC must check BIOS integrity(RootofTrust) before.

 

[thunderbolt93]

If anyone has a 3D-Printer, I've designed a protective cover for the VRM area.

Use at your own risk, it might cause some thermal issues.

If anyone has suggestions for improvements (add venting holes?) feel free.

 

[RolloZ170]

Use at your own risk, it might cause some thermal issues.

i expect that too. even after move of the tiny vrms to the underside there where damages probably by overheating.
you can install the cover before handling and remove after ?

 

[thunderbolt93]

the idea was to protect the VRMs during PCIe card installation.

removing it after that is probably difficult because it's buried under the PCIe cards

making the cover from aluminum might be an option (but I don't have a CNC mill)
but a thermal pad might put too much pressure on the chips and crack them...

 

[RolloZ170]

but a thermal pad might put too much pressure on the chips and crack them...

there are very soft ones like chewing gum but expensive.

 

[thunderbolt93]

started a RMA (and also created a support ticket, apparently chances of RMAs getting accepted is higher if you have a support ticket)

Well, at least my RMA got accepted after telling them I'm willing to pay for shipping and repair.
First response was "we can't accept the RMA, the item is out of warranty", but the RMA wasn't closed.

 

[RolloZ170]

Well, at least my RMA got accepted after telling them I'm willing to pay for shipping and repair.

That's the least they can do, since they're the ones who messed this up.

 

[bards1888]

Fark. Just killed my rev 1.01 while installing new 10Gb NICs.... eek

Seems like rev 1.10 will be a little bit less of a risk ?

 

[RolloZ170]

Seems like rev 1.10 will be a little bit less of a risk ?

correct, the tiny VRMs moved to the back, but you can brick them of by installing into the chassis,
or they can overheat if not installed in server chassis (no airflow)

 

[bards1888]

correct, the tiny VRMs moved to the back, but you can brick them of by installing into the chassis,
or they can overheat if not installed in server chassis (no airflow)

Yeah, was thinking of giving the asrock rome board a go but alas, I have a supermicro chassis. I've ordered rev 1.10 but if anything goes wrong I'll probs ditch SM and move elsewhere.