GIGABYTE MB10 Datto Motherbaord Xeon D-1521 with I/O Shield - $130 + shipping

[RageBone]

@acquacow I don't see an option to disable the bmc on those boards.
reading and thinking and reading about it a bit more, none of the postcode make any sense.

The only reasonable thing i can come up with from my 04 postcode experience is that the board is waiting for the BMC to come up, and that somehow fails, so it waits for ever.

Reasons can be many, but i think that physical damage to the hardware could be a reason.
Did you visually inspect the board on all sides for scratches and broken off parts ?
Maybe you could upload high res pictures of the board? Can't have enough of those when searching possible failures.

 

[acquacow]

Yeah, I went over it with a good light and magnifying glass on both sides, didn't see any thing funky.

I then washed it down with 99% isopropyl and a light brush in case something small was bridging pins and causing an issue.

No luck.

I'd like to drop it into an ultrasonic, but I don't have access to one anymore.

One thing I did note. When I feed power to the board and it tries to post, if I hold the power button, it will shut off, the BMC will come up and I can access it via browser.

Even though the system is powered off, the CPU and the other heatsink on the board get fairly hot to the touch.

 

[RageBone]

@acquacow Well, shit. Would have been nice if it were something easy.
Do all Sensors in the ipmi report "everything OK"?
Another guess would be faulty oscillators.
The one on my SM, didn't output any clock, 0V on output, another defective one, outputted 3.3V.
The working one, 1.5V. You could grab a multi-meter and simply probe all those little buggers and if one doesn't have such a odd voltage, it might be of interest.


The BMC is powered of off 5V standby, to allow system-management even when the server is powered off.
Since the Xeon Ds don't have a external PCH, i assume they have something along that duty and feature-set inside, maybe also powered by 5vsb.
Networking? no clue why that should be supplied by 5vsb.

Another possibility is janky "powerstates" My MD70 Turns on the PSU even if it is "OFF", if you press the power button, all fans start spinning, but it does not Post jet.
If you again press the button, it then spins faster and turns on fully, posts and does it's thing.

 

[jahsoul]

yeah, that and a couple of Xeon E5-26xx v3 motherboards; which i'm currently hunting for CPU and RAM so that I can test it soon... :-(

Well, my motherboard came in and everybody checked out. Updated the BIOS and BMC with the software on the Asrock Rack site (with it no longer showing anything Datto) and changed the LSI firmware to IT mode.

 

[BLinux]

Well, my motherboard came in and everybody checked out. Updated the BIOS and BMC with the software on the Asrock Rack site (with it no longer showing anything Datto) and changed the LSI firmware to IT mode.

i was not so lucky. My D-1541 board was already damaged when I opened it up. :-(

the other motherboards physically seem okay, but waiting to receive CPUs to test them.

 

[jahsoul]

i was not so lucky. My D-1541 board was already damaged when I opened it up. :-(

the other motherboards physically seem okay, but waiting to receive CPUs to test them.

That's sucks. I was thinking to myself that even if it was not working, it was packed very well. You think it was a shipping issue?

 

[BLinux]

That's sucks. I was thinking to myself that even if it was not working, it was packed very well. You think it was a shipping issue?

No, I don't think it was a shipping issue. However, I don't necessarily think it was packed well. For one, the seller did not place the motherboard in an anti-static bag. they just wrapped it in some foam or bubble wrap. That "foam" / "bubble wrap" was then placed in a good sized box, filled with formed foam bits to hold it in place, which was good. I just wished they did place it a a proper anti-static bag. On the other motherboards however, they left the CPU sockets LGA2011 unprotected with the I/O shield loose on top; I got lucky and the I/O shield did not bend any pins that i can see.

The damage on the D-1541 was very obvious and bad. It broke off one capacitor and damaged the M.2 socket next to it. This can only be caused by a significant impact, but the foam around it did not show signs of such an impact, so I don't think this happened in shipping. I suspect this seller does not test their stuff, as they claim, but probably figures it's a gamble and they win most of the time and just take a few losses as the "cost of business". Or, perhaps they know some of this stuff doesn't work, but they claim shipping damage and use insurance claims to turn a dead motherboard into cash.

I can pretty much say that "techassetsales" eBay seller is on my list of "do not buy" sellers. There may be a good deal here and there, but it's also a time waste.

 

[RageBone]

@acquacow i was pointed a freezing stuff by a friend.
So no kidding, i suggest you put it into a freezer for a few hours.

Might not do anything at all, but it is worth a try if nothing helps.
I think its better then baking since well, baking is the worst idea one can have.

 

[dstanding]

@acquacow i was pointed a freezing stuff by a friend.
So no kidding, i suggest you put it into a freezer for a few hours.

Might not do anything at all, but it is worth a try if nothing helps.
I think its better then baking since well, baking is the worst idea one can have.

I've only heard of that working for spinning drives.

 

[RageBone]

@dstanding well yes, and for Ram and i think that it might be worth a try, since what else is left to try?

 

[acquacow]

@acquacow i was pointed a freezing stuff by a friend.
So no kidding, i suggest you put it into a freezer for a few hours.

Might not do anything at all, but it is worth a try if nothing helps.
I think its better then baking since well, baking is the worst idea one can have.

Actually baking it (at least to warm it up) and then using a heat-gun to re-flow some chips would probably be the best option.

If I still had access to a hot air workstation and an ultrasonic cleaner, I'd do exactly that.

 

[RageBone]

Actually baking it (at least to warm it up) and then using a heat-gun to re-flow some chips would probably be the best option.

If I still had access to a hot air workstation and an ultrasonic cleaner, I'd do exactly that.

@acquacow edited for simplicity and spelling.

Well I rather say " freeze it for a day" instead of bake it for 5 min.

If you have a reflow oven, please don't hold back, but in any other case, I simply wouldn't, to Mich to go wrong.

 

[hnamanh]

No, I don't think it was a shipping issue. However, I don't necessarily think it was packed well. For one, the seller did not place the motherboard in an anti-static bag. they just wrapped it in some foam or bubble wrap. That "foam" / "bubble wrap" was then placed in a good sized box, filled with formed foam bits to hold it in place, which was good. I just wished they did place it a a proper anti-static bag. On the other motherboards however, they left the CPU sockets LGA2011 unprotected with the I/O shield loose on top; I got lucky and the I/O shield did not bend any pins that i can see.

The damage on the D-1541 was very obvious and bad. It broke off one capacitor and damaged the M.2 socket next to it. This can only be caused by a significant impact, but the foam around it did not show signs of such an impact, so I don't think this happened in shipping. I suspect this seller does not test their stuff, as they claim, but probably figures it's a gamble and they win most of the time and just take a few losses as the "cost of business". Or, perhaps they know some of this stuff doesn't work, but they claim shipping damage and use insurance claims to turn a dead motherboard into cash.

I can pretty much say that "techassetsales" eBay seller is on my list of "do not buy" sellers. There may be a good deal here and there, but it's also a time waste.

I bought quite a few of those D-1541 from the seller and no damage on my boards. The only thing is that one of the board was missing a CPU fan.

Sorry to hear that your board was damaged, have you reach out to the seller and see what they can do?

 

[PnoT]

No, I don't think it was a shipping issue. However, I don't necessarily think it was packed well. For one, the seller did not place the motherboard in an anti-static bag. they just wrapped it in some foam or bubble wrap. That "foam" / "bubble wrap" was then placed in a good sized box, filled with formed foam bits to hold it in place, which was good. I just wished they did place it a a proper anti-static bag. On the other motherboards however, they left the CPU sockets LGA2011 unprotected with the I/O shield loose on top; I got lucky and the I/O shield did not bend any pins that i can see.

The damage on the D-1541 was very obvious and bad. It broke off one capacitor and damaged the M.2 socket next to it. This can only be caused by a significant impact, but the foam around it did not show signs of such an impact, so I don't think this happened in shipping. I suspect this seller does not test their stuff, as they claim, but probably figures it's a gamble and they win most of the time and just take a few losses as the "cost of business". Or, perhaps they know some of this stuff doesn't work, but they claim shipping damage and use insurance claims to turn a dead motherboard into cash.

I can pretty much say that "techassetsales" eBay seller is on my list of "do not buy" sellers. There may be a good deal here and there, but it's also a time waste.

I'm sorry to hear about your damaged 1541 as I have 4 that are being delivered today and will report back. The X9 motherboard came in flawless for myself and we just put another system together for a buddy and it worked right out of the box. The pink "wrapping" stuff is just dumb.. use static bags and be done with it.

 

[BLinux]

I bought quite a few of those D-1541 from the seller and no damage on my boards. The only thing is that one of the board was missing a CPU fan.

Sorry to hear that your board was damaged, have you reach out to the seller and see what they can do?

i'm not saying that this seller sells only broken motherboards, clearly, some have gotten working ones here. what i'm saying is that i highly doubt they actually test them and that their claim that they do is false. they are betting on a statistics game that the odds favor them; which perhaps is true and that the time/labor savings from not testing it is worth more than the number of bad transactions. if my suspicion is correct, i just wished they would state it as so: "product pulled from working environment, but untested" or some such; it would be more honest.

i just noticed the v3 board i got from them has broken SAS 8643 connectors... they yanked the cables out pretty hard without releasing the latch. i'm now trying to figure out if that is replaceable; but still waiting on the v3 CPUs to test the board itself.

 

[croakz]

My DATTO board seemed to be working fine, but then ran Memtest86 overnight and it stopped posting past the BMC initializing stage. VGA display just turns off. At least the seller gave me a quick refund.

 

[hnamanh]

My DATTO board seemed to be working fine, but then ran Memtest86 overnight and it stopped posting past the BMC initializing stage. VGA display just turns off. At least the seller gave me a quick refund.

Which Datto board are you referring to?

 

[croakz]

Which Datto board are you referring to?

GIGABYTE MB10 Datto Motherbaord Xeon D-1521.

 

[Dzeinz]

I just want to hop in and give my account of my bad boards. I ordered 2 and they both have the same symptoms. When power is applied, the fans spin and the BMC begins loading. The board is unresponsive to the power button (at any point). Fans kick up after the BMC is ready and then the amber light comes on indicating an error. All the while it says "waiting to for BMC to initialize".

The BMC lists an error with 0volts detected @ VCC_P0. So it's detecting 0 volts on processor 0 vcore. However, when using a multimeter, the voltage is there at 1.25V (good voltage). This tells me that there is a problem in the voltage detection. My guess would be that these boards all overheated and baked themselves inside their NAS cases. They likely need a CPU reflow.

I received a quick refund when I contacted the seller and I won't be sinking any more time into trying to make them work. If anyone wants to keep messing with these, I have 2 available for the price of shipping. PM me.

 

[acquacow]

Yeah, I had the same voltage errors in my error log after I cleared it and let it log new events. The previous errors had a lot of dead fan events.

Baking is probably a likely thing.