[SOLVED] Board fans start before power

[doenn]

Hi, I have a RM850x Corsair PSU, ASUS WS C621E SAGE lga3647 board, XEON Gold 6130F 16C processor, Micron 16GB DDR4 2666Mhz 288 pin ECC LRDIMM, and Intel heatsink with fan.

When I switch the PSU on, the fans in the case, on the PSU itself, on the Intel heatsink fan, and on PCIE cards if I install them start spinning. LEDs light up, including the double-digit LED, which reads 00, and after 30+ seconds, the BMC heartbeat LED blinks. Pressing the power button doesn't seem to do anything.

I've tried connecting the VGA adapter included by ASUS with

https://www.amazon.com/FOINNEX-Conv...s=vga+to+hdmi&qid=1552516392&s=gateway&sr=8-9

to an HDMI ViewSonic projector. I don't have a VGA monitor. I also tried a USB type C to HDMI connector with the built in USB type C port and using an AMD FirePro workstation GPU with DisplayPort to HDMI, but that didn't show anything either.

There were no beeps once the fans started spinning or the BMC heartbeat started. Every part seems to fit snugly but I'm confused why things start spinning once I switch the Corsair PSU to ON. Shouldn't I have to press power on first from the ASUS board? I've already bought several spare memory DIMMs of different sizes, heatsinks, board batteries, adapters, and even two other Workstation boards, but this is as far as I've gotten. I would really not like to buy anything else, but I've read the manuals and am trying to figure out the issues. For example, if the VGA to HDMI seems to be the issue, but there are doubts if the board is even on, it seems prudent to resolve the power on first, then buy a monitor if needed. Will probably get rid of a VGA monitor after anyway.

I will provide any more details if anyone has any ideas on how to troubleshoot this.

Have a good one.

 

[Andrei N]

you are describing a serious issue. fans starting up on themselves is never a good sign. the 00 code on asus is a motherboard issue most of the time. since you have another board you can test this out by swapping the boards.

 

[doenn]

Yes, I agree it seems bizarre. I tried this with a separate board and the same fans starting on their own problem occurred. On the other board, sound codes beeped indicating no memory was present. This was not the case and I tried different memory sizes in different slots.

This is exactly where I bought the processor:
https://www.ebay.com/itm/Intel-Xeon...e=STRK:MEBIDX:IT&_trksid=p2060353.m2749.l2649
which seems relisted here:
https://www.ebay.com/itm/Intel-Xeon...kylake-SP-135W-/163455052736?oid=113628341501

This is where I bought the heatsink & fan:
https://www.amazon.com/Intel-BXSTS300C-Solution-Heat-sink-Removable/dp/B077BHLPKM
though I have two other heatsinks and fans also. They each showed the same problem.

Memory (only install one at a time):
https://www.ebay.com/itm/Micron-16G...e=STRK:MEBIDX:IT&_trksid=p2060353.m2749.l2649
and
https://www.ebay.com/itm/Crucial-4G...e=STRK:MEBIDX:IT&_trksid=p2060353.m2749.l2649

PSU from BestBuy:
https://www.corsair.com/us/en/Categ...ply-Units-Advanced/RMx-Series/p/CP-9020093-NA

PCIE GPU, M.2 SSD also purchased, but I've tried with & without them and don't think those come into play yet...

I have two of these also:
https://www.ebay.com/itm/SUPERMICRO...e=STRK:MEBIDX:IT&_trksid=p2060353.m2749.l2649

 

[Andrei N]

Wow, that CPU does not exist:

So that CPU is an Xeon-1800 with an added plate to fit in the newer socket.
You got this : https://ark.intel.com/content/www/u...-gold-6130f-processor-22m-cache-2-10-ghz.html

converted to socket 3647, socket P

EDIT: Wrong link: https://ark.intel.com/content/www/u...rocessor-1-80-ghz-512k-cache-400-mhz-fsb.html

 

[doenn]

I agree, the CPU was the most suspect part on the list. It was very inexpensive at the Gold level. I thought maybe because the 1.8GHz frequency was lower and since it was a bit older and used, I may have just gotten a better price. The reason we have difficulty finding it, I think is because ES must mean Engineering Sample, which is a test version that can be incompatible. A separate eBay user posted the same chip for sale and listed incompatibilities with the ASUS board and many others. It was compatible with Supermicro's board that I had, but without the -O qualifier on the end, not sure if that makes a difference.

I reached out to the original seller to get clarification this morning. I also just purchased:
Intel Xeon 4114 Deca-core (10 Core) 2.20 GHz Processor - Socket 3647 - Retail Pack:
https://www.amazon.com/gp/product/B074G22N14/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&psc=1

Please do not worry any further, I think this must be the problem or ruled out as a problem. I'll re-test once the new chip arrives. Not sure if my higher frequency memory will support the Silver chip, so I got a slower 288 pin DIMM pair as well.

The original seller should really specify what ES means and instead of listing the condition as working, perhaps listing compatibility charts like the other seller. I'll revisit this later.


Take care.

 

[Andrei N]

I'm happy to help but since people will be looking at this thread I need to add that there is no such thing as an ' Engineering Sample' especially when talking about Intel or AMD trademark.
These people are nothing more than scammers. (Misrepresenting a product)
I know that you'll be able to find a video from Linus tech tips when he tests out a similar CPU bought from china, but even in his case the CPU was sold with the original name and a different socket, it was a mobile chip added on a casing that fits in a Desktop socket.

EDIT: Wrong link on my post above: https://ark.intel.com/content/www/u...rocessor-1-80-ghz-512k-cache-400-mhz-fsb.html