Thanks for replying; yes, it does appear to have signalling back to the p/s. I tried plugging in a standard high-end p/s that didn't have that connector and got the 1-5-4-4- beep sequence indicating a Power control fault (power good assertion timeout), so I suspect it is indeed required. I'm assuming Intel is the only one who makes the power supplies that are compatible with this m/b?
Trying to follow their documentation, I feel like a rat in a maze. From what I can gather, the AXX750WPS would work, but I was never able to see a sample picture of any connectors even coming off this p/s, so I'm at a loss on this one.
This stuff shouldn't have to be this difficult.
The good news is, there are both Intel and non-Intel chassis that work. See http://www.intel.com/support/motherboards/server/sb/CS-030487.htm
The AXX750WPS will work fine, but you can also get some weird signal bounces off front panel switches, that generate the 1-5-4-4.
The aux power connector is not required, but without a power supply that supports the PMBus commands via this connector you are giving up the ability to mointor the systems power condition and usage. Not a big deal for most folks.
The beep code your reporting indicates that the board told the power to turn on but it did not get power on all power rails. (No Power Good)
Most often this means you did not connect the dual - 2 X 4 - 12v power to the processors.
Here is the link to the boards TPS (Technicial Product Specification) http://download.intel.com/support/motherboards/server/s5520hc/sb/e39529013_s5520hc_s5500hcv_s5520hct_tps_1_8.pdf
This is the handbook. Everything you want to know and some you likly don't want to know.
Thanks much for the feedback. Sorry for the delay. In the interim I decided to pick up a non-Intel m/b for comparison. Same thing. So I bought a comparable p/s from another mfr; same thing. Haven't tried the new p/s with the Intel m/b yet. That's next.
This is pretty puzzling. I'm wondering if the processor might be bad. Unfortunately I don't have more than one L5520 lying around.
Hope to have these tests done later today and will post how that went.
And yes, I'd come across the technical manual. Reminds me of the mainframe Principles of Operation guide IBM put out. Way TMI...
The TPS is a good reference book, but not something I would recommend for a easy bed time reading.
Strange that you see the same type behavior with different mother board and different power supply.
The Intel board error is indicating that power good never comes up so that the board does not start fetching code from BIOS.
(Stop me if I get too detailed)
I would verify that all 3 power connections are connected and correctly wired ( the 2 -- 2x4 pin 12v and the main SSI power connector)
When power is present at all three, the mother board should report a "Power Good" to the processor and start reading the BIOS instructions.
You should see the row of POST LEDs start flashing at this point. If the LED stop, make note of the light pattern and the TPS should tell us were it is hanging up. ( I don;t think your even getting as far as the LED's flashing)
A bad processor is a possibility, but it would be low on my list.
I would expect
1) one of the 3 power connector is missing or wired wrong for the mother board.
2) A component(s) on the edge of the mother board at the end of the PCIe slots has been damaged inserting a card (usually video)
There are 2 or 3 little resistors very near the edge of the board, that I have seen users wipe off when seating PCIe cards. They connect the power good signals to the BMC.
3) some chassis mounting screw is shorting the mother board so that power does not come up correctly.
4) Bent pin in CPU socket
5) PCIe or DIMM socket shorting the mother board.
6) power supply
7) BIOS corruption ( not likely if it has never booted)
8) Add in card failure (try removing any cards)
9) Processor - since all else is not fixing it and you seem to have the same issue on a second board.